Directed wireless charger deployment and energy distribution joint optimization method

Abstract

The invention discloses a directed wireless charger deployment and energy distribution joint optimization method, which comprises the following steps: (1) acquiring a sensor network diagram, a charger set, a candidate position set and a task set, and establishing a cooperative task model; (2) formalizing a joint optimization problem which is driven by a cooperative task and has connectivity constraints; (3) extracting a dominating strategy, constructing an auxiliary graph based on a dominating strategy set, and converting a complex node-level physical connectivity constraint in an original problem into a strategy-level topological connectivity constraint on the auxiliary graph; (4) decoupling the converted problem into two relatively independent sub-problems of energy distribution and charger deployment; and (5) calling a hierarchical joint optimization algorithm, and determining a charger deployment strategy and an energy distribution strategy. According to the charger deployment and energy distribution scheme, the cooperative task execution is taken as a purpose, and the total utility of the cooperative task can be maximized under the constraint of the deployment number and connectivity.

Description

Technical Field

[0001] This invention relates to the field of wireless rechargeable sensor network technology, and more particularly to a collaborative task-driven method for joint optimization of directed wireless charger deployment and energy distribution. Background Technology

[0002] Wireless sensor networks (WSNs) have been widely used in environmental monitoring, smart cities, military reconnaissance, and industrial control. However, sensor nodes in WSNs typically rely on batteries with limited capacity, and this energy constraint severely limits the continuity and reliability of mission execution. In recent years, wireless power transfer technologies have emerged, providing a novel solution for extending the power supply of sensor nodes.

[0003] Wireless rechargeable sensor networks (WNRs) built on wireless power transfer technology significantly extend network lifecycle and improve the continuity and reliability of task execution by introducing mobile or fixed chargers to wirelessly power sensor nodes. Compared to traditional methods that rely on manual battery replacement, WNRs offer greater network autonomy and deployment flexibility, making them particularly suitable for large-scale and long-term operating sensing systems, demonstrating broad prospects in numerous practical application scenarios. For example, in smart agriculture and ecological environment monitoring, WNRs can support long-term continuous data collection of soil and meteorological parameters; in smart city infrastructure construction, WNRs can support continuous sensing and intelligent management of building structural safety and environmental parameters, enabling efficient monitoring and maintenance of critical facility operating status; and in complex industrial IoT scenarios, WNRs can ensure long-term real-time status monitoring of core equipment in high-temperature or polluted environments.

[0004] Current research largely focuses on improving the resource utilization efficiency of wireless power transfer technologies, such as maximizing charging utility, minimizing latency, reducing charging costs, extending node lifetime, or improving coverage. While these studies have made significant progress in addressing their respective problems, they all neglect the high degree of coupling between node energy consumption and the task load they undertake, treating charging demand merely as a simple supplement to the node's battery energy. In practical applications, different nodes exhibit significant non-uniformity and dynamism in energy consumption due to the vast differences in the sensing or computing tasks they undertake. Focusing solely on node recharging while ignoring task load can easily lead to energy imbalance, affecting task completion. To alleviate this problem, some works have attempted to jointly model charging scheduling and task energy allocation to improve energy utilization efficiency, but these works still do not consider the inter-node communication topology upon which task execution depends. In fact, in multi-node collaborative task execution scenarios, the communication connectivity between nodes is crucial. For example, in environmental monitoring, nodes must locally fuse multimodal data such as temperature, humidity, and smoke to identify anomalies. Poor communication between nodes will force nodes to upload raw data to the cloud, increasing communication load, energy consumption, and response latency. In collaborative sensing or control, nodes must exchange control and status information in real time. Communication interruption will prevent some nodes from participating in the collaboration even if they have sufficient energy. In addition, communication connectivity enables nodes to sense changes in network status, collaboratively complete topology reconstruction and fault recovery, and maintain the normal operation of the network. Summary of the Invention

[0005] This invention aims to address the aforementioned shortcomings of existing technologies and provides a method for determining charger deployment and energy allocation strategies for collaborative task-driven wireless rechargeable sensor networks. This invention ensures that the total utility of collaborative tasks is maximized within constraints on deployment quantity and connectivity.

[0006] The method for joint optimization of directed wireless charger deployment and energy distribution according to the present invention includes the following steps:

[0007] (1) Let the sensor network graph G = {V, Π}, where the vertex set V = {v1, v2, ..., v} N} represents N static wireless rechargeable sensor nodes, and Π is the edge set. For any two nodes v i ,v i′ ∈V(i,i′=1,2,…,N,i≠i′), its Euclidean distance d(v i ,v i′ (Do not exceed the set maximum communication distance D) v Then the two can communicate directly, corresponding to an undirected communication link ξ. i,i′ ∈ Π; Charger set C = {c1, c2, ..., c B}, where B represents the number of directed wireless chargers; the candidate location set L = {l1, l2, ..., l W}, where W represents the number of candidate positions; let the collaborative task set be... Where M represents the number of collaborative tasks, and a collaborative task model is established;

[0008] (2) Formal collaborative task-driven joint optimization problem with connectivity constraints;

[0009] (3) Extract the dominating strategy and construct an auxiliary graph based on the dominating strategy set to transform the complex node-level physical connectivity constraints in the original problem into strategy-level topological connectivity constraints on the auxiliary graph;

[0010] (4) Decouple the transformed problem into two relatively independent sub-problems: energy allocation and charger deployment;

[0011] (5) Call the hierarchical joint optimization algorithm to determine the charger deployment strategy and energy distribution strategy.

[0012] The establishment of the collaborative task model in step (1) includes the following steps:

[0013] (1.1) Node v i From charger c j Received charging power for:

[0014]

[0015] Where α and β are constants related to the hardware and environment, A c D c These represent the charging sector angle and charging radius of the charger, respectively. v For the angle of the receiving sector of the node, They are nodes v i The receiving direction vector and the charger c j The charging direction vector, d(v i ,c j ) represents v i With c j Euclidean distance between them;

[0016] (1.2) Define the charger deployment strategy set Γ={<l1,θ1> ,<l2,θ2> ,…}, where the binary pair <l u ,θ u >Represents the deployment location and orientation angle of the charger (l u ∈L,θ u ∈[0,2π), u=1,2,…), when the charger deployment strategy is Γ, node v i The received energy is:

[0017]

[0018] Where T represents the maximum continuous operating time of the charger, and E max This indicates the upper limit of the battery capacity of the sensor node.

[0019] (1.3) Define energy allocation strategy Where, x i,k For node v i ∈V represents a collaborative task The allocated energy, when the energy allocation strategy is X, affects the collaborative task τ. k The utility of U k (X) is:

[0020]

[0021] in, For task τ k The upper limit of utility, w i,k For node v i For collaborative tasks τ k The utility produced per unit of energy;

[0022] (1.4) Under the charger deployment strategy Γ and energy allocation strategy X, the total utility of the entire network for the cooperative tasks is:

[0023]

[0024] (1.5) Define the subgraph formed by nodes within the charging range under deployment strategy Γ as G. sub ={V sub ,Π sub}, where V sub ={v i |v i ∈V,E i (Γ)>0} represents the set of nodes within the charging range under deployment strategy Γ, Π sub ={ξ i,j |v i ,v j ∈V sub ,d(v i ,v j )≤D v} represents the set of nodes V sub The corresponding edge set.

[0025] In the collaborative task model of this invention, the nodes performing this series of collaborative tasks need to exchange data or coordinate actions in real time. Therefore, all charged nodes V sub The graphs must form connected subgraphs.

[0026] Compared to existing task models that treat nodes as independent task units and do not consider the connectivity between nodes, the collaborative task model of this invention, by defining a communication connectivity subgraph composed of the nodes being charged, can effectively solve the problem of node collaboration failure caused by being in communication islands, thus wasting charging resources. It significantly improves the energy utilization rate and reliability of wireless charging networks when performing complex collaborative tasks.

[0027] Step (2) formalizes the collaborative task-driven joint optimization problem with connectivity constraints, including the following steps:

[0028] The collaborative task-driven joint optimization problem with connection constraint (CTOC) is formalized as follows:

[0029] (P1)maxU(Γ,X) (5)

[0030]

[0031] |Γ|≤B, (5-4)

[0032] G sub It is a connected graph (5-5)

[0033] Wherein, constraint (5-1) represents any node v i The energy allocated to each collaborative task cannot exceed the energy E it receives. i (Γ); Constraint (5-2) requires all energy allocation variables x i,k The value must be non-negative; constraint (5-3) restricts the charger deployment location to be selected only from the candidate location set L, and the range of the charger placement angle is [0, 2π); constraint (5-4) indicates that the actual number of chargers deployed |Γ| cannot exceed the maximum allowed deployment number B; constraint (5-5) is a connectivity constraint, requiring that under the deployment strategy Γ, the subgraph G formed by all nodes that receive charging must be connected. sub They are connected.

[0034] While existing technologies have attempted joint optimization of deployment and allocation, they often overlook the constraints of inter-node communication topology on the execution of collaborative tasks. The CTOC model established in this step introduces the constraint of inter-node communication connectivity into the joint optimization framework of charger deployment and energy allocation for the first time. This overcomes the defect in existing joint optimization models where energy allocation exists but nodes cannot communicate and cooperate, and directly guarantees the effective execution of collaborative tasks from the mathematical model level.

[0035] Step (3) involves extracting the dominance policy method and constructing an auxiliary graph based on the dominance policy set. This transforms the complex node-level physical connectivity constraints in the original problem into policy-level topological connectivity constraints on the auxiliary graph, including the following steps:

[0036] (3.1) Define the domination strategy and domination coverage set: Given the deployment strategy h of any charger = <l u ,θ> and its covered set of nodes V(h), if no strategy h′= <l u ,θ′>, making Then policy h is called the dominance policy of the charger, and V(h) is called the dominance cover set of policy h;

[0037] (3.2) Definition Candidate position l u The set of domination strategies corresponding to ∈L; for each candidate position l u The charger is rotated counterclockwise within the interval [0, 2π) to cover nodes one by one. During the rotation, if a node is about to leave the coverage area, the set of dominating strategies is traversed. For each policy h in the set, determine the current policy h′ = <l u If the set of sensors covered by θ is a subset of the coverage set of policy h, then add the current policy h′ to the dominant coverage policy set. Then continue rotating and executing the above process until the rotation angle is greater than or equal to 2π, at which point the process terminates; Define This is the set of domination strategies obtained after extracting domination strategies from all candidate positions;

[0038] (3.3) Define the connectivity of domination policies: for any two given domination policies and its covered node set If v exists i ∈V(h u ),v i′ ∈V(h u′ ), such that d(v) i ,v i′ )≤D v These two dominance strategies are called h u ,h u′ Interconnected, and denote the domination strategy h. u ,h u′ The edges between them are

[0039] (3.4) Define the domination strategy graph: Given a set of domination strategies Its corresponding dominance strategy diagram is defined as The vertex set in the graph is Each vertex Indicates a dominance strategy <l u ,θ u >, edge set The edge in The necessary and sufficient condition for existence is: two dominance strategies h u ,h u′ Interconnected;

[0040] (3.5) Constructing a Domination Strategy Map Initialize edge set The collection is empty; subsequently, nested loops iterate through the collection. All policy pairs, for each policy pair h u and h u′ Further traverse the nodes they cover, and once a node is detected that satisfies the distance condition d(v)... i ,v i′ )≤D v If the node pair is found, it will immediately be added to the edge set. Add corresponding edges

[0041] (3.6) In constructing the dominance strategy diagram Subsequently, the CTOC problem is transformed into a problem involving the set of dominance strategies. In problem P1, at most B strategies can be chosen to maximize the total utility of the collaborative task, and the set of nodes V is given by... sub The connectivity constraint (Equation 5-5) can be equivalently transformed into applying the dominance strategy graph G under the deployment strategy Γ. Γ ={Γ,Π Γ The connectivity constraints of}, where The CTOC problem, which derives the edge set from the deployment strategy Γ, is formulated as follows:

[0042] (P2)maxU(Γ,X) (6)

[0043]

[0044] Domination Strategy Diagram G Γ It is a connected graph (6-4)

[0045] The extraction of dominance strategies and the construction of auxiliary graphs in this step transform node-level connectivity constraints into topological association constraints at the strategy level, effectively reducing the solution space and lowering the problem complexity while strictly ensuring connectivity.

[0046] Step (4) decouples the transformed problem into two relatively independent sub-problems: energy allocation and charger deployment. This includes the following steps:

[0047] (4.1) Given a set of domination strategies Any node v in equation (7-1) iEnergy received by ∈V The energy allocation strategy X is uniquely determined at this point. This problem is called the energy allocation subproblem, and its formalization is as follows:

[0048]

[0049] (4.2) Given an energy allocation algorithm For any set of domination strategies Γ that satisfies the constraints, the energy allocation strategy for each node is uniquely determined. in For the set of dominance strategies The mapping to the energy allocation strategy X, where the decision dominance strategy Γ is then defined, is called the charger deployment subproblem, which is formalized as:

[0050]

[0051] Domination Strategy Diagram G Γ ={Γ,Π Γ} is a connected graph (8-2)

[0052] The original problem P2 is a mixed integer nonlinear programming problem with a high dependence of decision variables (Γ and X). This step effectively reduces the solution dimension and computational complexity by decoupling it into two relatively independent sub-problems: energy allocation and charger deployment.

[0053] Step (5) calls the hierarchical joint optimization algorithm to determine the charger deployment strategy and energy allocation strategy, including the following steps:

[0054] Step (5) describes the greedy energy allocation algorithm for problem P3. Includes the following steps:

[0055] (5.1) Input: Domination strategy set

[0056] (5.2) Initialize the energy allocation strategy X = 0, and the remaining available energy matrix of the nodes. Current cumulative utility matrix of collaborative tasks Energy sequence The total utility of the collaborative task is U = 0;

[0057] (5.3) Calculate each domination strategy For each node v i Energy provided by ∈V

[0058] (5.4) For each sensor node v i Construct a task priority sequence ∈V This sequence is based on the unit energy efficiency w of each node performing each cooperative task. i,k Perform non-increasing sorting;

[0059] (5.5) Traversing the energy sequence Every bit of energy in e u,i Perform the following procedure:

[0060] (5.5.1) Compute node v i Current actual available energy

[0061] (5.5.2) According to task sequence T i The tasks in the task list are traversed sequentially. When e>0 and the task Current cumulative utility Less than the task threshold The following procedures will be executed at that time:

[0062] (5.5.2.1) is the task distribute Energy;

[0063] (5.5.2.2) Update total utility In the energy allocation strategy, node v i For the task Energy distribution value x i,k =x i,k +Δe, node v i Remaining available energy Task Current cumulative utility node v i The current available energy is e = e - Δe;

[0064] (5.6) Output: Energy allocation strategy X, total utility of cooperative task.

[0065] This step, given the determined charger deployment strategy Γ, allows for the rapid and efficient determination of the energy allocation strategy X. Furthermore, the method described in this step achieves at least optimal performance in the worst-case scenario.

[0066] For the CTOC problem, the neighborhood greedy connectivity deployment algorithm for problem P4 is invoked, along with the greedy energy allocation algorithm for problem P4. As its input;

[0067] Step (5) for deploying the neighborhood greedy connectivity algorithm for problem P4 includes the following steps:

[0068] (6.1) Input: Domination strategy graph Number of chargers B, energy distribution algorithm

[0069] (6.2) Initialize the deployment strategy set Optimal local total utility U * =0;

[0070] (6.3) Traversal Domination Strategy Graph Each node in Perform the following procedure:

[0071] (6.3.1) Initialize the temporary candidate policy set

[0072] (6.3.2) Construct a local neighborhood candidate node set That is, with node h as the center, the distance does not exceed The nodes, excluding the central node itself;

[0073] (6.3.3) In set H h Iterative selection Each strategy node calls the energy allocation algorithm to calculate marginal utility in each iteration. Add the node h′ with the largest gain to the temporary candidate policy set Γ h In the middle, the central node h is added to the local candidate set Γ. h middle;

[0074] (6.3.4) Compare U * and like Then update the optimal local deployment scheme. Optimal center node h * and optimal local total utility

[0075] (6.4) For the optimal local set Traverse strategy nodes In the figure Find a path leading to the central node h. * Find the shortest path and add all relay policy nodes on the path to the final policy set Γ.

[0076] (6.5) Output: Charger deployment strategy and energy distribution strategy

[0077] This step proposes an efficient approximation algorithm to jointly deploy chargers and allocate node energy. This method effectively improves the overall task utility of the network while ensuring connectivity between nodes.

[0078] Beneficial Effects: This invention, for the first time, addresses the energy and network connectivity requirements of actual task execution. It proposes a method for jointly determining charger deployment and node energy allocation while maintaining limited deployment costs and strictly ensuring connectivity between nodes participating in the collaborative task, thereby maximizing overall task utility. This invention boasts advantages such as strong charging practicality and high efficiency, effectively ensuring stable information interaction and data transmission between participating nodes, thus effectively avoiding collaboration failures and energy waste caused by communication silos.

[0079] The dominance strategy extraction and auxiliary graph construction proposed in this invention effectively narrows the solution space, transforming the original node-level connectivity constraints into policy-level topological association constraints, thereby effectively reducing problem complexity while strictly ensuring connectivity; it also proposes a method with... The high-efficiency charger deployment and energy allocation hierarchical joint optimization algorithm with overall approximation can effectively determine the charger deployment scheme and the node energy allocation scheme for tasks.

[0080] This invention proposes a hierarchical joint optimization algorithm for collaborative task-driven joint optimization problems with connectivity constraints, with an overall approximation degree of [missing information].

[0081] For the energy allocation subproblem, the greedy energy allocation algorithm described in step (7) has an approximation degree of: That is, in the worst case, its performance is at least the optimal solution.

[0082] Let X be the energy allocation strategy generated by the greedy energy allocation algorithm; domination strategy Mapped energy sequence Allocate energy sequences according to allocation strategy X. Center front The utility produced by a unit of energy is denoted as U. g U 0 =0; Let the optimal energy allocation strategy be . Energy sequence The first g parts of energy are allocated according to the optimal strategy The utility generated by the distribution is denoted as

[0083] Suppose there is an energy sequence Recorded as forward The energy is allocated according to energy allocation strategy X, and the remaining energy is distributed according to strategy X. to Energy sequence according to optimal strategy Allocation, let this energy allocation strategy be denoted as Before The effect of energy production is exist and U g In, for any previous All energy is allocated according to strategy X, therefore it has the following properties:

[0084]

[0085] Assume there is a special energy allocation strategy For energy sequence The first g-1 portions of energy are allocated according to strategy X, and the g-th portion of energy is allocated according to the optimal strategy. Given that the first g-1 portions of energy are allocated according to strategy X, the greedy energy allocation algorithm's allocation strategy for the g-th portion of energy is a locally optimal greedy strategy, and therefore has the following properties:

[0086]

[0087] In addition, in accordance with strategy Distribute energy sequence In the middle, the first The effect of energy By strategy Distribute energy sequence The utility produced by the gth energy Satisfying Relationship:

[0088]

[0089] This is because, in the first g-1 energy allocations, strategy X+X * With strategy They are completely identical, therefore their corresponding utilities are also identical; however, for strategy X+X... * In the allocation of the first Before the first portion of energy, additional portions from the gth to the th were allocated. Each task has an upper limit to its utility; some tasks may reach utility saturation in the early stages, leading to... Even if the energy is allocated according to the same strategy, the actual marginal utility gained will not exceed the strategy's marginal utility. The utility increment when allocating the g-th portion of energy. Thus, for... We can obtain:

[0090]

[0091]

[0092] Obviously, therefore Proof obtained.

[0093] For the charger deployment subproblem, when the energy allocation strategy X is the greedy energy allocation algorithm described in step (7) (denoted as...) When determined, the approximation degree of the neighborhood greedy connectivity deployment algorithm described in step (7) is: That is, in the worst case, its performance is at least the optimal solution.

[0094] First, it is proved that when the energy allocation strategy X is determined by the greedy energy allocation algorithm described in step (7), the objective function in the charger deployment subproblem is... In the set The above is non-negative, monotonic, and submodular.

[0095] remember For the GEA algorithm on the energy sequence E Γ ={e1,e2,…,e N|Γ| After the first g∈[1,N|Γ|] portions of energy are greedily allocated, node v i For task τ k Distributed energy; For the first g units of energy, the task τ k The cumulative utility; The overall utility produced by the first g units of energy; Δ g (Γ)=U g (Γ)-U g-1 (Γ) represents the marginal utility gain of the g-th energy.

[0096] Nonnegativity: Obviously, for any x i,k ≥0, therefore

[0097] Monotonicity: Take any and The energy sequences mapped by the policy sets Γ and Γ∪{h} are respectively: Where g1 = N|Γ|.

[0098] The greedy energy allocation algorithm performs the exact same greedy allocation on the first g1 units of energy for these two sequences, i.e. Due to E Γ∪{h} There are N additional units of energy available for allocation, and the allocation of these N additional units of energy will not affect the previous allocation results; it will only provide more energy to be allocated. Therefore, node v i For task τ k The allocated energy satisfies

[0099]

[0100] Furthermore,

[0101]

[0102] Therefore, the utility generated by the strategy set Γ and Γ∪{h} and satisfy

[0103]

[0104] That is, monotonicity is proven.

[0105] Submodularity: Take any and a new strategy The energy sequences mapped by Γ and Γ′ are respectively: Energy sequence mapped by strategy h Then the utilities of adding policy h to policy sets Γ and Γ′ are respectively Its marginal utility can be expressed as

[0106]

[0107] Therefore, it is necessary to prove Immediate proof

[0108] right Energy sequence Both are energy sequences The subsequence, therefore we have

[0109]

[0110] Thus, the GEA algorithm in sequence P Γ′∪{e} When continuing to allocate the g2+gth energy, because the allocatable energy of a node may reach its upper limit prematurely or the utility of a task with high unit utility may saturate prematurely, sequence P... Γ′∪{e} With P Γ∪{e} energy The marginal utility gain satisfies:

[0111] Δg 1+g (Γ∪{e})≥Δg 2+g (Γ′∪{e})

[0112] Thus, set functions The submodel property is proved.

[0113] At this point, the charger deployment subproblem is a submodular function maximization problem with connectivity constraints. According to the theory of submodular function maximization problems with connectivity constraints, the approximation ratio of the neighborhood greedy connectivity deployment algorithm described in step (7) is:

[0114] For a collaborative task-driven joint optimization problem with connectivity constraints, the approximation of the hierarchical joint optimization algorithm combining the greedy energy allocation algorithm in joint step (7) and the neighborhood greedy connectivity deployment algorithm is:

[0115] Let the optimal charger deployment strategy be Γ * The optimal energy allocation algorithm is At this time, there are

[0116]

[0117] Assuming Γ′ is the optimal charger deployment strategy obtained by applying the greedy energy allocation algorithm, then we have

[0118]

[0119] The approximation of the neighborhood greedy connected deployment algorithm for the charger deployment subproblem is: The conclusion is that

[0120]

[0121] Therefore, the overall approximation of this invention is: Attached Figure Description

[0122] Figure 1 This is a flowchart illustrating the process of determining the charger deployment strategy and energy distribution strategy in this invention;

[0123] Figure 2 This is a schematic diagram of the network model in this invention;

[0124] Figure 3 This is a call relationship diagram of the hierarchical joint optimization algorithm in this invention;

[0125] Figure 4 This is a flowchart of the greedy energy allocation algorithm in this invention;

[0126] Figure 5 This is a flowchart of the neighborhood greedy connectivity deployment algorithm in this invention. Detailed Implementation

[0127] To more clearly illustrate the purpose, technical solution, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Those skilled in the art should realize that the embodiments described below are only for explaining the present invention and are not intended to limit the present invention.

[0128] Example: A joint optimization method for directed wireless charger deployment and energy distribution, the process is as follows: Figure 1 As shown.

[0129] Step (1) Let the sensor network graph G = {V, Π}, where the vertex set V = {v1, v2, ..., v} N} represents N static wireless rechargeable sensor nodes, and Π is the edge set. For any two nodes v i ,v i′ ∈V(i,i′=1,2,…,N,i≠i′), its Euclidean distance d(v i ,v i′ (Do not exceed the set maximum communication distance D) v Then the two can communicate directly, corresponding to an undirected communication link ξ. i,i′ ∈ Π, the set of chargers C = {c1, c2, ..., c B}, where B represents the number of directed wireless chargers, and the candidate location set L = {l1, l2, ..., l W}, where W represents the number of candidate positions. See the network diagram for reference. Figure 2 Let the set of collaborative tasks be defined. Where M represents the number of collaborative tasks, the collaborative task model is established, including the following steps:

[0130] (1.1) Node v i From charger c j Received charging power for:

[0131]

[0132] Where α and β are constants related to the hardware and environment, A c D c These represent the charging sector angle and charging radius of the charger, respectively. v For the angle of the receiving sector of the node, They are nodes v i The receiving direction vector and the charger c j The charging direction vector, d(v i ,c j ) represents v i With c j Euclidean distance between them;

[0133] (1.2) Define the charger deployment strategy set Γ={<l1,θ1> ,<l2,θ2> ,…}, where the binary pair <l u ,θ u >Represents the deployment location and orientation angle of the charger (l u ∈L,θ u ∈[0,2π), u=1,2,…), when the charger deployment strategy is Γ, node v i The received energy is:

[0134]

[0135] Where T represents the maximum continuous operating time of the charger, and E max This indicates the upper limit of the battery capacity of the sensor node.

[0136] (1.3) Define energy allocation strategy Where, x i,k For node v i ∈V represents a collaborative task The allocated energy, when the energy allocation strategy is X, affects the collaborative task τ. k The utility of U k (X) is:

[0137]

[0138] in, For task τ k The upper limit of utility, w i,k For node v i For collaborative tasks τ k The utility produced per unit of energy;

[0139] (1.4) Under the charger deployment strategy Γ and energy allocation strategy X, the total utility of the entire network for the cooperative tasks is:

[0140]

[0141] (1.5) Define the subgraph formed by nodes within the charging range under deployment strategy Γ as G. sub ={V sub ,Π sub}, where V sub ={v i |v i ∈V,E i (Γ)>0} represents the set of nodes within the charging range under deployment strategy Γ, Π sub ={ξ i,j |v i ,v j ∈V sub ,d(v i ,v j )≤D v} represents the set of nodes V sub The corresponding edge set.

[0142] Step (2) formalizes the collaborative task-driven joint optimization problem with connectivity constraints, including the following steps:

[0143] The collaborative task-driven joint optimization problem with connection constraint (CTOC) is formalized as follows:

[0144] (P1)maxU(Γ,X) (5)

[0145]

[0146] |Γ|≤B, (5-4)

[0147] G sub It is a connected graph (5-5)

[0148] Wherein, constraint (5-1) represents any node v i The energy allocated to each collaborative task cannot exceed the energy E it receives. i (Γ); Constraint (5-2) requires all energy allocation variables x i,k The value must be non-negative; constraint (5-3) restricts the charger deployment location to be selected only from the candidate location set L, and the range of the charger placement angle is [0, 2π); constraint (5-4) indicates that the actual number of chargers deployed |Γ| cannot exceed the maximum allowed deployment number B; constraint (5-5) is a connectivity constraint, requiring that under the deployment strategy Γ, the subgraph G formed by all nodes that receive charging must be connected. sub They are connected.

[0149] Step (3) Extract the dominating policies and construct an auxiliary graph based on the set of dominating policies. This transforms the complex node-level physical connectivity constraints in the original problem into policy-level topological connectivity constraints on the auxiliary graph, including the following steps:

[0150] (3.1) Define the domination strategy and domination coverage set: Given the deployment strategy h of any charger = <l u ,θ> and its covered set of nodes V(h), if no strategy h′= <l u ,θ′>, making Then policy h is called the dominance policy of the charger, and V(h) is called the dominance cover set of policy h;

[0151] (3.2) Definition Candidate position l u The set of domination strategies corresponding to ∈L; for each candidate position l u The charger is rotated counterclockwise within the interval [0, 2π) to cover nodes one by one. During the rotation, if a node is about to leave the coverage area, the set of dominating strategies is traversed. For each policy h in the set, determine the current policy h′ = <l u If the set of sensors covered by θ is a subset of the coverage set of policy h, then add the current policy h′ to the dominant coverage policy set. Then continue rotating and executing the above process until the rotation angle is greater than or equal to 2π, at which point the process terminates; Define This is the set of domination strategies obtained after extracting domination strategies from all candidate positions;

[0152] (3.3) Define the connectivity of domination policies: for any two given domination policies and its covered node set If v exists i ∈V(h u ),v i′ ∈V(h u′ ), such that d(v) i ,v i′ )≤D v These two dominance strategies are called h u ,h u′ Interconnected, and denote the domination strategy h. u ,h u′ The edges between them are

[0153] (3.4) Define the domination strategy graph: Given a set of domination strategies Its corresponding dominance strategy diagram is defined as The vertex set in the graph is Each vertex Indicates a dominance strategy <l u ,θ u >, edge set The edge in The necessary and sufficient condition for existence is: two dominance strategies h u ,h u′ Interconnected;

[0154] (3.5) Constructing a Domination Strategy Map Initialize edge set The collection is empty; subsequently, nested loops iterate through the collection. All policy pairs, for each policy pair h u and h u′ Further traverse the nodes they cover, and once a node is detected that satisfies the distance condition d(v)... i ,v i′ )≤D v The node pairs (line 4) are immediately added to the edge set. Add corresponding edges

[0155] (3.6) In constructing the dominance strategy diagram Subsequently, the CTOC problem is transformed into a problem involving the set of dominance strategies. In problem P1, at most B strategies can be chosen to maximize the total utility of the collaborative task, and the set of nodes V is given by... sub The connectivity constraint (Equation 5-5) can be equivalently transformed into applying the dominance strategy graph G under the deployment strategy Γ. Γ ={Γ,Π Γ The connectivity constraints of}, where The CTOC problem, which derives the edge set from the deployment strategy Γ, is formulated as follows:

[0156] (P2)maxU(Γ,X) (6)

[0157]

[0158] Domination Strategy Diagram G Γ It is a connected graph (6-4)

[0159] Step (4) decouples the transformed problem into two relatively independent sub-problems: energy allocation and charger deployment. This includes the following steps:

[0160] (4.1) Given a set of domination strategies Any node v in equation (7-1) i Energy received by ∈V The energy allocation strategy X is uniquely determined at this point. This problem is called the energy allocation subproblem, and its formalization is as follows:

[0161]

[0162] (4.2) Given an energy allocation algorithm For any set of domination strategies Γ that satisfies the constraints, the energy allocation strategy for each node is uniquely determined. in For the set of dominance strategies The mapping to the energy allocation strategy X, where the decision dominance strategy Γ is then defined, is called the charger deployment subproblem, which is formalized as:

[0163]

[0164] Domination Strategy Diagram G T ={Γ,Π Γ} is a connected graph (8-2)

[0165] Step (5) calls the hierarchical joint optimization algorithm to determine the charger deployment strategy and energy allocation strategy, including the following steps:

[0166] For the CTOC problem, the neighborhood greedy connectivity deployment algorithm for problem P4 is invoked, and the greedy energy allocation algorithm for problem P3 is also applied. As its input, the calling relationship is as follows: Figure 3 ;

[0167] Step (5) describes the greedy energy allocation algorithm for problem P3. The process is as follows Figure 5 This includes the following steps:

[0168] (5.1) Input: Domination strategy set

[0169] (5.2) Initialize the energy allocation strategy X = 0, and the remaining available energy matrix of the nodes. Current cumulative utility matrix of collaborative tasks Energy sequence The total utility of the collaborative task is U = 0;

[0170] (5.3) Calculate each domination strategy For each node v i Energy provided by ∈V

[0171] (5.4) For each sensor node v i Construct a task priority sequence ∈V This sequence is based on the unit energy efficiency w of each node performing each cooperative task. i,k Perform non-increasing sorting;

[0172] (5.5) Traversing the energy sequence Every bit of energy in e u,i Perform the following procedure:

[0173] (5.5.1) Compute node v i Current actual available energy

[0174] (5.5.2) According to task sequence T i The tasks in the task list are traversed sequentially. When e>0 and the task Current cumulative utility Less than the task threshold The following procedures will be executed at that time:

[0175] (5.5.2.1) is the task distribute Energy;

[0176] (5.5.2.2) Update total utility In the energy allocation strategy, node v i For the task Energy distribution value x i,k =x i,k +Δe, node v i Remaining available energy Task Current cumulative utility node v i The current available energy is e = e - Δe;

[0177] (5.6) Output: Energy allocation strategy X, total utility of cooperative task U.

[0178] The neighborhood greedy connectivity deployment algorithm for problem P4 described in step (5) is as follows: Figure 4 This includes the following steps:

[0179] (6.1) Input: Domination strategy graph Number of chargers B, energy distribution algorithm

[0180] (6.2) Initialize the deployment strategy set Optimal local total utility U * =0;

[0181] (6.3) Traversal Domination Strategy Graph Each node in Perform the following procedure:

[0182] (6.3.1) Initialize the temporary candidate policy set

[0183] (6.3.2) Construct a local neighborhood candidate node set That is, with node h as the center, the distance does not exceed The nodes, excluding the central node itself;

[0184] (6.3.3) In set H h Iterative selection Each strategy node calls the energy allocation algorithm to calculate marginal utility in each iteration. Add the node h′ with the largest gain to the temporary candidate policy set Γ h In the middle, the central node h is added to the local candidate set Γ. h middle;

[0185] (6.3.4) Compare U * and like Then update the optimal local deployment scheme. Optimal center node h* and optimal local total utility

[0186] (6.4) For the optimal local set Traverse strategy nodes In the figure Find a path leading to the central node h. * Find the shortest path and add all relay policy nodes on the path to the final policy set Γ.

[0187] (6.5) Output: Charger deployment strategy and energy distribution strategy

[0188] Example 1:

[0189] This embodiment takes a wireless rechargeable sensor network as an example. The joint optimization method for directed wireless charger deployment and energy distribution includes the following steps:

[0190] Let G = {V, Π} denote the sensor network graph, with node set V = {v1, v2, v3, v4, v5, v6, v7} and edge set Π = {ξ}. 1,2 ,ξ 1,3 ,ξ 1,5 ,ξ 2,3 ,ξ 2,5 ,ξ 3,4 ,ξ 5,6 ,ξ 6,7 The charger set C = {c1, c2, c3, c4} and the candidate location set L = {l1, l2, l3, l4, l5, l6} are given. The coordinates of these devices are shown in Table 1. Charging-related parameters are set as follows: α = 50, β = 10, D... c =10, D v =50, A v =π, T = 300, E max =100.

[0191] use This represents a set of collaborative tasks. Task-related parameters are set to: w 1,1 =0.7, w 1,2 =0.3, w 2,1 =0.6, w 2,2 =0.2, w 3,1 =0.4, w 3,2 =0.5, w 4,1 =0.3, w 4,2 =0.1, w 5,1 =0.2, w 5,2 =0.3, w 6,1 =w 6,2 =0.2, w7,1 =0.7, w 7,2 =

[0192] 0.8,

[0193] Table 1 Equipment Parameters

[0194]

[0195] Therefore, after steps (3.1) to (3.2), we obtain...

[0196] therefore,

[0197]

[0198] After constructing the auxiliary graph through steps (3.3) to (3.5), we obtain...

[0199] (3.6) In constructing the dominance strategy diagram Subsequently, the CTOC problem is transformed into a problem involving the set of dominance strategies. In problem P1, at most B strategies can be chosen to maximize the total utility of the collaborative task, and the set of nodes V is given by... sub The connectivity constraint (Equation 5-5) can be equivalently transformed into applying the dominance strategy graph G under the deployment strategy Γ. Γ ={Γ,Π Γ The connectivity constraints of}, where The CTOC problem, which derives the edge set from the deployment strategy Γ, can be formulated as follows:

[0200] (P2)maxU(Γ,X) (6)

[0201]

[0202] Domination Strategy Diagram G Γ It is a connected graph. (6-4)

[0203] Step (4) decouples the transformed problem into two relatively independent sub-problems: energy allocation and charger deployment. This includes the following steps:

[0204] (4.1) Given a set of domination strategies Any node v in equation (7-1) i Energy received by ∈V The energy allocation strategy X is uniquely determined at this point. This problem is called the energy allocation subproblem, and its formalization is as follows:

[0205]

[0206] (4.2) Given an energy allocation algorithm For any set of domination strategies Γ that satisfies the constraints, the energy allocation strategy of each node can be uniquely determined. in For the set of dominance strategies The mapping to the energy allocation strategy X. The decision dominance strategy Γ is then defined as the charger deployment subproblem, formalized as:

[0207]

[0208] Domination Strategy Diagram G Γ ={Γ,Π Γ} is a connected graph. (8-2)

[0209] Step (5): For the CTOC problem, call the neighborhood greedy connectivity deployment algorithm for problem P4, and use the greedy energy allocation algorithm for problem P3. As its input;

[0210] Step (5) describes the greedy energy allocation algorithm for problem P3. set up Includes the following steps:

[0211] (5.1) Input: Domination strategy set

[0212] (5.2) Initialize the energy allocation strategy X = 0, and the remaining available energy matrix of the nodes. Current cumulative utility matrix of collaborative tasks Energy sequence The total utility of the collaborative task is U = 0;

[0213] (5.3) Calculate each domination strategy For each node v i Energy provided by ∈V

[0214] therefore,

[0215] (5.4) For each sensor node v i Construct a task priority sequence ∈V This sequence is based on the unit energy efficiency w of each node performing each cooperative task. i,k Perform non-increasing sorting;

[0216] therefore,

[0217] (5.5) Traversing the energy sequence Every bit of energy in e u,i For example, e 1,1 =76.53, execute the following procedure:

[0218] (5.5.1) The current actual available energy of computing node v1

[0219] (5.5.2) By task sequence The tasks in the task list are traversed sequentially. For example, τ1, when e>0 and the current cumulative utility of task τ1... Less than the task threshold Perform the following procedure:

[0220] (5.5.2.1) Assign tasks to τ1 Energy;

[0221] (5.5.2.2) Update total utility U = U + Δe·w 1,1 =53.57, the energy allocation value x of node v1 for task τ1 in the energy allocation strategy. 1,1 =x 1,1 +Δe = 76.53, the remaining available energy at node v1 Current cumulative utility of task τ1 The current actual available energy of node v1 is e = e - Δe = 0;

[0222] Similarly, after step (5.5), energy e 1,2 =51.9 is all allocated to task τ1, i.e., x 1,2 =51.9, Energy e 6,7 =100 are all allocated to task τ2, i.e., x 7,2 =100, The final total utility U = 164.71;

[0223] (5.6) Output: Energy Allocation Strategy The total utility of the collaborative task is U = 164.71.

[0224] The neighborhood greedy connectivity deployment algorithm for problem P5 described in step (5) includes the following steps:

[0225] (6.1) Input: Domination strategy graph Number of chargers B = 4, energy distribution algorithm

[0226] (6.2) Initialize the deployment strategy set Optimal local total utility U* =0;

[0227] (6.3) Traversal Domination Strategy Graph Each node in For example, h1 executes the following procedure:

[0228] (6.3.1) Initialize the temporary candidate policy set

[0229] (6.3.2) Construct a local neighborhood candidate node set H1 = {h2, h3, h4, h5, h6}, which is a set of nodes centered at node h1 with a distance of no more than 3, excluding the center node itself;

[0230] (6.3.3) Iteratively select one policy node from set H1, and call the energy allocation algorithm to calculate the marginal utility in each iteration. Add the node h6 with the highest gain to the temporary candidate policy set Γ h In the local candidate set Γ, the central node h1 is added. h middle;

[0231] Therefore, Γ1 = {h1, h6};

[0232] (6.3.4) Compare U * and like Then update the optimal local deployment scheme. Optimal center node h * and optimal local total utility

[0233] Therefore, U * =164.71, h * =1,

[0234] Similarly, with h2 as the center, Γ2 = {h1, h2}; Centered on h3, Γ3 = {h1, h3}; Centered on h4, Γ4 = {h4, h6}; Centered on h5, Γ5 = {h5, h6}; Centered on h6, Γ6 = {h1, h6};

[0235] Therefore, after step (6.3)U * =133.57, h * =1,

[0236] (6.4) For the optimal local set Traverse strategy nodes In the figure Find a path leading to the central node h. * Find the shortest path and add all relay policy nodes on the path to the final policy set Γ.

[0237] Therefore, Γ = {h1,h4,h5,h6}

[0238] (6.5) Output: Charger deployment strategy Γ={h1,h4,h5,h6} and energy distribution strategy

[0239]

[0240] Example 2:

[0241] This embodiment uses a wireless rechargeable sensor network for simulation experiments. The joint optimization method for directed wireless charger deployment and energy distribution includes the following steps:

[0242] The simulation parameters are as follows: sensor nodes are uniformly distributed within a 400m × 400m rectangular area, α = 100, β = 40, D c =5, D v =20, The default number of sensors, candidate locations, and tasks are 60, 50, and 100, respectively; the node battery capacity is E. max =10, charger working time T=300; for The upper limit of utility for collaborative tasks The unit energy utility of the sensor for the collaborative task is randomly generated by a uniform distribution between [0,5].

[0243] In this embodiment, four comparison algorithms are presented: connected charger scheduling and greedy energy allocation algorithm, greedy charger scheduling and random energy allocation algorithm, random charger scheduling and greedy energy allocation algorithm, and random charger scheduling and random energy allocation algorithm. The main ideas of the four algorithms are as follows:

[0244] Greedy Charger Deployment and Greedy Energy Allocation (CCGE): GCGE employs a greedy approach to select the charger deployment strategy. In each iteration, it chooses the strategy that yields the maximum increase in charging energy from the connected neighborhood of the currently selected strategies. After selecting the charging strategy, a greedy energy allocation algorithm is used to perform the energy allocation process.

[0245] Greedy Charger Deployment and Random Energy Allocation (GCRE): GCRE works the same as GCGE in the charger deployment phase, also using a greedy strategy that maximizes marginal utility. After the charging strategy is selected, random energy allocation is performed, that is, several tasks are randomly selected for each node and energy is allocated to them as much as possible until the available energy of the node is exhausted.

[0246] Random Charger Scheduling and Greedy Energy Allocation (RCGE): The RCGE charger deployment strategy adopts a random selection method, that is, in each round of iteration, a strategy is randomly selected from the connected neighborhood of the currently selected charging strategy; after deployment is completed, a greedy energy allocation algorithm is used to allocate energy.

[0247] Random Charger Scheduling and Random Energy Allocation (RCRE): RCRE adopts the same approach as RCGE in the charger deployment strategy; and adopts the same approach as GCRE in the energy allocation phase.

[0248] As shown in Table 2, the experimental results show that the Hierarchical Joint Optimization (HJO) algorithm of the present invention outperforms the comparative algorithms under different numbers of chargers. Its cooperative task utility is improved by an average of 10.78%, 85.90%, 33.66%, and 167.68% compared with GCGE, GCRE, RCGE, and RCRE, respectively.

[0249] Table 2. Effectiveness of Collaborative Tasks

[0250]

[0251] The embodiments described above are intended to help readers understand the principles of the present invention, but the scope of protection of the present invention is not limited thereto. Various other substitutions or modifications made by those skilled in the art based on the technical solutions disclosed in the present invention without departing from the essence of the present invention are all within the scope of protection of the present invention.

Claims

1. A method for joint optimization of directed wireless charger deployment and energy distribution, characterized in that: Includes the following steps: (1) Let the sensor network graph G = {V, Π}, where the vertex set V = {v1, v2, ..., v} N } represents N static wireless rechargeable sensor nodes, and Π is the edge set. For any two nodes v i ,v i′ ∈V(i,i′=1,2,…,N,i≠i′), its Euclidean distance d(v i ,v i′ (Do not exceed the set maximum communication distance D) v If the two communicate directly, it corresponds to an undirected communication link ξ. i,i′ ∈ Π; Charger set C = {c1, c2, ..., c B }, where B represents the number of directed wireless chargers; the candidate location set L = {l1, l2, ..., l W }, where W represents the number of candidate positions; let the collaborative task set be... Where M represents the number of collaborative tasks, and a collaborative task model is established; (2) Formal collaborative task-driven joint optimization problem with connectivity constraints; (3) Extract the dominating strategy and construct an auxiliary graph based on the dominating strategy set to transform the complex node-level physical connectivity constraints in the original problem into strategy-level topological connectivity constraints on the auxiliary graph; (4) Decouple the transformed problem into two relatively independent sub-problems: energy allocation and charger deployment; (5) Call the hierarchical joint optimization algorithm to determine the charger deployment strategy and energy distribution strategy.

2. The method for joint optimization of directed wireless charger deployment and energy distribution according to claim 1, characterized in that: Step (1) includes the following steps: (1.1) Node v i From charger c j Received charging power for: Where α and β are constants related to the hardware and environment, A c D c These represent the charging sector angle and charging radius of the charger, respectively. v For the angle of the receiving sector of the node, They are nodes v i The receiving direction vector and the charger c j The charging direction vector, d(v i ,c j ) represents v i With c j Euclidean distance between them; (1.2) Define the charger deployment strategy set Γ={<l1,θ1> ,<l2,θ2> ,…}, where the binary pair <l u ,θ u >Represents the deployment location and orientation angle of the charger (l u ∈L,θ u ∈[0,2π), u=1,2,…), when the charger deployment strategy is Γ, node v i The received energy is: Where T represents the maximum continuous operating time of the charger, and E max This indicates the upper limit of the battery capacity of the sensor node. (1.3) Define energy allocation strategy Where, x i,k For node v i ∈V represents a collaborative task The allocated energy, when the energy allocation strategy is X, affects the collaborative task τ. k The utility of U k (X) is: in, For task τ k The upper limit of utility, w i,k For node v i For collaborative tasks τ k The utility produced per unit of energy; (1.4) Under the charger deployment strategy Γ and energy allocation strategy X, the total utility of the entire network for the cooperative tasks is: (1.5) Define the subgraph formed by nodes within the charging range under deployment strategy Γ as G. sub ={V sub ,Π sub }, where V sub ={v i |v i ∈V,E i (Γ)>0} represents the set of nodes within the charging range under deployment strategy Γ, Π sub ={ξ i,j |v i ,v j ∈V sub ,d(v i ,v j )≤D v } represents the set of nodes V sub The corresponding edge set.

3. The method for joint optimization of directed wireless charger deployment and energy distribution according to claim 2, characterized in that: Step (2) includes the following steps: The collaborative task-driven joint optimization problem with connection constraint (CTOC) is formalized as follows: (P1)maxU(Γ,X) (5) |Γ|≤B, (5-4) G sub It is a connected graph (5-5) Wherein, constraint (5-1) represents any node v i The energy allocated to each collaborative task cannot exceed the energy E it receives. i (Γ); Constraint (5-2) requires all energy allocation variables x i,k The value must be non-negative; constraint (5-3) restricts the charger deployment location to be selected only from the candidate location set L, and the range of the charger placement angle is [0, 2π); constraint (5-4) indicates that the actual number of chargers deployed |Γ| cannot exceed the maximum allowed deployment number B; constraint (5-5) is a connectivity constraint, requiring that under the deployment strategy Γ, the subgraph G formed by all nodes that receive charging must be connected. sub They are connected.

4. The method for joint optimization of directed wireless charger deployment and energy distribution according to claim 3, characterized in that: Step (3) includes the following steps: (3.1) Define the domination strategy and domination coverage set: Given the deployment strategy h of any charger = <l u ,θ> and its covered set of nodes V(h), if no strategy h′= <l u ,θ′>, making Then policy h is called the dominance policy of the charger, and V(h) is called the dominance cover set of policy h; (3.2) Definition Candidate position l u The set of domination strategies corresponding to ∈L; for each candidate position l u The charger is rotated counterclockwise within the interval [0, 2π) to cover nodes one by one. During the rotation, if a node is about to leave the coverage area, the set of dominating strategies is traversed. For each policy h in the set, determine the current policy h′ = <l u If the set of sensors covered by θ is a subset of the coverage set of policy h, then add the current policy h′ to the dominant coverage policy set. Then continue rotating and executing the above process until the rotation angle is greater than or equal to 2π, at which point the process terminates; Define This is the set of domination strategies obtained after extracting domination strategies from all candidate positions; (3.3) Define the connectivity of domination policies: for any two given domination policies and its covered node set If v exists i ∈V(h u ),v i′ ∈V(h u′ ), such that d(v) i ,v i′ )≤D v These two dominance strategies are called h u ,h u′ Interconnected, and denote the domination strategy h. u ,h u′ The edges between them are (3.4) Define the domination strategy graph: Given a set of domination strategies Its corresponding dominance strategy diagram is defined as The vertex set in the graph is Each vertex Indicates a dominance strategy <l u ,θ u >, edge set The edge in The necessary and sufficient condition for existence is: two dominance strategies h u ,h u′ Interconnected; (3.5) Constructing a Domination Strategy Map Initialize edge set The collection is empty; subsequently, nested loops iterate through the collection. All policy pairs, for each policy pair h u and h u′ Further traverse the nodes they cover, and once a node is detected that satisfies the distance condition d(v)... i ,v i′ )≤D v If the node pair is found, it will immediately be added to the edge set. Add corresponding edges (3.6) In constructing the dominance strategy diagram Subsequently, the CTOC problem is transformed into a problem involving the set of dominance strategies. In problem P1, at most B strategies can be chosen to maximize the total utility of the collaborative task, and the set of nodes V is given by... sub The connectivity constraint (Equation 5-5) can be equivalently transformed into applying the dominance strategy graph G under the deployment strategy Γ. Γ ={Γ,Π Γ The connectivity constraints of}, where The CTOC problem, which derives the edge set from the deployment strategy Γ, is formulated as follows: (P2)maxU(Γ,X) (6) Domination Strategy Diagram G Γ It is a connected graph (6-4).

5. The method for joint optimization of directed wireless charger deployment and energy distribution according to claim 4, characterized in that: Step (4) includes the following steps: (4.1) Given a set of domination strategies Any node v in equation (7-1) i Energy received by ∈V The energy allocation strategy X is uniquely determined at this point. This problem is called the energy allocation subproblem, and its formalization is as follows: (4.2) Given an energy allocation algorithm For any set of domination strategies Γ that satisfies the constraints, the energy allocation strategy for each node is uniquely determined. in For the set of dominance strategies The mapping to the energy allocation strategy X, where the decision dominance strategy Γ is then defined, is called the charger deployment subproblem, which is formalized as: Domination Strategy Diagram G T ={Γ,Π Γ } is a connected graph (8-2).

6. The method for joint optimization of directed wireless charger deployment and energy distribution according to claim 5, characterized in that: Step (5) Greedy energy allocation algorithm for problem P3 Includes the following steps: (5.1) Input: Domination strategy set (5.2) Initialize the energy allocation strategy X = 0, and the remaining available energy matrix of the nodes. Current cumulative utility matrix of collaborative tasks Energy sequence The total utility of the collaborative task is U = 0; (5.3) Calculate each domination strategy For each node v i Energy provided by ∈V (5.4) For each sensor node v i Construct a task priority sequence ∈V This sequence is based on the unit energy efficiency w of each node performing each cooperative task. i,k Perform non-increasing sorting; (5.5) Traversing the energy sequence Every bit of energy in e u,i Perform the following procedure: (5.5.1) Compute node v i Current actual available energy (5.5.2) According to task sequence T i The tasks in the task list are traversed sequentially. When e>0 and the task Current cumulative utility Less than the task threshold The following procedures will be executed at that time: (5.5.2.1) is the task distribute Energy; (5.5.2.2) Update total utility In the energy allocation strategy, node v i For the task Energy distribution value x i,k =x i,k +Δe, node v i Remaining available energy Task Current cumulative utility node v i The current available energy is e = e - Δe; (5.6) Output: Energy allocation strategy X, total utility of cooperative task U.

7. The method for joint optimization of directed wireless charger deployment and energy distribution according to claim 6, characterized in that: Step (5) includes the following steps: For the CTOC problem, the neighborhood greedy connectivity deployment algorithm for problem P4 is invoked, along with the greedy energy allocation algorithm for problem P4. As its input; Step (5) for deploying the neighborhood greedy connectivity algorithm for problem P4 includes the following steps: (6.1) Input: Domination strategy graph Number of chargers B, energy distribution algorithm (6.2) Initialize the deployment strategy set Optimal local total utility U * =0; (6.3) Traversal Domination Strategy Graph Each node in Perform the following procedure: (6.3.1) Initialize the temporary candidate policy set (6.3.2) Construct a local neighborhood candidate node set That is, with node h as the center, the distance does not exceed The nodes, excluding the central node itself; (6.3.3) In set H h Iterative selection Each policy node calls the energy allocation algorithm in each iteration. Calculate marginal utility Add the node h′ with the largest gain to the temporary candidate policy set Γ h In the middle, the central node h is added to the local candidate set Γ. h middle; (6.3.4) Compare U * and like Then update the optimal local deployment scheme. Optimal center node h * and optimal local total utility (6.4) For the optimal local set Traverse strategy nodes In the figure Find a path leading to the central node h. * Find the shortest path and add all relay policy nodes on the path to the final policy set Γ; (6.5) Output: Charger deployment strategy and energy distribution strategy 8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the directional wireless charger deployment and energy distribution joint optimization method as described in any one of claims 1 to 7.

Images