Illumination network node conflict resolution and collaboration method, equipment and medium

Abstract

The invention discloses a lighting network node conflict resolution and collaboration method, equipment and a medium, belongs to the technical field of smart city Internet of Things, and aims to solve the technical problem of how to balance user experience and energy efficiency while keeping a data topological structure and supporting novel conflict resolution of real-time distributed calculation. The adopted technical scheme is as follows: multi-source data acquisition: acquiring multi-dimensional data of each node in a lighting network; constructing a feature vector: constructing a high-dimensional feature vector of the multi-dimensional data through feature engineering; non-linear manifold embedding: carrying out dimension reduction on the high-dimensional feature vector to a low-dimensional manifold space by adopting a non-linear manifold embedding algorithm, and reserving a topological relation of an original data structure; 3D manifold space mapping: in a low-dimensional manifold space, identifying an instruction conflict area between nodes based on a density peak clustering algorithm, and quantifying conflict intensity; performing multi-objective collaborative optimization; solving a distributed instruction; and incremental manifold updating.

Description

Technical Field

[0001] This invention relates to the field of smart city Internet of Things (IoT) technology, specifically to a method, device, and medium for resolving and coordinating conflicts among lighting network nodes. Background Technology

[0002] Current smart lighting systems generally employ centralized control or simple priority arbitration mechanisms, resulting in high response latency, rigid conflict handling, and low energy efficiency. Traditional methods, such as static rule matching and linear dimensionality reduction techniques, struggle to effectively handle the inherent correlations in high-dimensional nonlinear data within lighting networks, leading to a failure to achieve global optimization when multiple control sources (such as the central system, user apps, and security linkages) conflict. While existing solutions can achieve basic coordination, they rely on preset rules and have poor scalability, failing to adapt to dynamically added nodes. As lighting networks expand to tens of thousands of nodes, balancing user experience and energy efficiency while maintaining data topology and supporting new conflict resolution methods for real-time distributed computing remains a pressing technical challenge. Summary of the Invention

[0003] The technical objective of this invention is to provide a method, device, and medium for resolving and coordinating conflicts among lighting network nodes, in order to address the problem of balancing user experience and energy efficiency while maintaining the data topology and supporting a novel conflict resolution mechanism for real-time distributed computing.

[0004] The technical objective of this invention is achieved as follows: a method for resolving and coordinating conflicts among nodes in a lighting network, the specific details of which are as follows:

[0005] Multi-source data acquisition: Acquiring multi-dimensional data from each node in the lighting network; the multi-dimensional data includes control command data, environmental parameters, equipment status, and topology information;

[0006] Constructing feature vectors: Constructing high-dimensional feature vectors for multi-dimensional data through feature engineering;

[0007] Nonlinear manifold embedding: The nonlinear manifold embedding algorithm is used to reduce the dimensionality of high-dimensional feature vectors to a low-dimensional manifold space while preserving the topological relationship of the original data structure;

[0008] 3D manifold space mapping: In a low-dimensional manifold space, the density peak clustering algorithm is used to identify instruction conflict regions between nodes and quantify the conflict intensity;

[0009] Multi-objective collaborative optimization: Constructing a multi-objective optimization function that integrates spatial consistency constraints, user intent fidelity, and energy efficiency targets;

[0010] Distributed instruction solving: A distributed asynchronous optimization algorithm is used to solve the multi-objective optimization function and generate cooperative control instructions;

[0011] Command issuance and execution: The coordinated control command is issued to each lighting node for execution;

[0012] Incremental manifold update: The manifold structure is dynamically adjusted using an incremental update mechanism to adapt to changes in network topology.

[0013] Preferably, the high-dimensional feature vector has 18 dimensions; among them, the control command data uses one-hot encoding; the environmental parameters are Z-score normalized; and the device status is converted into percentage values.

[0014] The control command data includes central control commands, user APP requests, and third-party system commands; environmental parameters include light intensity (Lux value) and pedestrian flow detection data; equipment status includes lamp brightness, energy consumption, and fault codes; and topology information includes node physical locations and network connection relationships.

[0015] More preferably, the nonlinear manifold embedding algorithm employs an improved Laplacian feature mapping algorithm, which reduces the dimensionality of high-dimensional feature vectors to a low-dimensional manifold space while preserving the topological relationships of the original data structure. The improved Laplacian feature mapping algorithm adds topological constraints to the Laplacian feature mapping algorithm and uses geodesic distance between nodes in similarity calculation to preserve the positional information of nodes in the high-dimensional space to the maximum extent.

[0016] More preferably, an improved Laplacian eigenmap algorithm is used to reduce the dimensionality of high-dimensional feature vectors to a low-dimensional manifold space, preserving the topological relationships of the original data structure, as follows:

[0017] Constructing a weighted adjacency matrix W: Calculating the similarity between nodes based on 18-dimensional feature vectors. Where σ = 0.5 is the Gaussian kernel bandwidth parameter, x i x j W represents the feature vectors of node i and node j, respectively. ij The larger the value, the more similar the two lighting nodes are in the feature space (command, environment, state);

[0018] The degree matrix D is a diagonal matrix with diagonal elements. D ii The larger the value, the more similar or closely related node i is to other nodes in the network.

[0019] Construct the Laplace matrix L = DW, where the diagonal elements L of the Laplace matrix L are... ii L represents the total connectivity strength of node i in the lighting network, which is the sum of the weights of all edges connected to it. The weight values ​​directly reflect the importance of the node; for example, the degree of a central control node is usually significantly higher than that of an edge node. The off-diagonal element L... ijThe connection weight W between nodes i and j ij The negative number; the negative sign indicates the "repulsion" effect between nodes, which is used to preserve the topological structure in subsequent eigenvalue decomposition (similar nodes are close to each other in the manifold space);

[0020] Based on the matrices L and D obtained above, solve the generalized eigenvalue problem L. v =λD v We obtain all n eigenvalues ​​and their corresponding eigenvectors. We select the eigenvectors v1, v2, and v3 corresponding to the first three smallest non-zero eigenvalues ​​λ1, λ2, and λ3. These three eigenvectors v1, v2, and v3 form a mapping from the original high-dimensional feature space to the 3D manifold space. The manifold coordinates of each node i are (v1(i), v2(i), v3(i)), where v1(i) = v2(i) + v3 ... k (i) represents the component of the k-th eigenvector at the i-th node;

[0021] Output 3D manifold coordinates. Each lighting node i is mapped to a point in the 3D manifold space. The coordinates of node i are determined by the projection values ​​of the corresponding node on the three feature vectors, i.e. (v1(i), v2(i), v3(i)), preserving the original data structure.

[0022] More preferably, in a low-dimensional manifold space, the density peak clustering algorithm is used to identify instruction conflict regions between nodes and quantify the conflict intensity as follows:

[0023] Density peak clustering: calculating local density Where, d ij Let ρ be the geodesic distance between nodes i and j in the 3D manifold space, and σ = 0.3 be the neighborhood radius; i The larger the value, the denser the nodes around node i are, and the more likely it is to become the center of a conflict region, i.e., a cluster.

[0024] Identify conflicting clusters: Traverse each divided cluster and check the control commands carried by all nodes in the cluster; when two or more logically mutually exclusive commands exist in a cluster (e.g., "turn on the light" and "turn off the light", "brighten" and "dimen"), the corresponding cluster is officially marked as a "command conflict area".

[0025] Conflict Intensity Quantification: To differentiate the severity of different identified conflict zones, a conflict intensity index is constructed.

[0026] More preferably, a multi-objective optimization function that integrates spatial consistency constraints, user intent fidelity, and energy efficiency targets is constructed as follows:

[0027] Constructing a multi-objective optimization function and introducing a dynamic weighting mechanism: The core of the optimization problem is to minimize the following function:

[0028] minF = W s ·‖XL‖ 2 +W u ·‖UC‖ 2 +W e ·E;

[0029] Among them, the spatial smoothing term ‖XL‖ 2 Used to ensure smooth brightness changes between physically adjacent nodes, avoiding abrupt brightness changes; X is the brightness matrix of the nodes to be optimized; L is the Laplace smoothing constraint matrix constructed based on the node physical topology; Intent fidelity term ||UC|| 2 Used to preserve the user's original control intent to the greatest extent; U is the user instruction matrix, C is the actual execution value matrix, and the energy consumption balancing term E = ∑(P i -P avg ) 2 To balance the overall energy consumption of the network and prevent excessive energy consumption in some nodes; P i P represents the power consumption of node i. avg The average power consumption of the network; the weight W of the spatial consistency term. s Conflict intensity index S with output I Dynamic related: W s =0.6+γ·S I γ is a preset gain coefficient (e.g., γ = 0.05) used to increase the collision intensity S. I This is mapped to the adjustment magnitude of the weights; when the conflict intensity is high, i.e., S... I When W exceeds the set threshold, s As the value increases, the importance of the "spatial consistency" objective is enhanced during the optimization process. The algorithm will be more inclined to have nodes in the conflict area execute similar brightness commands, thereby directly resolving the drastic spatial differences caused by mutually exclusive commands such as "turn on the light" and "turn off the light," achieving conflict reconciliation. When the conflict intensity is low, i.e., S... I When W is less than the set threshold, s Approaching the baseline value of 0.6, the optimization process considers the three objectives more evenly; user intent weight W u and energy consumption weight W e Minor adjustments can be made accordingly to ensure the total weight sums to 1, but the user intent weight W u and energy consumption weight W e Its main function is to ensure that, while resolving conflicts, it does not deviate excessively from the user's original intent and energy efficiency goals;

[0030] For multi-objective optimization functions, asynchronous gradient descent is used for solving, as follows:

[0031] Initialization: Each node maintains a local copy of the parameters, with the initial values ​​being the collected raw command data;

[0032] Iterative update: Gradient calculation and parameter update are performed every 50ms;

[0033] Neighbor communication: Exchange gradient information with neighboring nodes every 50ms to synchronize optimization progress;

[0034] Termination condition: Maximum number of iterations T = 100, or the rate of change of the objective function is less than the threshold of 10. -5 .

[0035] Even better, an incremental update mechanism is used to dynamically adjust the manifold structure to adapt to changes in network topology, as follows:

[0036] Triggering and Identification: When a new node joins the network, the new node reports its identity, physical location, and initial state information, and identifies the topology change event and the set of new nodes affected.

[0037] Local update: To avoid the high computational overhead and latency caused by global retraining, an incremental update strategy is adopted;

[0038] Manifold coordinate recalculation: Only the coordinates of nodes affected by topological changes (including newly added nodes and their nearest neighbors, accounting for about 3% of the total number of nodes) are recalculated in the manifold space, and the solution of the complete generalized eigenvalue problem is avoided by using partial eigenvalue recalculation technology, which greatly reduces the amount of computation.

[0039] Synchronization and Activation: The updated coordinates in the manifold space take effect immediately and are available for use in the next control cycle. This allows for seamless adaptation to network changes without downtime, maintaining high-precision collision detection and collaborative optimization capabilities.

[0040] More preferably, the incremental update strategy is as follows:

[0041] Local connectivity reconstruction: Calculate the k=5 nearest neighbor connections between the new node and the existing nodes, and only update the corresponding local sub-blocks in the weighted adjacency matrix W and degree matrix D;

[0042] Local Laplacian Matrix Update: Based on the updated W and D, locally reconstruct the affected portion of the Laplacian matrix L.

[0043] An electronic device includes: a memory and at least one processor;

[0044] The memory contains computer programs;

[0045] The at least one processor executes the computer program stored in the memory, causing the at least one processor to perform the lighting network node conflict resolution and coordination method as described above.

[0046] A computer-readable storage medium storing a computer program that can be executed by a processor to implement the lighting network node conflict resolution and coordination method described above.

[0047] The lighting network node conflict resolution and coordination method, device, and medium of the present invention have the following advantages:

[0048] (I) This invention uses an improved Laplace eigenmap algorithm to nonlinearly reduce the high-dimensional feature data (including control commands, environmental parameters, equipment status, etc.) of lighting nodes to a low-dimensional manifold space, effectively preserving more than 90% of the structural information of the original data. In the manifold space, a spectral clustering algorithm is used to identify command conflict regions, and a multi-objective optimization function that integrates spatial consistency, user intent fidelity, and energy efficiency targets is constructed. Distributed collaborative solution is achieved through an asynchronous gradient descent algorithm. Compared with traditional linear methods, this invention improves the conflict identification accuracy by 27% and controls the command processing latency to within 200ms when processing systems with tens of thousands of nodes. At the same time, it achieves an energy saving effect of 15-20%, overcoming problems such as command conflicts from multiple control sources, energy efficiency optimization, and real-time response in large-scale distributed lighting systems. It can be widely applied to scenarios such as smart streetlights in cities and lighting in commercial complexes, and has the characteristics of independent controllability, strong adaptability, and cost advantages.

[0049] (II) This invention maps high-dimensional node data to a low-dimensional manifold space to achieve accurate detection and dynamic resolution of multi-source command conflicts. It is a novel conflict resolution method that maintains the data topology structure and supports real-time distributed computing, thereby balancing user experience and energy efficiency. It solves the problems of multi-source heterogeneous control command conflicts, dynamic environmental adaptation and energy efficiency optimization.

[0050] (III) This invention solves the problems of delayed response, rigid priority arbitration mechanism, and distortion of linear dimensionality reduction technology in traditional centralized control. By integrating nonlinear manifold embedding and distributed optimization algorithm, it achieves real-time collaborative control of lighting networks while ensuring the integrity of the topology. It achieves the following objectives: 1) Improves conflict identification accuracy to over 90%; 2) Controls instruction processing latency of tens of thousands of nodes to within 200ms; 3) Simultaneously optimizes energy consumption during conflict resolution, achieving an energy saving effect of 15-20%; 4) Supports dynamic node addition and deletion without global model retraining, significantly improving system scalability and adaptability, and providing an autonomous and controllable efficient solution for smart city lighting.

[0051] (iv) This invention constructs an 18-dimensional node feature vector, integrates multi-source data such as control commands, environmental parameters and equipment status, and uses an improved Laplace feature mapping algorithm to nonlinearly reduce high-dimensional data to a 3D manifold space, retaining more than 90% of the original data structure. In the manifold space, it identifies command conflict areas through density peak clustering and quantifies conflict intensity based on geodesic distance. It designs a multi-objective optimization function that includes spatial consistency, user intent fidelity and energy efficiency targets, uses an asynchronous gradient descent algorithm for distributed solution and deploys an incremental manifold update mechanism to support local model adjustment when dynamic node changes occur. It achieves technical indicators such as 92% conflict detection accuracy, command processing latency <200ms, and energy saving of 15-20%, and the algorithm time only increases linearly by 12% when the system is expanded.

[0052] (v) This invention improves the traditional Laplace feature mapping algorithm by adding topological constraints and using geodesic distance between nodes instead of Euclidean distance in similarity calculation, thus preserving the positional information of nodes in high-dimensional space to the maximum extent.

[0053] (vi) This invention uses the quantified conflict intensity index (SI) as the key input, and dynamically adjusts the optimization weight to prioritize the resolution of high-intensity conflicts. Finally, it solves a set of optimal control commands that can effectively resolve conflicts while taking into account energy efficiency and user intent, thus solving the identified command conflict problem.

[0054] (vii) The present invention can ensure the continuous and stable operation of the system and the guarantee mechanism to adapt to changes in network topology; when changes such as the addition or deletion of nodes occur in the lighting network, the manifold space constructed in the aforementioned steps will become invalid; the present invention uses an incremental update algorithm to efficiently update the manifold space structure, thereby ensuring that the manifold space basis relied upon in subsequent control cycles always remains up-to-date and accurate.

[0055] (VIII) This invention achieves a breakthrough improvement in lighting network control through nonlinear manifold embedding technology, with the following specific effects: ① It increases the collision identification accuracy from 68% to over 92%, enabling real-time detection and resolution of various collision modes in scenarios such as commercial complexes, with a false judgment rate of less than 3%; ② Response performance is significantly optimized, with the instruction processing latency of a system with tens of thousands of nodes compressed to within 200ms, and the measured emergency response time for smart city street light projects is 156ms, an improvement of 83% compared to traditional solutions; ③ In terms of energy efficiency optimization, it achieves 15-20% energy savings while ensuring lighting quality. Benefits: ④ The innovative incremental manifold update mechanism significantly enhances system scalability. When nodes are expanded, the algorithm time only increases linearly by 12%, supporting dynamic changes of 50+ nodes per minute and reducing model retraining frequency by 90%. ⑤ Through nonlinear manifold embedding, energy consumption patterns, equipment degradation trends, and abnormal modes are automatically discovered, comprehensively improving the level of intelligence and achieving a fault prediction accuracy of 95%. ⑥ When deployed in smart street light networks, commercial complex lighting systems, or industrial park lighting networks, it supports expansion to a scale of tens of thousands of nodes, and the algorithm time increases linearly with a growth rate not exceeding 12% when the system is expanded. Attached Figure Description

[0056] The invention will be further described below with reference to the accompanying drawings.

[0057] Appendix Figure 1 A flowchart illustrating the conflict resolution and coordination method for lighting network nodes. Detailed Implementation

[0058] The lighting network node conflict resolution and coordination method, device and medium of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0059] Example 1:

[0060] As attached Figure 1 As shown in the figure, this embodiment provides a method for resolving and coordinating conflicts among nodes in a lighting network. The method is as follows:

[0061] S1. Multi-source data acquisition: Acquire multi-dimensional data from each node in the lighting network; the multi-dimensional data includes control command data, environmental parameters, equipment status, and topology information.

[0062] S2. Constructing Feature Vectors: Constructing high-dimensional feature vectors for multi-dimensional data through feature engineering;

[0063] S3. Nonlinear manifold embedding: The nonlinear manifold embedding algorithm is used to reduce the dimensionality of high-dimensional feature vectors to a low-dimensional manifold space while preserving the topological relationship of the original data structure.

[0064] S4, 3D manifold space mapping: In low-dimensional manifold space, the density peak clustering algorithm is used to identify instruction conflict regions between nodes and quantify the conflict intensity;

[0065] S5. Multi-objective collaborative optimization: Construct a multi-objective optimization function that integrates spatial consistency constraints, user intent fidelity, and energy efficiency targets;

[0066] S6. Distributed instruction solving: A distributed asynchronous optimization algorithm is used to solve the multi-objective optimization function and generate cooperative control instructions;

[0067] S7. Command Issuance and Execution: Issue collaborative control commands to each lighting node for execution;

[0068] S8. Incremental manifold update: The manifold structure is dynamically adjusted using an incremental update mechanism to adapt to changes in network topology.

[0069] In this embodiment, the high-dimensional feature vector in step S2 has an 18-dimensional dimension; the control command data uses one-hot encoding; the environmental parameters are Z-score normalized; and the device status is converted into a percentage value.

[0070] The control command data includes central control commands, user APP requests, and third-party system commands; environmental parameters include light intensity (Lux value) and pedestrian flow detection data; equipment status includes lamp brightness, energy consumption, and fault codes; and topology information includes node physical locations and network connection relationships.

[0071] In this embodiment, the nonlinear manifold embedding algorithm in step S3 adopts an improved Laplacian feature mapping algorithm, that is, the improved Laplacian feature mapping algorithm reduces the dimensionality of high-dimensional feature vectors to a low-dimensional manifold space, while preserving the topological relationship of the original data structure. The improved Laplacian feature mapping algorithm adds topological constraints on the basis of the Laplacian feature mapping algorithm and uses the geodesic distance between nodes in the similarity calculation to preserve the positional information of nodes in the high-dimensional space to the maximum extent.

[0072] In this embodiment, an improved Laplacian eigenmap algorithm is used to reduce the dimensionality of high-dimensional feature vectors to a low-dimensional manifold space, while preserving the topological relationships of the original data structure, as follows:

[0073] S301. Construct a weighted adjacency matrix W: Calculate the similarity between nodes based on 18-dimensional feature vectors. Where σ = 0.5 is the Gaussian kernel bandwidth parameter, x i x j W represents the feature vectors of node i and node j, respectively. ij The larger the value, the more similar the two lighting nodes are in the feature space (command, environment, state);

[0074] S302, Calculate the degree matrix D: the diagonal elements of the diagonal matrix. D ii The larger the value, the more similar or closely related node i is to other nodes in the network.

[0075] S303. Construct the Laplace matrix L = DW, where the diagonal elements of the Laplace matrix L are... ii L represents the total connectivity strength of node i in the lighting network, which is the sum of the weights of all edges connected to it. The weight values ​​directly reflect the importance of the node; for example, the degree of a central control node is usually significantly higher than that of an edge node. The off-diagonal element L... ij The connection weight W between nodes i and j ij The negative number; the negative sign indicates the "repulsion" effect between nodes, which is used to preserve the topological structure in subsequent eigenvalue decomposition (similar nodes are close to each other in the manifold space);

[0076] S304. Based on the matrices L and D obtained above, solve the generalized eigenvalue problem L. v =λD v We obtain all n eigenvalues ​​and their corresponding eigenvectors. We select the eigenvectors v1, v2, and v3 corresponding to the first three smallest non-zero eigenvalues ​​λ1, λ2, and λ3. These three eigenvectors v1, v2, and v3 form a mapping from the original high-dimensional feature space to the 3D manifold space. The manifold coordinates of each node i are (v1(i), v2(i), v3(i)), where v1(i) = v2(i) + v3 ... k (i) represents the component of the k-th eigenvector at the i-th node;

[0077] S305. Output 3D manifold coordinates. Each lighting node i is mapped to a point in the 3D manifold space. The coordinates of node i are determined by the projection values ​​of the corresponding node on the three feature vectors, i.e. (v1(i),v2(i),v3(i)), preserving the original data structure.

[0078] In this embodiment, step S4 involves identifying instruction conflict regions between nodes in the low-dimensional manifold space based on the density peak clustering algorithm and quantifying the conflict intensity as follows:

[0079] S401, Density Peak Clustering: Calculating Local Density Where, d ij Let ρ be the geodesic distance between nodes i and j in the 3D manifold space, and σ = 0.3 be the neighborhood radius; i The larger the value, the denser the nodes around node i are, and the more likely it is to become the center of a conflict region, i.e., a cluster.

[0080] S405. Identify conflicting clusters: Traverse each divided cluster and check the control commands carried by all nodes in the cluster; when there are two or more logically mutually exclusive commands (e.g., "turn on the light" and "turn off the light", "brighten" and "dimen") in a cluster, the corresponding cluster is officially marked as a "command conflict area".

[0081] S403. Conflict Intensity Quantification: To differentiate the severity of different identified conflict zones, a conflict intensity index is constructed.

[0082] In this embodiment, step S5 involves constructing a multi-objective optimization function that integrates spatial consistency constraints, user intent fidelity, and energy efficiency targets, as follows:

[0083] S501. Constructing a multi-objective optimization function and introducing a dynamic weighting mechanism: The core of the optimization problem is to minimize the following function:

[0084] minF = W s ·‖XL‖ 2 +W u ·‖UC‖ 2 +W e ·E;

[0085] Among them, the spatial smoothing term ‖XL‖ 2 Used to ensure smooth brightness changes between physically adjacent nodes, avoiding abrupt brightness changes; X is the brightness matrix of the nodes to be optimized; L is the Laplace smoothing constraint matrix constructed based on the node physical topology; Intent fidelity term ||UC|| 2 Used to preserve the user's original control intent to the greatest extent; U is the user instruction matrix, C is the actual execution value matrix, and the energy consumption balancing term E = ∑(P i -P avg ) 2 To balance the overall energy consumption of the network and prevent excessive energy consumption in some nodes; P i P represents the power consumption of node i. avg The average power consumption of the network; the weight W of the spatial consistency term. s Conflict intensity index S with output I Dynamic related: W s =0.6+γ·S I γ is a preset gain coefficient (e.g., γ = 0.05) used to increase the collision intensity S. I This is mapped to the adjustment magnitude of the weights; when the conflict intensity is high, i.e., S... I When W exceeds the set threshold, sAs the value increases, the importance of the "spatial consistency" objective is enhanced during the optimization process. The algorithm will be more inclined to have nodes in the conflict area execute similar brightness commands, thereby directly resolving the drastic spatial differences caused by mutually exclusive commands such as "turn on the light" and "turn off the light," achieving conflict reconciliation. When the conflict intensity is low, i.e., S... I When W is less than the set threshold, s Approaching the baseline value of 0.6, the optimization process considers the three objectives more evenly; user intent weight W u and energy consumption weight W e Minor adjustments can be made accordingly to ensure the total weight sums to 1, but the user intent weight W u and energy consumption weight W e Its main function is to ensure that, while resolving conflicts, it does not deviate excessively from the user's original intent and energy efficiency goals;

[0086] S502. For multi-objective optimization functions, asynchronous gradient descent is used to solve them, as follows:

[0087] ① Initialization: Each node maintains a local copy of the parameters, with the initial value being the collected raw command data;

[0088] ② Iterative update: Gradient calculation and parameter update are performed every 50ms;

[0089] ③ Neighbor communication: Exchange gradient information with neighboring nodes every 50ms to synchronize optimization progress;

[0090] ④ Termination condition: Maximum number of iterations T = 100, or the rate of change of the objective function is less than the threshold of 10. -5 .

[0091] In step S8 of this embodiment, the incremental update mechanism is used to dynamically adjust the manifold structure to adapt to changes in network topology, as detailed below:

[0092] S801, Triggering and Identification: When a new node joins the network, the new node reports its identity, physical location and initial state information, and identifies the topology change event and the set of new nodes affected.

[0093] S802, Local Update: To avoid the high computational overhead and latency caused by global retraining, an incremental update strategy is adopted;

[0094] S803, Manifold Coordinate Recalculation: Only nodes affected by topological changes (including newly added nodes and their nearest neighbors, accounting for about 3% of the total number of nodes) are recalculated in the manifold space. The partial eigenvalue recalculation technique avoids solving the complete generalized eigenvalue problem, greatly reducing the amount of computation.

[0095] S804 Synchronization and Activation: The updated coordinates in the manifold space take effect immediately and are available for use in the next control cycle. This allows for seamless adaptation to network changes without downtime, maintaining high-precision collision detection and collaborative optimization capabilities.

[0096] The incremental update strategy in step S802 of this embodiment is as follows:

[0097] S80201, Local Connection Reconstruction: Calculate the k=5 nearest neighbor connection between the new node and the existing node, and only update the corresponding local sub-blocks in the weighted adjacency matrix W and degree matrix D;

[0098] S80202, Local Laplacian Matrix Update: Based on the updated W and D, locally reconstruct the affected part of the Laplacian matrix L.

[0099] In this embodiment, nonlinear manifold embedding refers to mapping high-dimensional node data to an implicit low-dimensional nonlinear manifold space using mathematical methods. Its technical features include: 1) Topology preservation: using geodesic distance (rather than Euclidean distance) to preserve the local neighborhood relationships of the original high-dimensional data; 2) Dynamic adaptation: supporting online adjustment of the manifold structure through incremental updates.

[0100] Example 2:

[0101] This embodiment also provides an electronic device, including: a memory and a processor;

[0102] The memory stores the instructions executed by the computer.

[0103] The processor executes computer execution instructions stored in the memory, causing the processor to execute the lighting network node conflict resolution and coordination method in any embodiment of the present invention.

[0104] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor can be a microprocessor or any conventional processor.

[0105] Memory is used to store computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory. Memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, at least one application program required for a function, etc.; the data storage area can store data created based on the use of the terminal, etc. In addition, memory can also include high-speed random access memory, and can also include non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart memory cards (SMC), secure digital cards (SD cards), flash memory cards, at least one disk storage device, flash memory devices, or other volatile solid-state storage devices.

[0106] Example 3:

[0107] This embodiment also provides a computer-readable storage medium storing multiple instructions, which are loaded by a processor to cause the processor to execute the lighting network node conflict resolution and coordination method in any embodiment of the present invention. Specifically, a system or apparatus equipped with a storage medium may be provided, on which software program code implementing the functions of any of the above embodiments is stored, and the computer (or CPU or MPU) of the system or apparatus may read and execute the program code stored in the storage medium.

[0108] In this case, the program code read from the storage medium can itself implement the function of any of the above embodiments, and therefore the program code and the storage medium storing the program code constitute part of the present invention.

[0109] Storage media embodiments for providing program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RYM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, program code can be downloaded from a server computer via a communication network.

[0110] Furthermore, it should be clear that not only can the program code read by the computer be executed, but also the operating system or other components operating on the computer can be instructed based on the program code to perform some or all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0111] Furthermore, it is understood that the program code read from the storage medium is written to the memory set in the expansion board inserted into the computer or to the memory set in the expansion unit connected to the computer. Then, based on the instructions of the program code, the CPU or other components installed on the expansion board or expansion unit execute some and all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for conflict resolution and coordination among nodes in a lighting network, characterized in that, The method is as follows: Multi-source data acquisition: Acquiring multi-dimensional data from each node in the lighting network; the multi-dimensional data includes control command data, environmental parameters, equipment status, and topology information; Constructing feature vectors: Constructing high-dimensional feature vectors for multi-dimensional data through feature engineering; Nonlinear manifold embedding: The nonlinear manifold embedding algorithm is used to reduce the dimensionality of high-dimensional feature vectors to a low-dimensional manifold space while preserving the topological relationship of the original data structure; 3D manifold space mapping: In a low-dimensional manifold space, the density peak clustering algorithm is used to identify instruction conflict regions between nodes and quantify the conflict intensity; Multi-objective collaborative optimization: Constructing a multi-objective optimization function that integrates spatial consistency constraints, user intent fidelity, and energy efficiency targets; Distributed instruction solving: A distributed asynchronous optimization algorithm is used to solve the multi-objective optimization function and generate cooperative control instructions; Command issuance and execution: The coordinated control command is issued to each lighting node for execution; Incremental manifold update: The manifold structure is dynamically adjusted using an incremental update mechanism to adapt to changes in network topology.

2. The method for resolving and coordinating conflicts among lighting network nodes according to claim 1, characterized in that, The high-dimensional feature vector has 18 dimensions; among them, the control command data uses one-hot encoding; the environmental parameters are Z-score normalized; and the device status is converted into percentage values. The control command data includes central control commands, user APP requests, and third-party system commands; environmental parameters include light intensity (Lux value) and pedestrian flow detection data; equipment status includes lamp brightness, energy consumption, and fault codes; and topology information includes node physical locations and network connection relationships.

3. The method for resolving and coordinating conflicts among lighting network nodes according to claim 1 or 2, characterized in that, The nonlinear manifold embedding algorithm employs an improved Laplacian feature mapping algorithm, which reduces high-dimensional feature vectors to a low-dimensional manifold space while preserving the topological relationships of the original data structure. The improved Laplacian feature mapping algorithm adds topological constraints to the Laplacian feature mapping algorithm and uses geodesic distance between nodes in similarity calculation to preserve the positional information of nodes in the high-dimensional space to the greatest extent.

4. The method for resolving and coordinating conflicts among lighting network nodes according to claim 3, characterized in that, An improved Laplacian eigenmap algorithm is used to reduce high-dimensional feature vectors to a low-dimensional manifold space while preserving the topological relationships of the original data structure, as follows: Constructing a weighted adjacency matrix W: Calculating the similarity between nodes based on 18-dimensional feature vectors. Where σ = 0.5 is the Gaussian kernel bandwidth parameter, x i x j W represents the feature vectors of node i and node j, respectively. ij The larger the value, the more similar the two lighting nodes are in the feature space; The degree matrix D is a diagonal matrix with diagonal elements. D ii The larger the value, the more similar or closely related node i is to other nodes in the network. Construct the Laplace matrix L = DW, and the diagonal elements L of the Laplace matrix L are... ii This represents the total connection strength of node i in the lighting network, which is the sum of the weights of all edges connected to it. The weight value directly reflects the importance of the node; off-diagonal elements L ij The connection weight W between nodes i and j ij The opposite of the value; the negative sign indicates the "repulsion" effect between nodes, which is used to preserve the topology in subsequent eigenvalue decomposition; Based on the matrices L and D obtained above, solve the generalized eigenvalue problem L. v =λD v We obtain all n eigenvalues ​​and their corresponding eigenvectors. We select the eigenvectors v1, v2, and v3 corresponding to the first three smallest non-zero eigenvalues ​​λ1, λ2, and λ3. These three eigenvectors v1, v2, and v3 form a mapping from the original high-dimensional feature space to the 3D manifold space. The manifold coordinates of each node i are (v1(i), v2(i), v3(i)), where v1(i) = v2(i) + v3 ... k (i) represents the component of the k-th eigenvector at the i-th node; Output 3D manifold coordinates. Each lighting node i is mapped to a point in the 3D manifold space. The coordinates of node i are determined by the projection values ​​of the corresponding node on the three feature vectors, i.e. (v1(i), v2(i), v3(i)), preserving the original data structure.

5. The method for resolving and coordinating conflicts among lighting network nodes according to claim 4, characterized in that, In a low-dimensional manifold space, a density peak clustering algorithm is used to identify instruction conflict regions between nodes and quantify the conflict intensity as follows: Density peak clustering: calculating local density Where, d ij Let ρ be the geodesic distance between nodes i and j in the 3D manifold space, and σ = 0.3 be the neighborhood radius; i The larger the value, the denser the nodes around node i are, and the more likely it is to become the center of a conflict region, i.e., a cluster. Identify conflicting clusters: Traverse each divided cluster and check the control instructions carried by all nodes in the cluster; when two or more logically mutually exclusive instructions exist in a cluster, the corresponding cluster is officially marked as an "instruction conflict region". Conflict Intensity Quantification: To differentiate the severity of different identified conflict zones, a conflict intensity index is constructed.

6. The method for resolving and coordinating conflicts among lighting network nodes according to claim 5, characterized in that, The multi-objective optimization function that integrates spatial consistency constraints, user intent fidelity, and energy efficiency targets is constructed as follows: Constructing a multi-objective optimization function and introducing a dynamic weighting mechanism: The core of the optimization problem is to minimize the following function: minF=W s ·‖X-L‖ 2 +W u ·‖U-C‖ 2 +W e ·E; Among them, the spatial smoothing term ‖XL‖ 2 Used to ensure smooth brightness changes between physically adjacent nodes, avoiding abrupt brightness changes; X is the brightness matrix of the nodes to be optimized; L is the Laplace smoothing constraint matrix constructed based on the node physical topology; Intent fidelity term ||UC|| 2 Used to preserve the user's original control intent to the greatest extent; U is the user instruction matrix, C is the actual execution value matrix, and the energy consumption balancing term E = ∑(P i -P avg ) 2 To balance the overall energy consumption of the network and prevent excessive energy consumption in some nodes; P i P represents the power consumption of node i. avg The average power consumption of the network; the weight W of the spatial consistency term. s Conflict intensity index S with output I Dynamic related: W s =0.6+γ·S I γ is a preset gain coefficient used to increase the collision intensity S. I Mapped to the adjustment magnitude of the weights; when the conflict intensity value S I When W exceeds the set threshold, s The increase in the conflict intensity value S indicates that the importance of the "spatial consistency" objective is enhanced during the optimization process; I When W is less than the set threshold, s Approaching the baseline value of 0.6, the optimization process considers the three objectives more evenly; user intent weight W u and energy consumption weight W e Minor adjustments can be made accordingly to ensure the total weight sums to 1, but the user intent weight W u And energy consumption weight W e Its main function is to ensure that, while resolving conflicts, it does not deviate excessively from the user's original intent and energy efficiency goals; For multi-objective optimization functions, asynchronous gradient descent is used for solving, as follows: Initialization: Each node maintains a local copy of the parameters, with the initial values ​​being the collected raw command data; Iterative update: Gradient calculation and parameter update are performed every 50ms; Neighbor communication: Exchange gradient information with neighboring nodes every 50ms to synchronize optimization progress; Termination condition: Maximum number of iterations T = 100, or the rate of change of the objective function is less than the threshold of 10. -5 .

7. The method for resolving and coordinating conflicts among lighting network nodes according to claim 6, characterized in that, An incremental update mechanism is used to dynamically adjust the manifold structure to adapt to changes in network topology, as detailed below: Triggering and Identification: When a new node joins the network, the new node reports its identity, physical location, and initial state information, and identifies the topology change event and the set of new nodes affected. Local update: To avoid the high computational overhead and latency caused by global retraining, an incremental update strategy is adopted; Manifold coordinate recalculation: Only the coordinates of nodes affected by topological changes are recalculated in the manifold space, and the problem of solving the complete generalized eigenvalue problem is avoided by using partial eigenvalue recalculation techniques, which greatly reduces the amount of computation. Synchronization and Activation: The updated coordinates in the manifold space take effect immediately and are available for use in the next control cycle.

8. The method for resolving and coordinating conflicts among lighting network nodes according to claim 7, characterized in that, The incremental update strategy is as follows: Local connectivity reconstruction: Calculate the k=5 nearest neighbor connections between the new node and the existing nodes, and only update the corresponding local sub-blocks in the weighted adjacency matrix W and degree matrix D; Local Laplacian Matrix Update: Based on the updated W and D, locally reconstruct the affected portion of the Laplacian matrix L.

9. An electronic device, characterized in that, include: Memory and at least one processor; The memory contains computer programs; The at least one processor executes the computer program stored in the memory, causing the at least one processor to perform the lighting network node conflict resolution and coordination method as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that can be executed by a processor to implement the lighting network node conflict resolution and coordination method as described in any one of claims 1 to 8.

Images