Linkage dimming method, system and device based on end-edge cooperative calculation and medium

Abstract

The application discloses a kind of linkage dimming methods, systems, equipment and medium based on end side collaborative computing, belong to intelligent lighting control technical field.The method includes: sensor acquires environmental data to slave node controller, data is analyzed by it lightly, and according to task type local generation instruction or unloading to edge gateway;Slave node controller is linked with adjacent node by short-range wireless direct connection protocol;Edge gateway real-time monitoring wireless channel, and when interference, dynamic switching algorithm is executed to select optimal channel;Edge gateway is based on non-cooperative game task scheduling algorithm, and dynamic task allocation and load balancing scheduling are carried out between slave node and edge gateway, and task is dynamically rescheduled according to load state;Edge gateway periodically summarizes data to generate report and upload to cloud.The application improves communication reliability, system real-time and robustness, realizes efficient end side collaboration and resource utilization.

Description

Technical Field

[0001] This invention relates to the field of intelligent lighting control technology, specifically to a linkage dimming method, system, device, and medium based on edge-end collaborative computing. Background Technology

[0002] With the rapid development of smart city and IoT technologies, public lighting systems are undergoing a profound transformation from traditional single control to intelligent, networked, and energy-saving systems. As an important public infrastructure and sensing node in cities, smart streetlights are expected to have advanced functions beyond basic switching and dimming, such as on-demand lighting, dynamic linkage, efficient operation and maintenance, and data analysis, to meet the diverse needs of smart management, energy conservation, and improved public experience.

[0003] To achieve the above goals, existing technologies have proposed various networked lighting control schemes based on wired or wireless communication, and have gradually introduced the concept of edge computing to reduce cloud load and transmission latency. However, when facing dynamic linkage scenarios with high real-time requirements and high reliability, such as lights following vehicles and lights turning off when people leave, existing solutions still have a series of key bottlenecks in system architecture, communication methods, and task scheduling, which restrict their large-scale application and performance improvement. (1) Insufficient communication link stability makes it difficult to guarantee reliable control in complex environments. Existing control systems mostly rely on a single communication method. Although wired methods such as RS-485 and PLC are relatively reliable, they have problems such as complex wiring, high cost, and difficulty in cross-regional expansion. While wireless methods such as ZigBee, LoRa, and RF433 are flexible in deployment, they are easily affected by the complex electromagnetic environment of cities, building obstruction, and co-channel interference, resulting in control command loss, delay, or increased bit error rate, and failing to guarantee a high communication success rate. Although some solutions attempt to combine wired and wireless (dual-mode communication), they are usually only used for simple master-slave switching or fixed mode. They fail to design deep collaboration and adaptive optimization mechanisms (such as dynamic channel selection and cross-domain interference coordination) for the physical characteristics of dual-mode (such as PLC being unable to cross transformer areas but having strong anti-interference, and wireless being susceptible to co-channel interference but having high flexibility). They cannot give full play to the complementary advantages of dual-mode, and the overall communication reliability is still poor in complex urban environments.

[0004] (2) The system has a large response delay, poor real-time linkage, and insufficient robustness. The linkage logic of the traditional architecture relies heavily on the cloud platform or centralized gateway for processing. Sensor data needs to be transmitted and processed through a long link of "terminal → edge gateway → cloud → edge gateway → terminal", resulting in an end-to-end delay of more than 500 milliseconds from event perception to execution control. This cannot meet the millisecond-level response requirements of real-time lighting scenarios such as lights following the vehicle, which may cause lighting lag and lead to driver visual fatigue or safety hazards. More importantly, the terminal controller (often as a slave node) is usually designed only as an instruction execution unit and does not have local data parsing and decision-making capabilities. Once the network connection is unstable or interrupted, the entire linkage function will fail. The system lacks the ability to maintain basic intelligent services in the event of a network outage, resulting in poor robustness.

[0005] (3) Rigid allocation of edge computing tasks leads to unbalanced resource utilization, making it difficult to balance system efficiency and stability. Existing systems that introduce edge computing often fail to establish effective edge-edge collaborative computing mechanisms. The allocation of computing tasks is fixed, usually concentrating all complex calculations (such as data fusion and policy decision-making) on ​​the edge gateway, while terminal devices with limited resources only perform the simplest dimming actions. This leads to the edge gateway being prone to computing resource overload during peak hours (such as CPU utilization consistently exceeding 90%), becoming a performance bottleneck; while the microcontroller (MCU) computing resources of the terminal devices are largely idle (utilization is often below 30%). This "heavy on one end and light on the other" load pattern not only limits the improvement of the overall system processing capacity, but also reduces the stability of the system when dealing with sudden tasks or high loads in local areas, lacking the ability to dynamically schedule and flexibly allocate tasks based on real-time network status, task urgency, and device load.

[0006] (4) The difficulty in integrating lightweight computing into the terminal hinders the realization of local intelligence. Due to cost considerations for large-scale deployment, street light terminals typically use resource-constrained, low-cost microcontrollers (such as the ARM Cortex-M0 series), whose flash memory capacity is usually less than 256KB and whose computing power is limited (e.g., less than 50 DMIPS). Existing technical solutions lack optimized computing models and algorithms for such lightweight hardware, making it impossible to offload necessary data parsing, feature extraction, and simple decision-making functions to the terminal. This makes the terminal highly dependent on external instructions, further exacerbating the aforementioned dependence on the network and central nodes, creating a vicious cycle.

[0007] Therefore, there is an urgent need in this field for an innovative technical solution that can fundamentally solve the complex technical problems that are intertwined, such as communication reliability, real-time response, end-edge collaboration efficiency, and lightweight terminal computing capabilities, so as to achieve truly high-reliability, low-latency, adaptive, and cost-effective intelligent linkage dimming. Summary of the Invention

[0008] To address the problems of low communication reliability, large system response delay, imbalance between edge computing and terminal intelligence in existing smart lighting technologies, this invention provides a linkage dimming method, system, device, and medium based on edge collaborative computing to solve the aforementioned technical deficiencies.

[0009] In a first aspect, the present invention proposes a linkage dimming method based on edge-end collaborative computing, which includes the following steps: S1. The sensor collects environmental data in real time and transmits it to the slave node controller. The slave node controller performs lightweight data parsing on the received environmental data. If it is determined to be a simple real-time task, it generates a dimming control command locally. If it is determined to be a complex non-real-time task, it offloads the task to the edge gateway. S2. After generating the dimming control command, the slave node controller sends the linkage command to the associated adjacent slave node controller through the preset short-range wireless direct connection protocol to realize the linkage of area lighting. At the same time, the edge gateway monitors the wireless channel status in real time and executes the wireless channel dynamic switching algorithm to switch to the optimal idle channel when channel interference is detected. S3. The edge gateway receives non-real-time complex tasks offloaded by the slave node controllers and performs dynamic task allocation and load balancing scheduling between the slave node controllers and the edge gateway based on a non-cooperative game-theoretic task scheduling algorithm. The edge gateway monitors the load status of each slave node controller in real time. When the load of a slave node controller exceeds the dynamically adjusted offload threshold, some of the tasks allocated to that slave node controller will be rescheduled. S4. The edge gateway periodically summarizes the energy consumption and fault data uploaded from the node controller, generates statistical reports, and uploads the statistical reports to the cloud platform during periods of low network load.

[0010] Preferably, an initialization step is included before step S1, specifically: The sensors in the front-end perception layer complete power-on self-tests and calibrations, and output standardized digital signals to the slave node controllers in the terminal control layer; the slave node controllers load local linkage strategies and initialize the dual-mode communication module; the edge gateways in the edge collaboration layer start the task scheduling algorithm and set the load threshold. The dual-mode communication module includes a power line carrier communication submodule and a wireless communication submodule; the slave node controller directly connects with the adjacent slave node controller through the wireless communication submodule; and the edge gateway sends control commands to the slave node controller through the power line carrier communication submodule.

[0011] Preferably, in step S1, the slave node controller performs lightweight data parsing on the received data. If it is determined to be a real-time simple task, a dimming control command is generated locally. If it is determined to be a non-real-time complex task, the task is offloaded to the edge gateway. Specifically, this includes the following sub-steps: S11. Based on preset physical rules or range thresholds, verify the environmental data collected by the sensor, remove abnormal data, and obtain valid data. S12. Extract the preset key feature information for generating dimming instructions from the valid data. The key feature information includes the target existence state, target position and target movement direction. S13. Match the key feature information with the linkage strategy pre-stored locally on the slave node controller; if the matching result meets the real-time simple task judgment conditions, generate dimming control instructions locally; if the matching result meets the non-real-time complex task judgment conditions, offload the current data and processing context as a task to the edge gateway.

[0012] Preferably, in step S2, the edge gateway monitors the wireless channel status in real time, and when channel interference is detected, executes a dynamic wireless channel switching algorithm to switch to the optimal idle channel, specifically including the following sub-steps: S21. The edge gateway monitors the signal quality of the current wireless channel in real time and calculates the channel interference intensity based on the monitoring results. When the channel interference intensity exceeds the preset high interference threshold, it triggers channel switching. S22. The edge gateway scans multiple preset candidate wireless frequency bands and evaluates the idle status of each candidate frequency band according to the preset idle determination rules. S23. The edge gateway combines the frequency differences between each frequency band and the channels used by the adjacent slave node controllers, and selects the idle channel with the best comprehensive evaluation result from the candidate frequency bands as the target channel. S24. The edge gateway reports the target channel information to the cloud platform through the WAN communication module and obtains the channel information of the adjacent edge gateways to determine whether switching to the target channel will cause interference to the adjacent edge gateways; if not, proceed with the following steps. S25. The edge gateway sends a frequency band switching command to the slave node controller under its jurisdiction through the common channel of the power line carrier communication module, and controls its own and the slave node controller's wireless communication modules to switch to the target channel.

[0013] More preferably, in step S21, the channel interference intensity is calculated and determined based on the ratio of the real-time signal-to-noise ratio of the wireless channel to the theoretical maximum signal-to-noise ratio of its frequency band. In step S23, the comprehensive evaluation is performed as follows: for each candidate frequency band evaluated as idle, the weighted difference between the channel idle duration and the frequency difference between the channel used by the adjacent slave node controller is calculated, wherein the weight coefficient of the idle duration is greater than the weight coefficient of the frequency difference; the idle channel with the highest weighted difference is selected as the target channel.

[0014] Preferably, step S3 specifically includes the following sub-steps: S31. The edge gateway receives non-real-time complex tasks offloaded from the node controller and prioritizes them according to their real-time requirements. S32. Based on the non-cooperative game-theoretic task scheduling algorithm, evaluate the first benefit of non-real-time complex tasks when they are processed by the slave node controller and the second benefit when they are processed by the edge gateway. The evaluation of the first benefit is based on the task processing efficiency and resource consumption of the corresponding slave node controller, and the evaluation of the second benefit is based on the number of tasks processed by the edge gateway and the end-edge communication latency. S33. Perform iterative scheduling at a preset period. Based on the comparison between the first benefit and the second benefit, and in combination with the priority of the task, dynamically adjust the allocation ratio of non-real-time complex tasks between the slave node controller and the edge gateway until a balance between load and benefit is achieved. S34. The edge gateway monitors the load status of each slave node controller in real time and dynamically adjusts the offloading threshold according to the average load of slave node controllers in the region. When the load of a slave node controller exceeds the current offloading threshold, the rescheduling mechanism is triggered to reallocate part of the tasks undertaken by the slave node controller to other idle edge gateways.

[0015] More preferably, in step S33, the iterative scheduling continues until the preset convergence conditions are met; the convergence conditions include: in multiple consecutive iterations, the change in the allocation ratio of non-real-time complex tasks between the slave node controller and the edge gateway is less than a first threshold, and the difference between the average revenue of the slave node controller and the revenue of the edge gateway is less than a second threshold.

[0016] Secondly, this invention proposes a linkage dimming system based on edge-end collaborative computing for implementing any of the methods described above, the system comprising: The front-end perception layer includes at least one sensor, which is configured to collect environmental data in real time and transmit it to the slave node; The terminal control layer includes multiple slave node controllers. Each slave node controller is configured to perform lightweight data parsing on the received environmental data. If it is determined to be a real-time simple task, it generates dimming control instructions locally. If it is determined to be a non-real-time complex task, it offloads the task to the edge gateway. And after generating the dimming control command, according to the dimming control command, the linkage command is sent to the associated adjacent slave node controller through the preset short-range wireless direct connection protocol to realize the linkage of regional lighting; The edge collaboration layer includes at least one edge gateway, which is configured to: monitor the wireless channel status in real time and, when channel interference is detected, execute a dynamic wireless channel switching algorithm to switch to the optimal idle channel; receive non-real-time complex tasks offloaded by slave node controllers and perform dynamic task allocation and load balancing scheduling between slave node controllers and edge gateways based on a non-cooperative game-theoretic task scheduling algorithm; monitor the load status of each slave node controller in real time, and when the load of a slave node controller exceeds a dynamically adjusted offload threshold, reschedule some of the tasks allocated to that slave node controller; and periodically summarize the energy consumption and fault data uploaded by the slave node controllers, generate statistical reports, and upload the statistical reports to the cloud platform during periods of low network load.

[0017] Thirdly, the present invention proposes a terminal device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described edge-coordinated dimming methods.

[0018] Fourthly, the present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-described edge-coordinated dimming methods.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention effectively improves communication stability in complex environments by constructing a clearly defined dual-mode communication architecture and combining it with a centralized intelligent channel management mechanism. Specifically, terminal layer devices use wireless communication modules for low-latency direct connection and linkage, while the edge layer reliably sends key control commands through power line carrier communication modules, fully leveraging the complementary advantages of different communication modes. At the same time, the edge gateway continuously monitors the network environment and, upon detecting interference, can automatically and quickly switch the communication link to a better idle channel based on a comprehensive evaluation algorithm that includes factors such as idle time and frequency differences. This series of collaborative mechanisms enables the system to maintain a high success rate of command transmission when facing co-channel interference or complex electromagnetic environments.

[0020] (2) By endowing terminal devices with localized, lightweight data parsing and decision-making capabilities, this invention significantly shortens the response chain of linkage control. The terminal device can process the perceived data in real time and independently complete the judgment and instruction generation for simple trigger events, and then directly drive the actions of its own level and adjacent devices, thereby bypassing the multi-level uploading and callback process required in the traditional architecture. This not only achieves near-instantaneous lighting response to targets such as vehicles and pedestrians, meeting the needs of highly dynamic scenarios, but also enables the system to maintain basic intelligent linkage functions by relying on local strategies when network connectivity fluctuates or is interrupted, greatly enhancing the overall robustness of the system.

[0021] (3) This invention innovatively introduces a dynamic task scheduling mechanism based on game theory between the terminal and the edge. This mechanism rationally classifies computing tasks and dynamically adjusts the allocation ratio of tasks on the terminal and edge sides through real-time evaluation of core elements such as processing efficiency, resource consumption, and communication latency. This adaptive scheduling strategy can flexibly allocate computing resources according to the real-time load of the system, effectively utilizing the idle computing power of terminal devices and avoiding the edge gateway from becoming a performance bottleneck due to task overload, thereby achieving load balancing of terminal and edge resources and improving the overall stability and efficiency of the system when processing high-concurrency tasks.

[0022] (4) In view of the limited hardware resources of terminal devices, this invention designs an efficient data simplification and feature extraction algorithm, which makes it possible to realize key information processing and rapid decision-making on a low-cost microcontroller, breaking the terminal's complete dependence on the central node. Based on the above-mentioned precise and fast local control capabilities, the system can realize on-demand and zoned fine-grained lighting management, thereby achieving significant energy-saving effects. In addition, the local aggregation and preprocessing of operational data at the edge layer greatly reduces the direct occupation of cloud bandwidth and computing power, and reduces the overall operation and maintenance cost of the system. Attached Figure Description

[0023] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments, taken with reference to the accompanying drawings: Figure 1 This is a flowchart of a linkage dimming method based on edge-end collaborative computing; Figure 2 This is a schematic diagram of a linkage dimming system based on edge-end collaborative computing; Figure 3 This is a schematic diagram of the structure of a computer system suitable for implementing the embodiments of the present invention. Detailed Implementation

[0024] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] Figure 1 A flowchart of the coordinated dimming method based on edge-end collaborative computing is shown. (Refer to...) Figure 1 The present invention proposes a linkage dimming method based on edge-end collaborative computing, which includes the following: Before the system begins normal operation, the following initialization steps must be performed: The sensors in the front-end perception layer are powered on and complete self-tests and calibrations; the slave node controllers in the terminal control layer load local linkage strategies and initialize the dual-mode communication module (including scanning and locking the initial wireless operating frequency band); the edge gateways in the edge collaboration layer start the non-cooperative game-theoretic task scheduling algorithm and set initial load thresholds (e.g., the slave node controller MCU utilization threshold is 80%). After system initialization is complete, the following continuous operation process begins.

[0027] S1. The sensor collects environmental data in real time and transmits it to the slave node controller. The slave node controller performs lightweight data parsing on the received environmental data. If it determines that the task is a simple real-time task, it generates a dimming control command locally. If it determines that the task is a complex non-real-time task, it offloads the task to the edge gateway. The specific implementation process of this step is as follows: First, after completing power-on self-tests and calibrations, the sensors deployed in the front-end perception layer enter a continuous working state, collecting environmental data in real time within their coverage area, especially information on moving targets (such as vehicles and pedestrians). This information is typically in the form of raw data containing dimensions such as position, speed, and direction. The sensors transmit the collected environmental data in real time to their respective terminal control layer slave node controllers via fieldbuses such as RS485 at a rate of, for example, 9600bps. This data transmission process typically takes milliseconds.

[0028] Subsequently, the slave node controller receives data and initiates its core lightweight data parsing process. This process is specifically optimized for the resource-constrained microcontrollers (MCUs, such as the ARM Cortex-M0 with a clock speed of 48MHz F_mcu) used by the slave node controller, aiming to quickly complete data understanding and task decision-making with extremely low computational overhead. The entire parsing process consists of three closely linked sub-steps: S11. Data Filtering: The node controller performs rapid verification of the incoming raw environmental data based on preset physical rules or validity range thresholds. For example, the system can preset a speed limit (e.g., 120 km / h). Moving target data reporting speeds exceeding this threshold is directly identified as abnormal or invalid data and discarded. This step aims to remove obvious noise or erroneous data, ensuring the input quality for subsequent processing and outputting verified, valid data.

[0029] S12. Feature Extraction: The node controller efficiently extracts pre-defined key feature information used to drive dimming decisions from the filtered valid data. This key feature information typically includes the target's presence status (e.g., whether there are vehicles / pedestrians), target location (vehicle / pedestrian location), and target movement direction (vehicle / pedestrian direction of approach). By retaining only these core features, the amount of data to be processed is significantly reduced. The data compression ratio (C) can be measured by the following formula:

[0030] Where V raw V represents the amount of raw data from the sensor. comp This refers to the amount of extracted key feature data. In one embodiment of the invention, the compression rate can reach over 60%, significantly reducing storage and computational burden.

[0031] S13. Task Judgment and Command Generation / Unloading: The node controller quickly matches the extracted key feature information (target presence, location, direction) with the linkage strategy pre-stored in its local memory (such as Flash). The linkage strategy defines the response rules for different scenarios, such as "when a vehicle is detected coming from direction A, dim the brightness to 100%".

[0032] Preferably, the linkage strategy adopts a hierarchical storage scheme: basic strategies (such as default dimming parameters) are permanently stored in non-volatile memory (such as Flash); temporary strategies (such as adjustment plans for holidays or special events) are stored in volatile memory (such as RAM) and are lost after power failure. To handle strategy conflicts, the system adopts a priority determination mechanism: temporary strategies have higher priority than basic strategies. Temporary strategies must be bound to a clear effective time and applicable area, and only override basic strategies within a limited scope. When a temporary strategy expires or is manually terminated, the strategy data in RAM is automatically cleared from the node controller, and the execution of the basic strategy in Flash is resumed without manual intervention. Strategy conflict events and execution results are recorded and synchronized to the edge gateway for operation and maintenance traceability.

[0033] If the matching result meets the criteria for a real-time simple task (e.g., rule matching is successful, and the task requires extremely low latency and simple computation), the slave node controller immediately generates the corresponding dimming control command locally and prepares for execution. This local decision-making and command generation process is rigorously optimized, and the execution time (Texec) can be controlled to, for example, within 30 milliseconds. The MCU resource utilization (U) of the entire parsing process can be estimated using the following formula:

[0034] Where T exec The execution time of a single lightweight algorithm (in milliseconds); F mcu For slave node MCU master frequency (unit: MHz, slave node: 48MHz); T cycle The task scheduling period for slave nodes (unit: ms, default 50ms). Through algorithm optimization, U can be kept at a low level (e.g., below 60%), thus leaving sufficient resources for other concurrent tasks.

[0035] If the matching result meets the criteria for non-real-time complex tasks (e.g., requiring complex processing such as historical data analysis or energy efficiency calculation, or involving collaboration across multiple slave node controllers), the slave node controller will not perform local complex calculations. Instead, it will package the current valid data along with the necessary processing context information as a calculation task and offload it to the upper-layer edge gateway through the communication module, requesting it to process the task.

[0036] At this point, step S1 is complete. The system has achieved intelligent processing and task allocation of preliminary perception data on the terminal side, laying the foundation for subsequent rapid linkage or collaborative computing.

[0037] Continue to refer to Figure 1 The proposed method for coordinated dimming based on edge-end collaborative computing further includes the following steps: S2. After generating the dimming control command, the slave node controller sends a linkage command to the associated adjacent slave node controllers through a preset short-range wireless direct connection protocol to realize the linkage of area lighting. At the same time, the edge gateway monitors the wireless channel status in real time and executes a wireless channel dynamic switching algorithm to switch to the optimal idle channel when channel interference is detected.

[0038] This step involves two concurrent core processes: direct connection and linkage between nodes and communication optimization led by the edge gateway. The specific implementation process is as follows.

[0039] 1. Direct connection and linkage process from the node controller When a slave node controller determines in step S13 that a real-time simple task is being performed and generates a local dimming control command, it will immediately activate a direct connection linkage mechanism to achieve coordinated lighting in the area (such as pre-lighting several streetlights ahead). This mechanism is based on a preset custom short-range wireless direct connection protocol (e.g., based on an RF433 module). The linkage command is encapsulated into a specific frame structure for transmission. This frame structure includes at least: a start bit (1 bit), a target slave node address bit (8 bits, addressable for up to 128 associated nodes), a command bit (8 bits, carrying the dimming command), and a check bit (1 bit).

[0040] The slave node controller (initiating node) that generates the instruction sends the encapsulated linkage instruction directly to one or more adjacent slave node controllers (target nodes) through its wireless communication submodule, based on a preset neighbor association table, without needing to go through an edge gateway or cloud relay. Upon receiving the instruction, the target node parses it and immediately executes the corresponding dimming action. Because this process avoids multi-level network transmission, the end-to-end linkage response time from the initiating node generating the instruction to the target node starting execution can be shortened to less than 100 milliseconds, achieving an instantaneous area lighting effect where "lights follow the vehicle."

[0041] 2. Dynamic switching process of wireless channels in edge gateways To ensure the quality of the wireless link for the aforementioned direct connection and daily communication, the edge gateway continuously monitors and optimizes the wireless channel status within its management scope. This process specifically includes the following sub-steps: S21. Interference Determination and Triggering: The edge gateway monitors the signal quality of the managed wireless channel (current operating channel) in real time through its wireless communication module. The core indicator is the signal-to-noise ratio (SNR). Based on the monitoring results, the edge gateway calculates the channel interference intensity (I) of the current channel. This intensity can be quantified using the following formula: I = 1 - SNR / SNR max Where SNR is the real-time monitored signal-to-noise ratio (unit: dB), SNR max The theoretical maximum signal-to-noise ratio (SNR) for the adopted wireless frequency band (e.g., RF433) is set to 20 dB. The calculated I value ranges from 0 to 1. The system presets a high interference threshold (e.g., 0.7). When the calculated I value exceeds this threshold, the edge gateway determines that there is high interference on the current channel and immediately triggers the channel switching procedure.

[0042] S22. Idle Channel Scan: After the handover is triggered, the edge gateway controls its wireless module to perform a rapid scan on multiple preset candidate frequency bands (e.g., 17 RF433 bands). Based on preset idle determination rules (e.g., if no co-channel signal is detected on a channel for 100 milliseconds, it is determined to be idle), the availability of each candidate frequency band is evaluated. All candidate frequency bands determined to be idle and their corresponding idle duration (T) are recorded. idle ).

[0043] S23. Optimal Channel Decision: For all available candidate frequency bands identified in step S22, the edge gateway will perform a comprehensive evaluation to select the optimal target channel. The evaluation considers not only the channel's availability but also the interference risk with neighboring devices after handover. Therefore, a comprehensive score (S) is introduced for quantitative evaluation, calculated as follows: S = α T idle - β D adj Among them, T idle D represents the idle time of the channel (in milliseconds). adj This is the center frequency difference (in MHz) between the candidate channel and the channel currently used by the adjacent slave node controller. α and β are preset weighting coefficients used to balance channel quality and handover impact. Typically, α (e.g., 0.7) is set greater than β (e.g., 0.3) to prioritize more idle and stable channels. The edge gateway calculates the S-value for each idle candidate frequency band and selects the idle channel with the highest overall score S as the target channel for this handover.

[0044] S24. Cross-domain Coordination Confirmation: To avoid new co-channel interference to networks managed by other edge gateways in adjacent areas caused by the selected target channel, the edge gateway performs cross-domain coordination before making the final decision. Specifically, the edge gateway reports the selected target channel information to the cloud platform through its WAN communication module (such as a 4G module). The cloud platform maintains the channel usage information of each edge gateway and feeds back the channel information of relevant adjacent edge gateways to the requesting gateway. Based on the feedback information, the edge gateway determines whether switching to the target channel will cause significant interference to adjacent gateways. If the determination result is no impact or the impact is controllable, the switch is confirmed; otherwise, the process returns to step S23 to select a suboptimal channel for re-evaluation.

[0045] S25. Switching Command Execution: Once the target channel is confirmed, the edge gateway broadcasts a frequency band switching command to all slave controllers within its jurisdiction via the common channel (e.g., channel 0) of its power line carrier communication (PLC) submodule. This command includes the target channel information. The broadcast command leverages the characteristics of PLC communication—immunity to wireless interference and cross-transformer frequency synchronization—to ensure reliable delivery of the command. Upon receiving the command, the edge gateway itself and all slave controllers synchronously switch the operating frequency band of their wireless communication modules to the target channel. The entire switching process (from trigger to completion) can be controlled within 50 milliseconds, thereby minimizing the impact on ongoing communication services.

[0046] Through the aforementioned concurrent direct connection linkage and dynamic channel switching mechanism, the system achieves millisecond-level regional lighting linkage response while ensuring high reliability and adaptive anti-interference capability of the wireless communication link.

[0047] Continue to refer to Figure 1 The proposed method for coordinated dimming based on edge-end collaborative computing further includes the following steps: S3. The edge gateway receives non-real-time complex tasks offloaded by the slave node controllers and performs dynamic task allocation and load balancing scheduling between the slave node controllers and the edge gateway based on a non-cooperative game-theoretic task scheduling algorithm. The edge gateway monitors the load status of each slave node controller in real time. When the load of a slave node controller exceeds the dynamically adjusted offload threshold, some of the tasks allocated to that slave node controller will be rescheduled.

[0048] This step is the core of realizing system-end collaborative computing and adaptive load balancing, and its specific implementation process includes the following sub-steps: S31. The edge gateway receives non-real-time complex tasks unloaded from the node controller and prioritizes them according to their real-time requirements.

[0049] The edge gateway continuously listens to its communication interface, receiving non-real-time complex tasks proactively offloaded by various slave node controllers within its jurisdiction. These tasks are typically unsuitable for real-time processing on resource-constrained terminals, such as: statistical analysis and trend analysis of energy consumption data for historical periods (e.g., the past hour); fault diagnosis and prediction of equipment operating status (e.g., abnormal current or voltage); and optimization calculations of lighting patterns that require coordination of data from multiple road segment slave node controllers.

[0050] Upon receiving a task, the edge gateway first performs initialization processing and priority classification based on preset classification rules. The core basis for classification is the task's real-time requirements and business importance. For example, periodic statistical tasks with low real-time requirements but high computational load can be classified as lower priority; while diagnostic tasks that may indicate potential faults, even if not real-time, can be assigned higher priority to obtain analysis results as quickly as possible. This classification result provides an important basis for differentiated processing in subsequent scheduling decisions.

[0051] S32. Based on the non-cooperative game-theoretic task scheduling algorithm, evaluate the first benefit of non-real-time complex tasks when processed by the slave node controller and the second benefit when processed by the edge gateway.

[0052] This step is the core decision-making process of the scheduling algorithm. The edge gateway treats itself and each of its subordinate slave node controllers as participants in a task processing ecosystem and runs a non-cooperative game-theoretic task scheduling algorithm. This algorithm quantifies the comprehensive value of tasks processed on different nodes by defining and calculating payoffs, providing a basis for dynamic allocation.

[0053] First Benefit (Slave Node Controller Processing Benefit) Evaluation: When considering assigning a non-real-time complex task to a specific slave node controller, the algorithm evaluates its first benefit. This benefit evaluation is based on the task processing efficiency and resource consumption of the corresponding slave node controller. In a specific embodiment, the first benefit (denoted as R) s This can be quantified using the following formula: R s = λ E t - μ U mcu Among them, E t The task processing efficiency of this slave node controller can be obtained by statistically analyzing its average rate of successfully processing similar tasks recently (e.g., tasks per second); U mcu R represents the current microcontroller resource utilization rate of the slave node controller (in %), which is provided by the status information packets reported by the slave node controller to the edge gateway periodically (e.g., every 10 seconds); λ and μ are preset weighting coefficients (e.g., 0.6 and 0.4 respectively) used to balance the importance of processing efficiency and resource consumption in the benefit assessment. s The higher the value, the higher the overall benefit (cost-effectiveness) of the task being processed on this slave node controller.

[0054] Secondary benefit (edge ​​gateway processing benefit) evaluation: When considering the task being processed by the edge gateway itself, the algorithm evaluates its secondary benefit. This benefit evaluation is based on the number of tasks processed by the edge gateway and the end-edge communication latency. In one specific embodiment, the secondary benefit (denoted as R) e This can be quantified using the following formula: R e = γ N t - δ D comm Where, N t This represents the number of pending and currently being processed tasks at the edge gateway, reflecting the real-time load of the edge gateway itself; D comm R represents the edge-to-edge communication latency (unit: ms), which is the average time taken for data transmission between the node controller and the edge gateway, and can be obtained through network probing; γ and δ are preset weighting coefficients (for example, both can be 0.5) used to balance task processing volume and communication overhead. e The higher the value, the greater the overall benefit of the task being processed at the edge gateway.

[0055] S33. Perform iterative scheduling at a preset period. Based on the comparison between the first benefit and the second benefit, and in combination with the priority of the task, dynamically adjust the allocation ratio of non-real-time complex tasks between the slave node controller and the edge gateway until a balance between load and benefit is achieved.

[0056] The edge gateway initiates an iterative scheduling round at a fixed preset period (e.g., every 10 seconds). Each iteration is a process of seeking the optimal task allocation scheme for the system. Initial allocation and profit calculation: based on the current state of each node in the system (from the node controller's U...) mcu E t N of the edge gateway t D comm The algorithm takes a task queue and, according to an initial strategy (such as tentatively allocating some tasks to the less loaded slave node controllers), forms an allocation scheme. Then, based on the profit formula from step S32, it calculates the profit of each involved node under the current scheme (for tasks allocated to slave node controllers, calculate their R...). s For tasks assigned to the edge gateway, calculate its R. e ), and then calculate the average first revenue (R) of all slave node controllers. s _avg) and the second benefit of edge gateways (R e ).

[0057] Return Comparison and Dynamic Adjustment: The core of the algorithm is comparing R. s _avg and Re If the difference between the two is (ΔR = |R) e - R s If the value of _avg| is greater than a preset threshold (e.g., 0.1), the system is considered to be unbalanced and task adjustments are needed to improve overall returns.

[0058] If R e Significantly higher than R s If _avg means the marginal benefit of the task being processed by the edge gateway is higher, then the algorithm will attempt to reassign some tasks from the edge gateway queue, combined with task priority (prioritizing the migration of lower priority tasks), to the current R. s Slave node controllers with lower values ​​and no overload limits are used to improve overall revenue.

[0059] Conversely, if R s _avg is significantly higher than R e This indicates that the slave node controller has a greater advantage in processing, and the algorithm will delegate some of the slave node controller's tasks (especially R). s Tasks on nodes with high values ​​are migrated back to the edge gateway.

[0060] Iterative Convergence: After the above adjustments, a new allocation scheme is formed and the payoff is recalculated. This process is repeated iteratively. Iterative scheduling continues until preset convergence conditions are met. These convergence conditions ensure that the system reaches a stable equilibrium state, for example: In multiple iterations, the value of ΔR is always less than the preset threshold.

[0061] In multiple iterations, the distribution ratio of non-real-time complex tasks between the slave node controller and the edge gateway changes less than a first threshold (e.g., 5%).

[0062] When all convergence conditions are met, the system is considered to have reached a state of load-reward balance, and the current round of iterative scheduling ends. The system then executes task processing according to the final allocation scheme. During the iteration process, if the above convergence conditions are not met even after reaching the preset maximum number of iterations (e.g., 10), the system will trigger a forced convergence mechanism, which will adopt the current task allocation ratio and restart the iteration process in the next scheduling cycle to avoid resource waste.

[0063] S34. The edge gateway monitors the load status of each slave node controller in real time and dynamically adjusts the offloading threshold according to the average load of slave node controllers in the region. When the load of a slave node controller exceeds the current offloading threshold, the rescheduling mechanism is triggered to reallocate part of the tasks undertaken by the slave node controller to other idle edge gateways.

[0064] To ensure system stability during dynamic changes and prevent overload of individual nodes, the edge gateway performs real-time load monitoring and protective rescheduling in parallel, in addition to periodic iterative scheduling.

[0065] Dynamically Adjusted Unloading Threshold: The system does not use a fixed load threshold as the overload judgment standard, but instead employs a dynamically adjusted unloading threshold (Th). This threshold adaptively fluctuates based on the overall system load condition, and its calculation formula is as follows: Th = Th0 + η (U avg - U s ) Where Th0 is the preset initial unload threshold (e.g., 80% MCU utilization); U avg is the average MCU utilization rate of all slave node controllers in the region, reflecting the overall load level; Us is the MCU utilization rate of a specific slave node controller currently being monitored; η is a preset adjustment coefficient (e.g., 0.3) used to control the sensitivity of the threshold to load differences.

[0066] This formula means that when the load (U) of a slave node controller... s ) is much lower than the regional average load (U avg When the load on a node is close to or exceeds the average load, its offload threshold (Th) will be increased accordingly, enabling it to handle more tasks; conversely, when the load on the node is close to or exceeds the average load, its offload threshold will be decreased, making it more sensitive and more likely to trigger the protection mechanism.

[0067] Triggering the rescheduling mechanism: The edge gateway continuously receives status information packets periodically reported by each slave node controller. These packets include information such as the current microcontroller resource utilization rate (U... s In addition to the node ID, it can also include the recent task processing efficiency (E). t ) Statistical information, edge communication delay (D comm The sampled values, along with markers indicating whether the load exceeds internal thresholds, provide comprehensive data for revenue calculation and overload assessment. When the real-time load U of a particular slave node controller is monitored... s When the current unloading threshold Th, which is dynamically calculated for it, is exceeded, the node is determined to be at risk of overload. At this time, the system will immediately trigger the rescheduling mechanism, interrupting or independently of the current periodic iterative scheduling process.

[0068] This mechanism selects a subset of assigned but incomplete tasks from the overloaded slave node controllers (typically prioritizing tasks with minimal impact on real-time performance based on task priorities in step S31), and reassigns these tasks. The primary target node for reassignment is other idle edge gateways (i.e., edge-to-edge collaboration). If this is not possible, other slave node controllers with lower loads within the region can be considered. This immediate intervention quickly mitigates localized overload risks and ensures the stable operation of the entire system.

[0069] Through the complete process of steps S31 to S34 above, the present invention realizes all-round end-edge collaborative task scheduling, from intelligent task classification, refined benefit evaluation based on game theory, periodic iterative optimization, to real-time load monitoring and elastic rescheduling, which significantly improves system resource utilization and overall stability.

[0070] Continue to refer to Figure 1 The proposed method for coordinated dimming based on edge-end collaborative computing further includes the following steps: S4. The edge gateway periodically summarizes the energy consumption and fault data uploaded from the node controller, generates statistical reports, and uploads the statistical reports to the cloud platform during periods of low network load.

[0071] This step is crucial for the system's data aggregation, operation and maintenance management, and cloud collaboration. It aims to achieve a macro-level understanding of the system's operational status, energy efficiency assessment, and fault early warning, while optimizing network resource utilization. The specific implementation process is as follows: During normal system operation, each slave node controller not only performs dimming control and calculation tasks, but also continuously monitors and records its local key operational data. This data mainly includes two categories: 1. Energy consumption data: Records real-time and historical power consumption information of the controlled streetlights, which may include power consumption at different times and under different brightness levels.

[0072] 2. Fault data: Record the abnormal state of the equipment itself, such as communication abnormalities, sensor failures, drive circuit failures, and continuous abnormal data records that were filtered out in step S1.

[0073] Each slave node controller proactively reports this data to its associated edge gateway via its dual-mode communication module (usually prioritizing the wireless submodule due to its smaller data transmission volume and less stringent real-time requirements) at preset intervals (e.g., every minute or after each major operation) or upon the occurrence of a fault event. Data reporting follows a lightweight principle, typically undergoing initial formatting or aggregation by the slave node controller to reduce transmission overhead.

[0074] As a data aggregation point for a region, the edge gateway continuously receives and caches data reported from all slave node controllers within its jurisdiction. To achieve efficient data management and reduce real-time pressure on the cloud, the edge gateway employs a periodic batch processing and aggregation mechanism. 1. Data Aggregation and Storage: The edge gateway initiates a data aggregation task at a fixed aggregation period (e.g., hourly). Within this period, it classifies, cleans (e.g., removes duplicate reports, corrects format errors), and aggregates all cached energy consumption and fault data from slave node controllers. For energy consumption data, it may calculate statistical indicators such as total electricity consumption, average power, and peak power for each node and segment within that hour; for fault data, it categorizes and counts them, and associates the occurrence time with the node location.

[0075] 2. Generate Statistical Reports: Based on the aggregated data, the edge gateway runs its built-in report generation engine to generate structured statistical reports according to predefined templates. Report content typically includes: Energy efficiency analysis report: Displays total regional energy consumption, year-on-year / month-on-month changes, energy consumption per unit of illuminance, etc., within a specified time period.

[0076] Equipment Health Status Report: Lists the online status of all active slave node controllers and their sensors, recent fault statistics (such as fault type and frequency of occurrence), and whether key parameters exceed limits.

[0077] Operation and maintenance recommendations: Based on the data analysis results, preliminary diagnostic or prompt information may be generated, such as "The energy efficiency of the lights in a certain section is low, and it is recommended to check" or "The communication stability of a certain node has decreased".

[0078] The generated report files (which may be in JSON, CSV, or PDF format) are temporarily stored on the edge gateway's local storage device.

[0079] 3. Cloud upload during low-load periods: To reduce interference with critical business communications (such as real-time control commands and channel switching commands) and to leverage the potentially lower bandwidth costs during network off-peak hours, the edge gateway does not upload reports in real time. Instead, it implements intelligent upload scheduling.

[0080] Network load assessment: The edge gateway monitors the network load status of its uplink (such as the exit to the metropolitan area network) in real time or periodically, for example by monitoring bandwidth utilization, network latency, or according to preset time period rules (for example, the default definition is that 1:00 am to 5:00 am is a low load period).

[0081] Triggered Upload: When the network is detected to be in a low-load period (e.g., bandwidth utilization is below 20%, or it enters a preset early morning off-peak period), the edge gateway automatically initiates an upload task. It securely transmits locally stored statistical report files to the cloud platform via its WAN communication module (such as 4G / 5G or Ethernet).

[0082] Cloud-based processing: After receiving reports, the cloud platform stores them in a database, which can be used for long-term historical data analysis, cross-regional energy efficiency comparisons, and the generation of more comprehensive operation and maintenance dashboards, providing decision support for managers. In addition, based on deeper analysis, the cloud platform may also release updated dimming strategies, fault handling plans, or system parameter optimization suggestions to lower levels, thus forming an optimization closed loop of "end-edge-cloud".

[0083] By implementing step S4, this invention achieves efficient local aggregation and preprocessing of operational data, with cloud synchronization only occurring when the network is idle. This significantly reduces reliance on real-time cloud computing and storage resources, lowers the cost and pressure of continuous network transmission, and ensures the integrity and availability of operational data, providing crucial support for the sustainable and low-cost operation of large-scale smart lighting systems. The entire methodology, from sensing, computation, and communication optimization to data management, constitutes a complete, self-consistent, and efficient intelligent linkage dimming solution.

[0084] Further reference Figure 2 As an implementation of the above method, this invention also proposes an embodiment of a linkage dimming system 200 based on edge-end collaborative computing, which can be applied to various electronic devices. The linkage dimming system 200 based on edge-end collaborative computing includes the following modules: The front-end perception layer 210 includes at least one sensor, which is configured to collect environmental data in real time and transmit it to the slave node. The terminal control layer 220 includes multiple slave node controllers. Each slave node controller is configured to perform lightweight data parsing on the received environmental data. If it is determined to be a real-time simple task, it generates a dimming control command locally. If it is determined to be a non-real-time complex task, it offloads the task to the edge gateway. And after generating the dimming control command, according to the dimming control command, the linkage command is sent to the associated adjacent slave node controller through the preset short-range wireless direct connection protocol to realize the linkage of regional lighting; The edge collaboration layer 230 includes at least one edge gateway, which is configured to: monitor the wireless channel status in real time and, when channel interference is detected, execute a dynamic wireless channel switching algorithm to switch to the optimal idle channel; receive non-real-time complex tasks offloaded by slave node controllers and perform dynamic task allocation and load balancing scheduling between slave node controllers and edge gateways based on a non-cooperative game-theoretic task scheduling algorithm; monitor the load status of each slave node controller in real time, and when the load of a slave node controller exceeds a dynamically adjusted offload threshold, reschedule some of the tasks allocated to that slave node controller; and periodically summarize the energy consumption and fault data uploaded by the slave node controllers, generate statistical reports, and upload the statistical reports to the cloud platform during periods of low network load.

[0085] Thirdly, the present invention proposes a terminal device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described edge-coordinated dimming methods.

[0086] Fourthly, the present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-described edge-coordinated dimming methods.

[0087] The following is for reference. Figure 3 It shows a schematic diagram of the structure of a computer system 300 suitable for implementing terminal devices or servers of the present invention. Figure 3 The terminal device or server shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.

[0088] like Figure 3 As shown, the computer system 300 includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 302 or programs loaded from storage section 308 into random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the computer system 300. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0089] The following components are connected to I / O interface 305: an input section 306 including a keyboard, mouse, etc.; an output section 307 including a liquid crystal display (LCD) and speakers, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN card and a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to I / O interface 305 as needed. A removable medium 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 310 as needed so that computer programs read from it can be installed into storage section 308 as needed.

[0090] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit (CPU) 301, it performs the functions defined in the methods of this invention. It should be noted that the computer-readable medium described in this invention can be a computer-readable signal medium or a computer-readable medium or any combination thereof. The computer-readable medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on a computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0091] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0092] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0093] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention is not limited to the specific combination of the above-described technical features, but also includes other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in this invention.

Claims

1. A linkage dimming method based on edge-end collaborative computing, characterized in that, Includes the following steps: S1. The sensor collects environmental data in real time and transmits it to the slave node controller. The slave node controller performs lightweight data parsing on the received environmental data. If it is determined to be a real-time simple task, it generates a dimming control command locally. If it is determined to be a non-real-time complex task, it offloads the task to the edge gateway. S2. After generating a dimming control command, the slave node controller sends a linkage command to the associated adjacent slave node controllers through a preset short-range wireless direct connection protocol to realize regional lighting linkage. At the same time, the edge gateway monitors the wireless channel status in real time and executes a wireless channel dynamic switching algorithm to switch to the optimal idle channel when channel interference is detected. S3. The edge gateway receives non-real-time complex tasks offloaded from the slave node controller and performs dynamic task allocation and load balancing scheduling between the slave node controller and the edge gateway based on a non-cooperative game-theoretic task scheduling algorithm. The edge gateway monitors the load status of each slave node controller in real time. When the load of a slave node controller exceeds the dynamically adjusted offload threshold, some of the tasks assigned to that slave node controller will be rescheduled. S4. The edge gateway periodically summarizes the energy consumption and fault data uploaded from the node controller, generates statistical reports, and uploads the statistical reports to the cloud platform during periods of low network load.

2. The linkage dimming method based on edge-end collaborative computing according to claim 1, characterized in that, An initialization step is included before step S1, specifically: The sensors in the front-end perception layer complete power-on self-test and calibration, and output standardized digital signals to the slave node controller in the terminal control layer; the slave node controller loads the local linkage strategy and initializes the dual-mode communication module; the edge gateway in the edge collaboration layer starts the task scheduling algorithm and sets the load threshold. The dual-mode communication module includes a power line carrier communication submodule and a wireless communication submodule; the slave node controller directly connects with adjacent slave node controllers through the wireless communication submodule; and the edge gateway sends control commands to the slave node controllers through the power line carrier communication submodule.

3. The linkage dimming method based on edge-end collaborative computing according to claim 1, characterized in that, In step S1, the slave node controller performs lightweight data parsing on the received data. If it is determined to be a real-time simple task, it generates a dimming control command locally. If it is determined to be a non-real-time complex task, it offloads the task to the edge gateway. Specifically, this includes the following sub-steps: S11. Based on preset physical rules or range thresholds, verify the environmental data collected by the sensor, remove abnormal data, and obtain valid data. S12. Extract preset key feature information for generating dimming instructions from the effective data. The key feature information includes the target existence state, target position, and target movement direction. S13. Match the key feature information with the linkage strategy pre-stored locally on the slave node controller; if the matching result meets the real-time simple task judgment conditions, generate a dimming control command locally. If the matching result meets the criteria for non-real-time complex tasks, the current data and processing context will be offloaded to the edge gateway as a task.

4. The linkage dimming method based on edge-end collaborative computing according to claim 1, characterized in that, In step S2, the edge gateway monitors the wireless channel status in real time, and when channel interference is detected, executes a dynamic wireless channel switching algorithm to switch to the optimal idle channel, specifically including the following sub-steps: S21. The edge gateway monitors the signal quality of the current wireless channel in real time and calculates the channel interference intensity based on the monitoring results. When the channel interference intensity exceeds a preset high interference threshold, channel switching is triggered. S22. The edge gateway scans multiple preset candidate wireless frequency bands and evaluates the idle status of each candidate frequency band according to preset idle determination rules. S23. The edge gateway combines the frequency differences between each frequency band and the channels used by the adjacent slave node controllers, and selects the idle channel with the best comprehensive evaluation result from the candidate frequency bands as the target channel. S24. The edge gateway reports the target channel information to the cloud platform through the wide area network communication module and obtains the channel information of adjacent edge gateways to determine whether switching to the target channel will cause interference to the adjacent edge gateways; if not, proceed with the following steps. S25. The edge gateway sends a frequency band switching command to the slave node controller under its jurisdiction through the common channel of the power line carrier communication module, and controls its own and the slave node controller's wireless communication modules to switch to the target channel.

5. The linkage dimming method based on edge-end collaborative computing according to claim 4, characterized in that, In step S21, the channel interference intensity is calculated and determined based on the ratio of the real-time signal-to-noise ratio of the wireless channel to the theoretical maximum signal-to-noise ratio of its frequency band. In step S23, the comprehensive evaluation is performed as follows: for each candidate frequency band evaluated as idle, the weighted difference between the channel idle duration and the channel frequency difference used by the adjacent slave node controller is calculated, wherein the weighting coefficient of the idle duration is greater than the weighting coefficient of the frequency difference. The idle channel with the highest weighted difference is selected as the target channel.

6. The linkage dimming method based on edge-end collaborative computing according to claim 1, characterized in that, Step S3 specifically includes the following sub-steps: S31. The edge gateway receives the non-real-time complex tasks offloaded from the node controller and prioritizes them according to the real-time requirements of the tasks. S32. Based on the non-cooperative game task scheduling algorithm, evaluate the first benefit of the non-real-time complex task when it is processed by the slave node controller and the second benefit when it is processed by the edge gateway; wherein, the evaluation of the first benefit is based on the task processing efficiency and resource consumption of the corresponding slave node controller, and the evaluation of the second benefit is based on the number of tasks processed by the edge gateway and the end-edge communication latency. S33. Perform iterative scheduling at a preset period, and dynamically adjust the allocation ratio of the non-real-time complex task between the slave node controller and the edge gateway based on the comparison between the first benefit and the second benefit, and in combination with the priority of the task, until a balance between load and benefit is achieved. S34. The edge gateway monitors the load status of each slave node controller in real time and dynamically adjusts the offloading threshold according to the average load of the slave node controllers in the region. When the load of a slave node controller exceeds the current offloading threshold, a rescheduling mechanism is triggered to reallocate some of the tasks undertaken by the slave node controller to other idle edge gateways.

7. The linkage dimming method based on edge-end collaborative computing according to claim 6, characterized in that, In step S33, the iterative scheduling continues until a preset convergence condition is met; the convergence condition includes: in multiple consecutive iterations, the change in the allocation ratio of non-real-time complex tasks between the slave node controller and the edge gateway is less than a first threshold, and the difference between the average revenue of the slave node controller and the revenue of the edge gateway is less than a second threshold.

8. A synchronized dimming system based on edge-to-edge collaborative computing, used to implement the method as described in any one of claims 1 to 7, characterized in that, The system includes: The front-end perception layer includes at least one sensor, which is configured to collect environmental data in real time and transmit it to the slave node; The terminal control layer includes multiple slave node controllers. Each slave node controller is configured to perform lightweight data parsing on the received environmental data. If it is determined to be a real-time simple task, it generates dimming control instructions locally. If it is determined to be a non-real-time complex task, it offloads the task to the edge gateway. And after generating the dimming control command, according to the dimming control command, a linkage command is sent to the associated adjacent slave node controller through a preset short-range wireless direct connection protocol to realize the linkage of regional lighting; The edge collaboration layer includes at least one edge gateway, which is configured to: monitor the wireless channel status in real time and, when channel interference is detected, execute a dynamic wireless channel switching algorithm to switch to the optimal idle channel; receive non-real-time complex tasks offloaded by slave node controllers and perform dynamic task allocation and load balancing scheduling between the slave node controllers and the edge gateway based on a non-cooperative game-theoretic task scheduling algorithm; monitor the load status of each slave node controller in real time, and when the load of a slave node controller exceeds a dynamically adjusted offload threshold, reschedule some of the tasks allocated to that slave node controller; and periodically summarize the energy consumption and fault data uploaded by the slave node controllers, generate statistical reports, and upload the statistical reports to the cloud platform during periods of low network load.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the linkage dimming method based on edge-to-edge collaborative computing as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the linkage dimming method based on edge-to-edge collaborative computing as described in any one of claims 1 to 7.

Images