Obstacle avoidance mobile charging scheduling algorithm based on fresnel model

Abstract

The application discloses a kind of based on Fresnel model's obstacle avoidance mobile charging scheduling algorithm, belong to wireless rechargeable sensor network technical field.The application is established by wireless rechargeable sensor network model, and dynamic planning charging path is combined with improved algorithm.Specifically, the charging signal enhancement zone, normal zone and weakening zone are divided by Fresnel diffraction model, and the charging power is optimized;The initial charging path is generated by simulated annealing, and the starting point and heuristic function are dynamically updated when obstacles are detected, and the path is corrected in real time;At the same time, the path smoothing processing is introduced to reduce the redundant inflection point and reduce the mobile energy consumption.The application significantly improves the charging efficiency and network life cycle, uses obstacles to enhance signals rather than simply avoid, achieves the balance of high charging utility and low energy consumption, and effectively adapts to real-time changes in complex environment through dynamic obstacle avoidance and local re-planning capability.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of wireless rechargeable sensor networks, and particularly relates to an obstacle-avoiding mobile charging scheduling algorithm based on a Fresnel model. BACKGROUND

[0002] Obstacles can affect the propagation of wireless charging signals. When a mobile charger (MC) wirelessly transmits energy to a sensor node through WPT technology, if the wireless charging signal encounters an obstacle during propagation, part of the signal will be reflected, diffracted, scattered, etc., so that it cannot penetrate, and the strength of the charging signal may change, affecting the charging effect.

[0003] In practical applications, there are usually many obstacles in the coverage area of a wireless rechargeable sensor network (WRSN), causing instability of the charging signal. If the influence of obstacles on the charging process is not specifically analyzed, the applicability of the WRSN will be greatly reduced, thereby hindering its development and practical application.

[0004] Therefore, how to propose an obstacle-avoiding mobile charging scheduling algorithm based on a Fresnel model, describe the influence of obstacles on the charging process according to the actual situation, and at the same time focus on planning the obstacle-avoiding path of the MC to improve the utilization rate of limited energy, is a problem that those skilled in the art need to solve. SUMMARY

[0005] Therefore, the application provides an obstacle-avoiding mobile charging scheduling algorithm based on a Fresnel model, which improves the charging effect and realizes global planning and dynamic obstacle avoidance of the charging path in combination with multiple algorithms.

[0006] In order to achieve the above purpose, the application adopts the following technical solutions:

[0007] The application provides an obstacle-avoiding mobile charging scheduling algorithm based on a Fresnel model, which includes the following steps:

[0008] A wireless rechargeable sensor network model is established based on the positions of sensor nodes, a mobile charger and obstacles;

[0009] In the wireless rechargeable sensor network model, a charging optimization objective function is constructed according to the effective energy received by the sensor nodes from the mobile charger, and a charging plan of the mobile charger is planned with the limited energy of the mobile charger as a constraint;

[0010] The charging area of the mobile charger for the sensor nodes is discretized, and the charging sequence and sub-path length of the to-be-charged nodes are obtained by using the algorithm with the goal of maximizing the effective amount of charging;

[0011] designing a charging position optimization target, determining an optimal charging position of each sensor node according to the charging position optimization target;

[0012] According to the optimal charging position, an initial charging path is obtained with the goal of minimizing the number of dead nodes; the mobile charger charges according to the charging plan along the initial charging path, and if an unknown obstacle is detected during movement, the initial charging path is corrected by using an improved algorithm.

[0013] Preferably, a wireless rechargeable sensor network model is established based on the sensor nodes, the mobile charger and the obstacle positions, comprising:

[0014] A two-dimensional network area is constructed, the two-dimensional network area contains a plurality of fixed known obstacles, the sensor nodes are randomly distributed in the obstacle-free area of the two-dimensional network area, and the base station is deployed at the center of the two-dimensional network area;

[0015] In the two-dimensional network area, n concentric ellipses corresponding to n Fresnel zones are constructed with the position coordinates of the mobile charger and the sensor nodes as the focus;

[0016] Based on the Fresnel zone, the charging signal influence area of different charging positions of the mobile charger is constructed, which is divided into an enhanced zone, a normal zone and a weakened zone; the enhanced zone, the normal zone and the weakened zone are drawn on the concentric circles with the sensor nodes as the center; wherein, a reference circle is made with the obstacle as the center, the boundaries of the enhanced zone and the weakened zone are determined according to the boundary of the first Fresnel zone and the reference circle, and the boundaries of the enhanced zone and the normal zone are determined according to the boundary of half of the second Fresnel zone and the reference circle;

[0017] In different charging signal influence areas, the i-th sensor node s i The charging power at the i-th candidate charging position l i is as follows: p l i , s i

[0018]

[0019] wherein, is a constant greater than 1; is a constant less than 1; α is a hardware influence parameter; β is an environmental influence parameter; is the distance between the i-th sensor node s i and the candidate charging position l ​​​i The distance between them.

[0020] Preferably, the charging optimization objective function is as follows:

[0021] ;

[0022] In the formula, U i For sensor nodes s i At candidate charging locations l i The effective energy received from the mobile charger , The rechargeable battery capacity of the sensor node. For mobile chargers to sensor nodes s i The energy transmitted , For mobile chargers at candidate charging locations l i Charging time;

[0023] Constrained by the limited energy of a mobile charger, including:

[0024] ;

[0025] In the formula, M represents the number of sensor nodes; The maximum energy that a mobile charger can use for wireless charging; e i For mobile chargers at candidate charging locations l i Give sensor nodes s i The energy sent; The initial charging path length of the mobile charger. Battery energy threshold for portable chargers that can be used for mobile purposes E move Maximum driving distance of the mobile charger under drive; For the candidate charging location set; This is the set of locations to be charged.

[0026] Preferably, the mobile charger discretizes the charging area of ​​the sensor node, aiming to maximize the effective charging amount, and utilizes... The algorithm obtains the charging order of the nodes to be charged and the sub-path length, including:

[0027] For each sensor node, with sensor node s i Concentric circles drawn with as the center are RA radius is Concentric circles, such that the difference in charging power between adjacent concentric circles does not exceed a threshold. The charging power of the same ring is equal, and each ring is equivalent to a charging power layer; the first j The radii of the concentric circles are: ;

[0028] Minimum charging power within each charging power level This refers to the charging power of the entire charging power layer; among which... D represents the maximum effective charging coverage distance of the mobile charger;

[0029] Distance sensor node s i The nearest known obstacle is abstracted as a grid, with the grid's circumcircle serving as the reference circle;

[0030] The enhancement region of the charging power layer is determined based on the reference circle and Fresnel zone;

[0031] according to The algorithm iterates through all possible mobile charger locations within the discretized charging area, aiming to maximize the effective charging amount, and obtains the charging sequence and sub-path length of the nodes to be charged.

[0032] Preferably, the objective function for optimizing the charging location is as follows:

[0033] ;

[0034] For mobile chargers located at candidate charging locations l i Time sensor node s i Received charging power; For mobile chargers located at candidate charging locations l i The distance between the current node and the next candidate charging position in the charging sequence. The weighting coefficients are used to balance charging power and mobility costs;

[0035] Each sensor node determines the optimal charging position from all candidate charging positions in the first charging power layer signal enhancement region according to the objective function of the charging position optimization objective, and outputs the optimal charging positions of all sensor nodes to form a set of positions to be charged. .

[0036] Preferably, based on the optimal charging position, and with the goal of minimizing the number of dead nodes, the following method is employed: The SA generates the initial charging path.

[0037] Preferably, if an unknown obstacle is detected during movement, an improved method is adopted. The algorithm corrects the initial charging path, including:

[0038] Improve the heuristic function by setting the starting point of the improved heuristic function. It updates in real time with the current location of the mobile charger;

[0039] The improved heuristic function is as follows:

[0040] ;

[0041] ;

[0042] In the formula, x , y Indicates the current node s The x and y coordinates, , Indicates the starting point The x and y coordinates, , Represents the target node The horizontal and vertical coordinates; For weighting functions;

[0043] Generate the current node s child nodes Based on the current node s To the target node The actual cost g ( s ) and from the current node s Through child nodes Reach the target node Minimum estimated cost rhs ( s The improved heuristic function is used to calculate the value of k, the node priorities are arranged according to the value of k, and the first correction node is determined according to the node priorities.

[0044] Update the first correction node to the current node, and repeat the steps of determining the correction node until the current node coincides with the target node;

[0045] A local obstacle avoidance path is generated based on all correction nodes, and the initial charging path is corrected.

[0046] The formula for calculating the value of k is as follows:

[0047] ;

[0048] ;

[0049] representing a set of all child nodes of the current node s , representing a moving cost of the current node s to a child node , representing a generation value of the child node to a target point, being an actual distance of each movement;

[0050] arranging the k values in ascending order, preferentially arranging the k1 values, and arranging the k2 values in the case of the same k1 values, modifying the candidate node corresponding to the k value arranged first, and the candidate node including the child node generated based on the current node.

[0051] Preferably, the improved algorithm shortens the path distance by adding smoothing processing, including:

[0052] Step 1. Marking n path points on the modified charging path, sequentially marked as 1, 2,...n from the starting point to the target point;

[0053] Step 2. Connecting the path point 1 and the path point 2, and judging whether the connecting line of the two points passes through the obstacle;

[0054] Step 3. Checking the next path point until the connecting line between the current path point q and the path point 1 passes through the obstacle;

[0055] Step 4. Connecting the path point 1 and the path point q- 1 as the end points, and replacing the original path from the path point 1 to the path point q- 1;

[0056] Step 5. Judging whether the starting point is the target point; if yes, the smoothing processing is ended; if not, taking the path point q- 1 as the new starting point, and repeating the steps 1-5.

[0057] According to the technical solution, compared with the prior art, the application provides an obstacle-avoiding mobile charging scheduling algorithm based on the Fresnel model, which dynamically plans a charging path by establishing a wireless rechargeable sensor network model and combining the improved algorithm. Specifically, the Fresnel diffraction model is used to divide a charging signal enhancement zone, a normal zone and a weakening zone, and the charging power is optimized; and the The initial charging path is generated by simulated annealing, and the starting point is dynamically updated when obstacles are detected, and the path is corrected in real time by using a heuristic function; meanwhile, path smoothing processing is introduced to reduce redundant inflection points and reduce mobile energy consumption. The charging efficiency and network life cycle are significantly improved. The balance between high charging efficiency and low energy consumption is achieved by using obstacle enhancement signals instead of simply avoiding obstacles. And through dynamic obstacle avoidance and local re-planning capability, it effectively adapts to real-time changes in complex environments. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 Method flowchart provided by the present application;

[0059] Figure 2 Wireless rechargeable sensor network model diagram;

[0060] Figure 3 Algorithm diagram of the present application;

[0061] Figure 4 Path smoothing processing step diagram;

[0062] Figure 5 Fresnel zone diagram;

[0063] Figure 6 First Fresnel zone (FFZ) tangent to obstacle diagram;

[0064] Figure 7 Charging signal influence partition diagram of mobile charger (MC) at different charging positions;

[0065] Figure 8 CSOF charging path simulation diagram of the algorithm of the present application;

[0066] Figure 9 Trend chart of charging efficiency of different algorithms. DETAILED DESCRIPTION

[0067] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0068] As shown in Figure 1 and Figure 3 , the embodiment of the present application proposes an obstacle avoidance mobile charging scheduling algorithm based on the Fresnel model, including:

[0069] S1. A wireless rechargeable sensor network model is established based on the positions of sensor nodes, mobile chargers and obstacles, including:

[0070] S11. Construct a two-dimensional network area, a plurality of fixed known obstacles are contained in the two-dimensional network area, the sensor nodes are randomly distributed in the non-obstacle area of the two-dimensional network area, and the base station is deployed at the center of the two-dimensional network area.

[0071] In the embodiment, parameters of the sensor nodes (SN) and the mobile charger (MC) are acquired, and the positions of the obstacles are acquired, and a wireless chargeable sensor network model is established. In the initialization, the wireless chargeable sensor network model is established, as shown in FIG. 1. A plurality of three-dimensional static known and fixed obstacles are distributed in the two-dimensional network area, and temporary unknown obstacles may also appear. Figure 2 L L A plurality of three-dimensional static known and fixed obstacles are distributed in the two-dimensional network area, and temporary unknown obstacles may also appear. A plurality of three-dimensional static known and fixed obstacles are distributed in the two-dimensional network area, and temporary unknown obstacles may also appear. A plurality of three-dimensional static known and fixed obstacles are distributed in the two-dimensional network area, and temporary unknown obstacles may also appear. A plurality of three-dimensional static known and fixed obstacles are distributed in the two-dimensional network area, and temporary unknown obstacles may also appear.

[0072] S12. In the two-dimensional network area, n concentric ellipses corresponding to n Fresnel zones are constructed with the position coordinates of the mobile charger and the sensor nodes as the focus.

[0073] As shown in FIG. 2, the Fresnel zone is a concentric ellipse with the MC and the sensor node as the focus, the signal sending end is Figure 5 , the receiving end is , the wavelength of the charging signal is , and the nth ellipse can be expressed as:

[0074] The first Fresnel zone (FFZ) is the main contribution area of the charging signal propagation, and the influence on the energy transmission is much greater than that of the external area. According to Figure 5 , the boundary of the FFZ can be obtained as: , the radius of the FFZ is , since , so , , and , therefore, ; the radius of the nth Fresnel zone is: n .

[0075] ​​​​​​S13. Based on the Fresnel region, the charging signal influence area of different charging positions of the mobile charger is constructed, which is divided into an enhanced region, a normal region and a weakened region; the enhanced region, the normal region and the weakened region are on the concentric circles drawn with the sensor node as the center; wherein, taking the obstacle position as a reference circle, the boundaries of the enhanced region and the weakened region are determined according to the boundaries of the first Fresnel region and the reference circle, and the boundaries of the enhanced region and the normal region are determined according to the boundaries of half of the second Fresnel region and the reference circle.

[0076] Specifically, in Figure 5 , adjacent Fresnel regions have opposite effects on the energy transmission of the FFZ: odd-numbered Fresnel regions have a positive effect on energy transmission, and even-numbered Fresnel regions have a negative effect on energy transmission. If the obstacle is placed in the second Fresnel region, it will block the counteracting wireless charging signal, and the total charging signal will be significantly enhanced; in the WRSN, the sensor node (focus) and the obstacle are fixedly placed, and the MC (another focus) is schedulable, and by adjusting the position of the MC, the obstacle is placed in the second Fresnel region, and the charging signal can be enhanced by the obstacle. At this time, the MC charging scheduling problem is transformed into: given a set of fixed-position sensors and obstacles, how to determine the optimal charging position of the MC, so as to further improve the charging power with the help of the obstacle;

[0077] In order to solve the above problems, the enhanced region, the normal region and the weakened region are divided based on the Fresnel region, as shown in Figure 7 . The enhanced / weakened region refers to the area where the charging signal of the MC is enhanced / weakened due to the influence of the obstacle, and the normal region refers to the area where the MC signal propagation is not affected by the obstacle;

[0078] Firstly, the boundaries of the charging signal enhanced region and the weakened region are constructed. When the obstacle is located on the boundary of the FFZ (i.e. mainly covering the second Fresnel region), the charging signal enhancement effect is the best. And since the FFZ is the main charging signal propagation area, once the obstacle blocks the FFZ, the charging signal will be significantly reduced. Therefore, the tangent point of the FFZ and the obstacle is taken as the boundary of the enhanced region and the weakened region.

[0079] As shown in Figure 6 , it is abstracted as a mathematical model: given a fixed circle (i.e. the circumscribed circle of the obstacle grid) and one focus of the ellipse (i.e. the sensor node), how to determine the position of the other focus (i.e. the MC) so that the circle is tangent to the ellipse. Figure 6 In , the sensor coordinates are , the coordinates of the circular obstacle with a radius of , and the coordinates of the tangent point are , then when the obstacle is tangent to the FFZ, the optimal position of the MC is represented as follows:

[0080] .

[0081] Secondly, the boundary of the enhanced area and the normal area is constructed. When the obstacle is far away from the boundary of the FFZ, the area of the second Fresnel zone blocked by the obstacle is gradually reduced, and the signal enhancement effect is gradually weakened. At this time, the obstacle is not in the propagation path of the charging signal, and this part of the area will not be affected by the obstacle, which is called the signal normal area. When the MC is located at the boundary of the enhanced area and the normal area, only the third formula of the above formula is replaced by .

[0082] The difference between the process of determining the boundary of the enhanced area and the weakened area is that , which means that the obstacle blocks half of the second Fresnel zone.

[0083] Figure 7 The charging signal influence partition of different charging positions of the MC is shown in the figure. Therefore, in order to maximize the charging utility of the network, the MC transmits energy in the charging signal enhanced area (green area), or tries to be located in the normal area, avoiding the MC located in the weakened area to supplement energy for the sensor nodes.

[0084] In different charging signal influence areas, the i-th sensor node s i When the i-th candidate charging position l i The charging power p ( l i , s i ) is as follows:

[0085] ;

[0086] In the formula, , are constants, , ; α is the hardware influence parameter; β is the environmental influence parameter; is the distance between the sensor node s i and the candidate charging position l i .

[0087] S2. In the model of wireless rechargeable sensor network, a charging optimization objective function is constructed according to the effective energy received by the sensor node from the mobile charger, and the charging plan of the mobile charger is planned with the limited energy of the mobile charger as the constraint.

[0088] In the wireless rechargeable sensor network, the increased charging utility is equivalent to the additional energy of the sensor node, which delays the sensor node to send the charging request again, improves the income created by the MC once charging tour, and how to improve the charging utility is very important to reduce the energy cost and maintain the normal operation of the network. Therefore, based on the Fresnel model, under the energy constraint of the MC, the total energy (i.e. effective energy) received by the sensor node from the MC is maximized as the optimization target to improve the network charging utility and prolong the network life cycle.

[0089] Specifically, the charging optimization objective function is as follows:

[0090]

[0091] In the formula, U i is the sensor node s i at the candidate charging position l i The effective energy received from the mobile charger, when the sensor node s i After being fully charged, the excess energy transmitted by the MC will be wasted, so , is the chargeable battery capacity of the sensor node, is the energy transmitted by the mobile charger to the sensor node s i , , is the charging time of the mobile charger at the candidate charging position l i ; the charging time needs to be controlled within , represents the remaining chargeable battery capacity of the sensor node at time .

[0092] With the limited energy of the mobile charger as the constraint, including:

[0093]

[0094] In the formula, M is the number of sensor nodes; is the maximum energy that can be used for wireless charging by the mobile charger; e i is the energy transmitted by the mobile charger to the sensor node l i at the candidate charging position s i ; is the initial charging path length of the mobile charger,​​ Battery energy threshold for mobile charger to be used E move Maximum driving distance of mobile charger under driving; Set of candidate charging locations; Set of to-be-charged locations.

[0095] S3. Discretize the charging area of the mobile charger to the sensor nodes, and obtain the charging sequence and sub-path length of the to-be-charged nodes by using the algorithm, including:

[0096] S31. For each sensor node, draw concentric circles with the sensor node as the center, and draw a same number of concentric circles with a radius of R , so that the charging power difference between adjacent concentric circles does not exceed the threshold , and the charging power of the same ring is equal, and each ring corresponds to a charging power layer; the radius of the th concentric circle is: j ; ;

[0097] S32. The maximum and minimum charging power are respectively: , , and D is the maximum effective charging coverage distance of the mobile charger; the minimum charging power in each charging power layer is taken as the charging power of the entire charging power layer.

[0098] S33. Abstract the known obstacles closest to the sensor node to a grid, and take the circumscribed circle of the grid as the reference circle.

[0099] S34. Based on the charging signal influence area division process in S13, determine the enhanced area of the charging power layer according to the reference circle and the Fresnel zone.

[0100] In this embodiment, if multiple obstacles appear in the perception area of the sensor node, the obstacle closest to the node is selected as the charging signal influencing factor, and the Fresnel signal enhanced area is constructed based on this, and the influence of other obstacles is ignored.

[0101] S34. According to the algorithm, traverse all possible mobile charger location points in the discretized charging area, and obtain the charging sequence and sub-path length of the to-be-charged nodes by maximizing the charging effective amount.

[0102] S4. Design a charging location optimization target, and determine the optimal charging location of each sensor node according to the charging location optimization target.

[0103] ​The charging position of the MC is determined from all candidate charging positions in the first charging power layer signal enhancement area of the to-be-charged sensor node by using a greedy strategy. First, since the charging power attenuates with the increase of the charging distance, the charging utility is improved and the movement consumption is reduced as the optimization objective of determining the charging position, and the objective function of the charging position optimization objective is defined as follows: ;

[0104] wherein, is the movement cost of the mobile charger at the candidate charging position l i when the sensor node s i receives the charging power; is the distance between the mobile charger at the candidate charging position l i and the next candidate charging position in the charging sequence of the to-be-charged node, is the weight coefficient for balancing the charging power and the movement cost.

[0105] Each sensor node determines the optimal charging position corresponding thereto from all candidate charging positions in the first charging power layer signal enhancement area according to the objective function of the charging position optimization objective, and outputs the optimal charging positions of all sensor nodes to form the to-be-charged position set .

[0106] The specific steps are as follows:

[0107] S41. Input the positions of the to-be-charged sensor nodes, the candidate charging positions and the charging power thereof, the charging sequence of the to-be-charged nodes and the sub-path lengths thereof;

[0108] S42. Determine whether there is a candidate charging position located in the signal enhancement area of the to-be-charged sensor node , and calculate the objective function according to the formula ;

[0109] S43. If the objective function is larger, update as the charging position of the to-be-charged sensor node, and if not, keep the original position;

[0110] S44. Output the optimal charging positions of all sensor nodes to obtain the to-be-charged position set .

[0111] S5. According to the optimal charging position, the initial charging path is obtained with the objective of minimizing the number of dead nodes; the mobile charger charges along the initial charging path according to the charging plan, and if an unknown obstacle is detected during the movement, the initial charging path is corrected by using the improved algorithm.

[0112] The obstacle avoidance charging path planning in this embodiment is divided into offline global planning and online dynamic adjustment.

[0113] First, based on the optimal charging position, with the goal of minimizing the number of dead nodes, the following method is used: Generate the initial global charging path.

[0114] The objective function is: Where Z represents the network node mortality rate. This indicates the number of dead nodes in the network.

[0115] Then, the MC moves along the initially planned path, relying on onboard sensors to monitor the network environment in real time during charging. Upon detecting an obstacle, it triggers a local planner, employing an improved method. The algorithm updates the local path cost and corrects the initial path.

[0116] In improvement In algorithms, heuristic functions starting point The system dynamically updates to the current position in real time, and the priority queue is adjusted accordingly. When a new obstacle is detected and the path is replanned, the system focuses on correcting the path segments near the current position. rhs ( s The expression represents the minimum value predicted based on the child node g value of the parent node (in the reverse search), as follows:

[0117] ;

[0118] in, succ ( s ) represents the current node s The set of all child nodes, Indicates the current node s child nodes, Indicates the current node s arrive The cost of moving, express The cost of reaching the target point The actual distance moved each time, initially set to 0; in traditional In China, use k The priority of value-ordered nodes k Smaller values ​​have higher priority. Sort first. k If the values ​​are equal, continue arranging. k The value of 2 is given by the following formula:

[0119] ;

[0120] However, as MC gradually approaches the target point from the new starting point, the heuristic function value continuously decreases, and its proportion also gradually declines, making it prone to multiple nodes. k When values ​​are the same, more nodes are added, reducing search efficiency. Therefore, improvements are needed. Modify its heuristic function to balance its performance in k The proportion of the value is expressed by the following formula:

[0121] ;

[0122] ;

[0123] In the formula, x , y Indicates the current node s The x and y coordinates, , Indicates the starting point The x and y coordinates, , Represents the target node The horizontal and vertical coordinates; As a weighting function, to balance the traditional The proportion of the algorithm's heuristic function in the value of k.

[0124] Adopting improved The algorithm updates the local path cost and corrects the initial path, including:

[0125] Generate the current node s child nodes Based on the current node s To the target node s goal The actual cost g ( s ) and from the current node s Through child nodes Reach the target node s goal Minimum estimated cost rhs ( s The improved heuristic function is used to calculate the value of k, the node priorities are arranged according to the value of k, and the first correction node is determined according to the node priorities.

[0126] Update the first correction node to the current node, and repeat the steps of determining the correction node until the current node coincides with the target node;

[0127] A local obstacle avoidance path is generated based on all correction nodes, and the initial charging path is corrected.

[0128] Further, the step of determining the modified node according to the node priority comprises: arranging the k values in ascending order, preferentially arranging the k1 values, and arranging the k2 values in the case of the same k1 values, and the modified node is a candidate node corresponding to the k value ranked first, and the candidate node includes a child node generated based on the current node.

[0129] Preferably, the improved algorithm shortens the path distance by adding smoothing processing, and the reference Figure 4 includes:

[0130] Step 1. Label the n path points on the modified charging path, sequentially labeled as 1, 2,..., n from the starting point to the target point.

[0131] Step 2. Connect path point 1 and path point 2, and determine whether the connecting line between the two points passes through an obstacle.

[0132] Step 3. Check the next path point until the connecting line between the current path point q ( q is an integer, q n ) and path point 1 passes through an obstacle.

[0133] Step 4. Connect path point 1 and path point q- 1 as endpoints, and replace the original path from path point 1 to path point q- 1.

[0134] Step 5. Determine whether the starting point is the target point; if so, the smoothing processing is completed; if not, path point q- 1 is taken as a new starting point, and steps 1-5 are repeated.

[0135] Figure 8 is a charging path simulation diagram of the algorithm (CSOF) of this embodiment, in which the black grid represents a static obstacle known from the map, and the yellow grid represents a temporary unknown obstacle from the map after the MC starts charging patrol. In order to clearly show the Fresnel modeling process and the charging position of the MC, a Fresnel model partition diagram is overlaid in the simulation diagram in the later stage, wherein the green area represents a Fresnel signal enhancement zone, and the gray circular shape represents an obstacle substituted in the Fresnel modeling. As can be seen from the diagram, the circumscribed circles of the grid obstacles in the perception areas of s 4, s 7 and s 6 are substituted into the Fresnel modeling, and the MC accurately drives to the Fresnel signal enhancement zones of s 4, s 7 and s 6 for charging. At the same time, the CSOF also makes real-time obstacle avoidance response to the unknown obstacle temporarily appearing in the network map, and dynamically adjusts the charging path.​

[0136] Figure 9 To compare the charging utility trend of the algorithms, the embodiment defines three algorithms: a maximum charging utility priority (MUP) algorithm, an ignoring positive effects of obstacles (IPO) algorithm, and a random (RAN) algorithm. The MUP is a charging utility greedy algorithm, which ignores the cost of changing the charging position when determining the charging position. The IPO only focuses on the attenuation effect caused by obstacles and ignores the charging signal enhancement. The RAN randomly selects a charging position that is not covered by obstacles. In the shown graph, the effect of the MC energy budget on the charging utility under different algorithms is mainly investigated. It can be seen that, as the energy budget increases, the energy of the MC can charge more sensor nodes, and the charging utility of the CSOF and other algorithms all shows an upward trend. Figure 9 Specifically, in the process of increasing the energy budget from 1000 J to 4000 J, the charging utility of the CSOF always leads other algorithms, and increases by 44.8%, and tends to be stable when the energy budget is close to 3000 J. Since the MUP is a charging utility greedy algorithm, it preferentially selects the charging position with the maximum charging utility and ignores the moving cost of changing the charging position. When the energy budget of the MC grows enough for charging and moving consumption (such as 4000 J in the graph), the disadvantage caused by ignoring the moving cost is compensated, and the charging utility is basically the same as the CSOF algorithm proposed in the embodiment when the energy budget is 4000 J. This phenomenon further illustrates that the CSOF can better balance the charging utility and the moving cost of the MC when the energy budget of the MC is low, and is more suitable for the network environment with limited energy of the MC.

[0137] In summary, the embodiment proposes an obstacle-avoiding mobile charging scheduling algorithm (CSOF) based on the Fresnel model. The algorithm constructs a Fresnel charging model using obstacles, effectively improves the charging utility, and realizes global planning of the MC charging path and obstacle avoidance for unknown obstacles.

[0138] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same and similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the related parts can be referred to the method part.

[0139] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and that the appended claims are intended to cover all such modifications that do not depart from the true spirit and scope of the application. Therefore, the application is not limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A mobile charging scheduling algorithm based on the Fresnel model with obstacle avoidance, characterized in that, include: A wireless rechargeable sensor network model is established based on the locations of sensor nodes, mobile chargers, and obstacles. In the wireless rechargeable sensor network model, a charging optimization objective function is constructed based on the effective energy received by the sensor nodes from the mobile charger, and the charging plan of the mobile charger is planned with the limited energy of the mobile charger as a constraint. The discrete mobile charger provides charging areas to sensor nodes, aiming to maximize the effective charging amount. The algorithm obtains the charging order of the nodes to be charged and the sub-path length; Design a charging location optimization objective, and determine the optimal charging location for each sensor node based on the charging location optimization objective; Based on the optimal charging location, the initial charging path is obtained with the goal of minimizing the number of dead nodes; The mobile charger charges along the initial charging path according to the charging plan. If an unknown obstacle is detected during the movement, an improved method is used. The algorithm corrects the initial charging path; A wireless rechargeable sensor network model is established based on the locations of sensor nodes, mobile chargers, and obstacles, including: A two-dimensional network region is constructed, which contains multiple fixed known obstacles. Sensor nodes are randomly distributed in the obstacle-free areas of the two-dimensional network region, and the base station is deployed at the center of the two-dimensional network region. Within the two-dimensional network region, n concentric ellipses corresponding to n Fresnel zones are constructed with the position coordinates of the mobile charger and sensor nodes as the focal points. Based on the Fresnel zone, the charging signal influence area at different charging positions of the mobile charger is constructed and divided into an enhancement zone, a normal zone, and a weakening zone. The enhancement zone, normal zone, and weakening zone are on concentric circles drawn with the sensor node as the center. Among them, the obstacle is used as the center of the reference circle. The boundary of the enhancement zone and the weakening zone is determined according to the boundary of the first Fresnel zone and the reference circle. The boundary of the enhancement zone and the normal zone is determined according to half of the boundary of the second Fresnel zone and the reference circle. In different regions affected by charging signals, the i-th sensor node s i At the i-th candidate charging position l i Charging power p ( l i , s i )as follows: ； In the formula, It is a constant greater than 1; It is a constant less than 1; α These are parameters that affect the hardware. β Environmental impact parameters; For sensor nodes s i With candidate charging locations l i The distance between them; The discrete mobile charger provides charging areas to sensor nodes, aiming to maximize the effective charging amount. The algorithm obtains the charging order of the nodes to be charged and the sub-path length, including: For each sensor node, with sensor node s i Concentric circles drawn with as the center are R A radius is Concentric circles, such that the difference in charging power between adjacent concentric circles does not exceed a threshold. The charging power of the same ring is equal, and each ring is equivalent to a charging power layer; the first j The radii of the concentric circles are: ; Minimum charging power within each charging power level This refers to the charging power of the entire charging power layer; among which... D represents the maximum effective charging coverage distance of the mobile charger; Distance sensor node s i The nearest known obstacle is abstracted as a grid, with the grid's outer circle serving as the reference circle; The enhancement region of the charging power layer is determined based on the reference circle and Fresnel zone; according to The algorithm iterates through all possible mobile charger locations within the discretized charging area, aiming to maximize the effective charging amount, and obtains the charging sequence and sub-path length of the nodes to be charged. The objective function for optimizing the charging location is as follows: ； For mobile chargers located at candidate charging locations l i Time sensor node s i Received charging power; For mobile chargers located at candidate charging locations l i The distance between the current node and the next candidate charging position in the charging sequence. To balance the weighting coefficients between charging power and mobility costs; Each sensor node determines the optimal charging position from all candidate charging positions in the first charging power layer signal enhancement region according to the objective function of the charging position optimization objective, and outputs the optimal charging positions of all sensor nodes to form a set of positions to be charged. ; If an unknown obstacle is detected during movement, an improved method is used. The algorithm corrects the initial charging path, including: Improve the heuristic function by setting the starting point of the improved heuristic function. It updates in real time with the current location of the mobile charger; The improved heuristic function is as follows: ； ； In the formula, x , y Indicates the current node s The x and y coordinates, , Indicates the starting point The x and y coordinates, , Represents the target node The x and y coordinates; For weighting functions; Generate the current node s child nodes Based on the current node s To the target node The actual cost g ( s ) and from the current node s Through child nodes Reach the target node Minimum estimated cost rhs ( s The improved heuristic function is used to calculate the value of k, the node priorities are arranged according to the value of k, and the first correction node is determined according to the node priorities. Update the first correction node to the current node, and repeat the steps of determining the correction node until the current node coincides with the target node; A local obstacle avoidance path is generated based on all correction nodes, and the initial charging path is corrected. The formula for calculating the value of k is as follows: ； ； succ ( s ) represents the current node s The set of all child nodes, Indicates the current node s To child nodes The cost of moving, Represents child nodes The cost of reaching the target point This represents the actual distance traveled each time. Arrange the k values ​​in ascending order, prioritizing the k1 value. If the k1 values ​​are the same, continue arranging the k2 values. The corrected node is the candidate node corresponding to the first k value in the sort. The candidate nodes include the child nodes generated based on the current node.

2. The obstacle avoidance mobile charging scheduling algorithm based on the Fresnel model according to claim 1, characterized in that, The objective function for charging optimization is as follows: ； In the formula, U i For sensor nodes s i At candidate charging locations l i The effective energy received from the mobile charger , The rechargeable battery capacity of the sensor node. For mobile chargers to sensor nodes s i The energy transmitted , For mobile chargers at candidate charging locations l i Charging time; Constrained by the limited energy of a mobile charger, including: ； In the formula, M represents the number of sensor nodes; The maximum energy that a mobile charger can use for wireless charging; e i For mobile chargers at candidate charging locations l i Give sensor nodes s i The energy sent; The initial charging path length of the mobile charger. Battery energy threshold for portable chargers that can be used for mobile purposes E move Maximum driving distance of the mobile charger under drive; For the candidate charging location set; This is the set of locations to be charged.

3. The obstacle avoidance mobile charging scheduling algorithm based on the Fresnel model according to claim 1, characterized in that, Based on the optimal charging position, and with the goal of minimizing the number of dead nodes, the following is applied: The initial charging path is generated with SA.

4. The obstacle avoidance mobile charging scheduling algorithm based on the Fresnel model according to claim 1, characterized in that, Improved The algorithm shortens the path distance by incorporating smoothing techniques, including: Step 1. Mark the n path points on the corrected charging path, and label them as 1, 2, ... n from the starting point to the target point; Step 2. Connect path point 1 and path point 2, and determine whether the line connecting the two points passes through an obstacle; Step 3. Check the next waypoint, up to the current waypoint. q The line connecting to path point 1 passes through an obstacle; Step 4. Using path point 1 and path point q- 1 is the endpoint, connecting two points, replacing the original path from point 1 to point 2. q- The path to 1; Step 5. Determine if the starting point is the target point; if yes, the smoothing process ends; otherwise, adjust the path points. q- Starting from point 1, repeat steps 1 through 5.

Images