Intelligent lighting integrated circuit cooperative control method based on edge computing

Abstract

The invention discloses an intelligent lighting integrated circuit cooperative control method based on edge computing, and the method comprises the following steps: deploying an edge computing unit, and collecting environment data in real time; utilizing an improved artificial fish swarm algorithm to generate a local control result; a lighting control instruction is generated according to the local optimal control result, and working state feedback is obtained; establishing a structured state space model which comprises an edge layer, a cloud layer and a data storage and management layer; a self-adaptive collaborative mapping function is introduced, and a global optimization control strategy and an increment parameter packet are generated; issuing the incremental parameter packet to an edge node, and dynamically balancing the parameter according to the adaptive fusion coefficient; and adjusting parameters and interlayer weights of the improved artificial fish swarm algorithm to obtain globally collaborative intelligent illumination. According to the invention, global cooperative control and adaptive optimization adjustment of intelligent illumination are realized through the improved artificial fish swarm algorithm and the adaptive cooperative mapping function.

Description

Technical Field

[0001] This invention relates to the field of intelligent lighting technology, and in particular to a collaborative control method for intelligent lighting integrated circuits based on edge computing. Background Technology

[0002] In modern smart lighting applications, with the continuous development of IoT and AI technologies, smart lighting has gradually become a key technology for improving lighting efficiency and user experience. Smart lighting collects environmental data in real time through sensors and uses advanced control algorithms to automatically adjust parameters such as brightness, on / off status, and color temperature of lighting equipment to achieve a balance between energy saving and comfort. Existing smart lighting is typically based on a centralized control architecture, with a central control unit managing all lighting nodes. While centralized control simplifies management and control, it still has significant problems in terms of response speed, communication load, and resource consumption. Especially in large-scale and high-density applications, centralized architecture struggles to meet the demands for real-time performance and efficiency.

[0003] Current smart lighting generally faces the following technical challenges: High response latency due to centralized architecture, especially with a large number of nodes or a large scale, places an excessive burden on the central control unit's computation and communication, affecting real-time performance and flexibility; existing smart lighting typically relies on preset control rules or thresholds for environmental monitoring and adjustment, resulting in a lack of adaptability and an inability to precisely adjust according to environmental changes, user needs, and energy consumption requirements; existing control methods are mostly static rules or simple threshold controls, unable to flexibly respond to different environmental changes and user needs, leading to energy waste and insufficient comfort; although existing methods collect data through multiple sensors, in the process of fusion and analysis of multi-sensor data, current technologies often rely on simple weighted averaging or local optimization, lacking in-depth exploration of the correlations and overall synergy between data, affecting the accuracy of control and optimization efficiency.

[0004] Therefore, how to provide a collaborative control method for intelligent lighting integrated circuits based on edge computing is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] One objective of this invention is to propose a collaborative control method for intelligent lighting integrated circuits based on edge computing. This invention fully utilizes edge computing technology, intelligent optimization algorithms, and multi-level data collaborative control strategies to design an intelligent lighting control strategy that integrates local optimization decision-making, global collaborative control, and adaptive optimization adjustment. The complete process from environmental data acquisition, local optimal control decision-making, state-space model construction to global optimization strategy generation is described in detail, highlighting the dynamic adjustment capabilities based on an improved artificial fish swarm algorithm and an adaptive collaborative mapping function. This method can significantly reduce the communication load and response latency in traditional centralized control architectures, improve the performance and energy efficiency of intelligent lighting in complex environments, and is particularly suitable for real-time control and optimization of large-scale intelligent lighting networks. It can provide efficient and reliable technical support for energy saving and lighting comfort in intelligent buildings, public facilities, and industrial applications.

[0006] According to an embodiment of the present invention, a collaborative control method for intelligent lighting integrated circuits based on edge computing includes: Edge computing units are deployed at each lighting node to collect environmental data in real time and construct local environmental status information for the node. Based on the collected environmental data, an improved artificial fish swarm algorithm is used to make local optimization decisions for each lighting node, generating a locally optimal control result that adapts to the current environment. Within the edge computing unit, lighting control commands are generated based on local optimal control results, the parameters of the lighting equipment are adjusted, the lighting control commands are executed, and feedback on the actual working status of the local lighting equipment is obtained. A structured state space model is established based on feedback from actual working status. The structured state space model includes an edge layer, a cloud layer, and a data storage and management layer. An adaptive collaborative mapping function is introduced between the edge layer and the cloud layer to facilitate information exchange, generate a global optimization control strategy and form an incremental parameter package, and store historical data in the data storage and management layer. Incremental parameter packages are distributed to each edge node. Each node dynamically balances the old and new parameters based on the adaptive fusion coefficient. When edge-cloud communication is interrupted, the edge node enters autonomous operation mode to maintain short-term state prediction and independent control. By leveraging the feedback mechanism of edge computing nodes, the lighting control effect is evaluated periodically. Based on the operating status, the parameters of the improved artificial fish swarm algorithm and the inter-layer mapping weights are dynamically adjusted to adaptively optimize the allocation of computing resources and control strategies. Historical data is then updated to achieve globally coordinated intelligent lighting and optimal energy efficiency.

[0007] Optionally, the edge computing unit includes a sensor module, a data processing unit, and a control execution unit.

[0008] Optionally, the environmental data includes light intensity, temperature and humidity, human activity, and energy consumption data.

[0009] Optionally, the local environment status information of the building node includes: An edge computing node is deployed at each lighting node. The edge computing unit connects to multiple sensor modules and collects environmental data in real time through the sensors, including light sensors, temperature and humidity sensors, human activity sensors, and energy consumption sensors. Each sensor collects data periodically or continuously and transmits the collected raw data to the data processing unit of the edge computing unit. The collected environmental data is input into the data processing unit for noise reduction, filtering, and normalization. The processed environmental data is integrated into the local environmental status information of the node, which serves as the real-time status input of the lighting node. By fusing data from multiple sensors, a multi-dimensional dataset that comprehensively reflects the current environmental status of the lighting node is generated and transmitted to the control execution unit to generate lighting control commands adapted to the current environment.

[0010] Optionally, generating locally optimal control results adapted to the current environment includes: Initialize the artificial fish swarm, setting the position and speed of each fish. The position represents the control parameter of the lighting node, and the speed represents the change of the control parameter. The initial control parameters include lighting brightness, on / off status, color temperature, and lighting area. The position of each fish represents the combination of control parameters of the lighting node, and the speed represents the change of the control parameter. During initialization, the control parameter position is the initial value of brightness, on / off status, color temperature, and lighting area, and the speed is the initial change of the control parameter. A local search is performed on each artificial fish, specifically: Calculate the distance between the current fish position and the target position, and calculate the speed of the search direction based on the difference between the current local optimum and the global optimum. The search speed is determined by the inertia weight, the difference between the local optimal position and the current fish position, and the difference between the global optimal position and the current fish position. The search speed is equal to the speed of the previous moment multiplied by the inertia weight, plus the difference between the local optimal position and the current fish position multiplied by the local learning factor, plus the difference between the global optimal position and the current fish position multiplied by the global learning factor. The inertia weight controls the influence of the previous velocity, while the local and global learning factors control the influence of the local and global optimal positions on the search direction; A fuzzy dynamic decision factor is introduced to dynamically adjust the weighting coefficients in the speed calculation. The fuzzy dynamic decision factor is adjusted in real time based on environmental data, user needs and energy efficiency requirements. The fuzzy dynamic decision factor adjusts the weighting coefficients in the speed calculation, optimizes the position update process of each fish, and limits the position and speed update range of each fish. The position and speed of the fish swarm are updated through local and global searches. The fitness of each fish is calculated based on the fitness function, which includes lighting effect and energy efficiency assessment. The weight coefficients are adjusted according to the needs. The fitness function comprehensively considers the current lighting brightness and energy consumption data, adjusts the weight of lighting brightness to optimize the lighting effect, and controls energy efficiency to reduce energy consumption. Based on the fitness calculation results, the optimal control parameters are selected, the global optimal position and the local optimal position are updated, and the iterative search continues until the stopping conditions are met. The stopping conditions include the fitness reaching a preset threshold or reaching the maximum number of iterations. When the stopping conditions are met, the optimal control parameters are output as the final control strategy for lighting control.

[0011] Optionally, establishing a structured state-space model based on actual working state feedback includes: The edge layer takes real-time collected environmental data as input and uses local state space equations to estimate the state of the current lighting node in real time, generating local state information of the node. The local state information generated by each edge node is uploaded to the cloud layer, and a global state space equation is constructed in the cloud layer. By aggregating the local state information from each edge node, a global optimization control strategy is formed. The data storage and management layer stores historical data, control policies, status information, and device logs for each edge node; Based on the calculation results of the local state space equation and the global state space equation, the control strategy of each node is dynamically adjusted, and the nodes are optimized and controlled in real time. Based on the real-time estimation results of the edge layer and the global optimization strategy of the cloud layer, the control parameters of each lighting node are adjusted.

[0012] Optionally, the step of introducing an adaptive collaborative mapping function between the edge layer and the cloud layer for information interaction, generating a global optimization control strategy, and forming an incremental parameter package includes: The adaptive cooperative mapping function receives local state information from the edge layer and dynamically adjusts it based on real-time input data to generate a simplified state summary. The adaptive cooperative mapping function consists of three main processing parts: an input processing part, a dynamic mapping part, and an output generation part, specifically: The input processing section receives local state information from the edge layer and adjusts the data preprocessing strategy according to the current load, real-time requirements, and environmental changes. The dynamic mapping part maps the input data into a simplified state summary using a weighted averaging method based on a real-time adjusted processing strategy. The output generation section converts the simplified state summary into a format suitable for uploading and sends it to the cloud layer; The cloud layer uses a collaborative attention mechanism to automatically adjust the weights of the state summaries of each node. The collaborative attention mechanism adaptively adjusts the weights based on the current environmental state, real-time feedback and needs of each node, generating a global optimization control strategy. The cloud layer calculates and updates the control parameters of each lighting node based on the global optimization control strategy. It then combines the updated control parameter values ​​of each lighting node with the current environmental data to generate an incremental parameter package, which contains the updated control strategy content for each lighting node.

[0013] Optionally, the adaptive fusion coefficients include environmental factor coefficients, control stability coefficients, energy efficiency optimization coefficients, and node load coefficients.

[0014] Optionally, the step of periodically evaluating the lighting control effect through the feedback mechanism of edge computing nodes, dynamically adjusting the improved artificial fish swarm algorithm parameters and inter-layer mapping weights according to the operating status, adaptively optimizing the allocation of computing resources and control strategies, and updating historical data to obtain globally coordinated intelligent lighting and optimal energy efficiency includes: Edge computing nodes periodically collect real-time operating data of lighting equipment, compare it with preset target states, calculate control deviations, and evaluate the effectiveness of the current control strategy. Based on the evaluation results, the edge computing nodes determine whether the control strategy needs to be adjusted. If the deviation exceeds the set threshold, the nodes adjust the relevant parameters in the improved artificial fish swarm algorithm through the feedback mechanism. Edge computing nodes dynamically adjust their local control strategies based on the adjusted artificial fish swarm algorithm and inter-layer mapping weights; Edge computing nodes record and store the control strategies, optimization parameters, and device performance data after each adjustment. They update the control strategy changes and optimization results in historical data and periodically upload historical data to the cloud for global optimization.

[0015] The beneficial effects of this invention are: The edge computing-based collaborative control method for intelligent lighting integrated circuits proposed in this invention effectively solves the prominent problems caused by the centralized architecture of existing intelligent lighting, such as response latency, heavy communication load, insufficient energy efficiency optimization, and poor adaptability. By introducing edge computing technology, this invention moves data processing and decision-making calculations from the cloud to the vicinity of the lighting nodes, thereby reducing communication burden and network latency, and significantly improving response speed and real-time performance. Through an improved artificial fish swarm algorithm, local optimization decisions are made for each lighting node, enabling real-time adjustments based on environmental data, user needs, and energy efficiency requirements. This ensures that parameters such as brightness, on / off status, and color temperature of the lighting equipment always meet the current optimization goals, thereby improving energy efficiency and comfort.

[0016] The method of this invention, through the introduction of a structured state-space model, effectively integrates local state information with global control strategies, avoiding the problems of low efficiency and inaccurate control effects caused by multi-sensor data fusion in traditional smart lighting. The generation of a global optimized control strategy enables coordinated operation among lighting nodes, thereby improving the overall efficiency and stability of the method. Through an adaptive collaborative mapping function, smart lighting can dynamically adjust according to real-time environmental changes, device feedback, and user needs, improving adaptability and flexibility. This invention, while ensuring efficient lighting operation, also provides an efficient and reliable solution for the optimized control of large-scale smart lighting networks, with broad application prospects, especially suitable for smart lighting in smart cities, public facilities, and industrial applications. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a collaborative control method for intelligent lighting integrated circuits based on edge computing proposed in this invention; Figure 2 This is a schematic diagram of an improved artificial fish swarm algorithm for a collaborative control method of intelligent lighting integrated circuits based on edge computing proposed in this invention. Detailed Implementation

[0018] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0019] refer to Figure 1 and Figure 2 A collaborative control method for intelligent lighting integrated circuits based on edge computing, comprising: Edge computing units are deployed at each lighting node to collect environmental data in real time and construct local environmental status information for the node. Based on the collected environmental data, an improved artificial fish swarm algorithm is used to make local optimization decisions for each lighting node, generating a locally optimal control result that adapts to the current environment. Within the edge computing unit, lighting control commands are generated based on local optimal control results, the parameters of the lighting equipment are adjusted, the lighting control commands are executed, and feedback on the actual working status of the local lighting equipment is obtained. A structured state space model is established based on feedback from actual working status. The structured state space model includes an edge layer, a cloud layer, and a data storage and management layer. An adaptive collaborative mapping function is introduced between the edge layer and the cloud layer to facilitate information exchange, generate a global optimization control strategy and form an incremental parameter package, and store historical data in the data storage and management layer. Incremental parameter packages are distributed to each edge node. Each node dynamically balances the old and new parameters based on the adaptive fusion coefficient. When edge-cloud communication is interrupted, the edge node enters autonomous operation mode to maintain short-term state prediction and independent control. By leveraging the feedback mechanism of edge computing nodes, the lighting control effect is evaluated periodically. Based on the operating status, the parameters of the improved artificial fish swarm algorithm and the inter-layer mapping weights are dynamically adjusted to adaptively optimize the allocation of computing resources and control strategies. Historical data is then updated to achieve globally coordinated intelligent lighting and optimal energy efficiency.

[0020] In this embodiment, the edge computing unit includes a sensor module, a data processing unit, and a control execution unit.

[0021] In this embodiment, the environmental data includes light intensity, temperature and humidity, human activity, and energy consumption data.

[0022] In this embodiment, the local environment status information of the construction node includes: An edge computing node is deployed at each lighting node. The edge computing unit connects to multiple sensor modules and collects environmental data in real time through the sensors, including light sensors, temperature and humidity sensors, human activity sensors, and energy consumption sensors. Each sensor collects data periodically or continuously and transmits the collected raw data to the data processing unit of the edge computing unit. The collected environmental data is input into the data processing unit for noise reduction, filtering, and normalization. The processed environmental data is integrated into the local environmental state information of the node, which serves as the real-time state input of the lighting node. By fusing data from multiple sensors, a multi-dimensional dataset that comprehensively reflects the current environmental state of the lighting node is generated and transmitted to the control execution unit to generate lighting control commands adapted to the current environment. The process of fusing data from multiple sensors is achieved by using principal component analysis to effectively integrate multi-dimensional data from different sensors.

[0023] In this embodiment, generating a locally optimal control result adapted to the current environment includes: Initialize the artificial fish swarm, setting the position and speed of each fish. The position represents the control parameter of the lighting node, and the speed represents the change of the control parameter. The initial control parameters include lighting brightness, on / off status, color temperature, and lighting area. The position of each fish represents the combination of control parameters of the lighting node, and the speed represents the change of the control parameter. During initialization, the control parameter position is the initial value of brightness, on / off status, color temperature, and lighting area, and the speed is the initial change of the control parameter. A local search is performed on each artificial fish, specifically: Calculate the distance between the current fish position and the target position, and calculate the speed of the search direction based on the difference between the current local optimum and the global optimum. The search speed is jointly determined by the inertia weight, the difference between the local optimum and the current fish position, and the difference between the global optimum and the current fish position. The search speed equals the speed of the previous moment multiplied by the inertia weight, plus the difference between the local optimum and the current fish position multiplied by the local learning factor, plus the difference between the global optimum and the current fish position multiplied by the global learning factor, where: The local learning factor is usually used to adjust the intensity of the artificial fish's search near the local optimum, ensuring that the optimization process can effectively explore the neighborhood of the current local solution in order to further refine the current solution. When the local learning factor is large, the artificial fish will tend to explore the area near the local optimum more, thereby accelerating the convergence of the local optimum. However, if it relies too much on local search, it will cause the algorithm to get stuck in the local optimum. The global learning factor controls the intensity of the artificial fish's search near the global optimum, guiding the optimization process to avoid the predicament of local optima. By exploring the global optimum through global search, when the global learning factor is large, the artificial fish will tend to move closer to the global optimum, which helps to break out of the range of local optima and conduct a global search. However, an excessively high global learning factor will cause the algorithm to search too broadly in the global scope, resulting in an insufficiently refined search process. The inertia weight controls the influence of the previous velocity, while the local and global learning factors control the influence of the local and global optimal positions on the search direction; A fuzzy dynamic decision factor is introduced to dynamically adjust the weighting coefficients in speed calculation. This fuzzy dynamic decision factor is adjusted in real time based on environmental data, user needs, and energy efficiency requirements. The fuzzy dynamic decision factor adjusts the weighting coefficients in speed calculation, optimizing the position update process for each fish and limiting the position and speed update range for each fish. Specifically, the introduction of the fuzzy dynamic decision factor to dynamically adjust the weighting coefficients in speed calculation involves: Based on real-time environmental data, user needs, and energy efficiency requirements, the data is processed according to preset fuzzy rules to obtain a fuzzy decision factor value between 0 and 1. The local and global learning factors are dynamically adjusted based on the fuzzy decision factor value. When the fuzzy decision factor is large, it indicates a greater tendency to perform local search. Therefore, the local learning factor is increased to enhance the fine adjustment capability of the local optimum. When the fuzzy decision factor is small, it indicates that global search is more important. The global learning factor is increased to guide the search toward the global optimum and avoid the dilemma of local optima. Through dynamic adjustment, the search process is optimized, and the capabilities of local and global search can be flexibly adjusted in different environments. After adjusting the local and global learning factors, the position is updated by substituting them into the search speed calculation formula, and the speed of each artificial fish is calculated. The position and speed of the fish swarm are updated through local and global searches. The fitness of each fish is calculated based on a fitness function that includes lighting effect and energy efficiency assessment. The weighting coefficients are adjusted according to requirements. The fitness function comprehensively considers the current lighting brightness and energy consumption data, adjusting the weight of lighting brightness to optimize the lighting effect, while controlling energy efficiency to reduce energy consumption. Specifically, the fitness function includes lighting effect and energy efficiency assessment. Lighting effect evaluation mainly includes the quantification of brightness, color temperature, and illumination uniformity. Brightness evaluation is calculated based on the difference between the current lighting intensity and the expected lighting standard. If the lighting brightness is too low, it will affect user comfort, and the fitness function will impose a high penalty. Color temperature evaluation considers the gap between the current color temperature and user needs. If the color temperature is too cold or too warm, it will reduce comfort, and the fitness function will optimize the color temperature deviation. Illumination uniformity evaluation also affects the fitness, especially in large-area lighting control. The fitness function will be adjusted according to the distribution of light to ensure that there are no excessively dark areas. The energy efficiency assessment is mainly conducted using the collected energy consumption data. The fitness function will calculate the energy consumption of the current lighting node and compare it with the target energy efficiency. If the energy consumption of the lighting is too high, the fitness value will decrease, prompting overall optimization. The fitness function will also consider the energy efficiency ratio, which is the ratio of lighting effect to energy consumption. The optimization goal is to improve the energy efficiency ratio and reduce unnecessary energy consumption while ensuring the lighting effect. The fitness function is finally calculated by weighting the lighting effect assessment and energy efficiency assessment. The weighting coefficients are dynamically adjusted according to the needs of actual application to ensure that the algorithm prioritizes different objectives under different operating conditions. During periods of high demand, the weighting coefficient of lighting effect will increase to ensure comfortable lighting conditions, while during periods of low demand, the weighting coefficient of energy efficiency will increase to reduce unnecessary energy consumption. Based on the fitness calculation results, the optimal control parameters are selected, the global optimal position and the local optimal position are updated, and the iterative search continues until the stopping conditions are met. The stopping conditions include the fitness reaching a preset threshold or reaching the maximum number of iterations. When the stopping conditions are met, the optimal control parameters are output as the final control strategy for lighting control.

[0024] In this embodiment, establishing a structured state-space model based on actual working state feedback includes: The edge layer takes real-time collected environmental data as input and uses local state space equations to estimate the state of the current lighting node in real time, generating local state information of the node, specifically: Based on the real-time collected environmental data, the edge layer uses the local state space equation to estimate the state of the lighting nodes. The local state space equation estimates the current state of the nodes in real time by taking into account the input environmental data and control input, combined with the measurement error. Through the local state space equation, the lighting nodes can be dynamically predicted, the current working state can be identified, and adjustments can be made. By calculating the local state space equation, the edge layer generates local state information, including the real-time state of the lighting nodes, and dynamically adjusts the parameters according to environmental changes and control requirements, which serve as the local state information of the nodes for subsequent decision-making and control. The local state information generated by each edge node is uploaded to the cloud layer, where a global state space equation is constructed. By aggregating the local state information from each edge node, a global optimization control strategy is formed. The global state space equation includes global state variables and global control inputs, wherein: Global state variables are a set of variables that uniformly describe the state of all lighting nodes, including the weighted average of lighting brightness, color temperature, on / off status, and energy consumption of all nodes. Global state variables reflect the overall operating status of the entire smart lighting system. By aggregating the local state information of each lighting node, it is possible to comprehensively understand the lighting needs and energy efficiency status of different areas or spaces. Global state variables are obtained by aggregating the local state information of all nodes and using a weighted average fusion method to calculate the global state based on the state importance and control priority of each node. Global control input refers to the control parameters applied to all lighting nodes to achieve the optimal control objective. Global control input can optimize the overall lighting effect and energy efficiency by uniformly allocating the control parameters of all lighting nodes. Global control input includes the transmission of control commands, the adjustment range of lighting brightness and color temperature, and energy efficiency optimization strategies. The data storage and management layer stores historical data, control policies, status information, and device logs for each edge node; Based on the calculation results of the local state space equation and the global state space equation, the control strategy of each node is dynamically adjusted, and the nodes are optimized and controlled in real time. The control parameters of each lighting node are adjusted based on the real-time estimation results of the edge layer and the global optimization strategy of the cloud layer. Specifically, the dynamic adjustment of the control strategy of each node is as follows: Each lighting node adjusts its control parameters to ensure optimal performance based on the local state estimated in real time by the edge computing unit and the global optimization strategy provided by the cloud. If the light intensity in a certain area is insufficient, the edge computing unit will adjust the brightness of the area to improve user comfort. If the cloud optimization strategy indicates that a certain area needs to reduce energy consumption, the edge computing unit will reduce the brightness of the area or automatically adjust the color temperature. Each lighting node will also dynamically adjust its control strategy according to real-time environmental changes. In densely populated areas, smart lighting can increase brightness and color temperature to ensure sufficient illumination, while in uninhabited areas, it will automatically reduce brightness to reduce energy consumption.

[0025] In this embodiment, the step of introducing an adaptive collaborative mapping function between the edge layer and the cloud layer for information interaction, generating a global optimization control strategy, and forming an incremental parameter package includes: The adaptive cooperative mapping function receives local state information from the edge layer and dynamically adjusts it based on real-time input data to generate a simplified state summary. The adaptive cooperative mapping function consists of three main processing parts: an input processing part, a dynamic mapping part, and an output generation part, specifically: The input processing section receives local state information from the edge layer and adjusts the data preprocessing strategy according to the current load, real-time requirements, and environmental changes. The dynamic mapping part maps the input data into a simplified state summary using a weighted averaging method based on a real-time adjusted processing strategy. The output generation section converts the simplified state summary into a format suitable for uploading and sends it to the cloud layer; The cloud layer uses a collaborative attention mechanism to automatically adjust the weights of the state summaries of each node. The collaborative attention mechanism adaptively adjusts the weights based on the current environmental state, real-time feedback and needs of each node, generating a global optimization control strategy. The cloud layer calculates and updates the control parameters of each lighting node based on the global optimization control strategy. It then combines the updated control parameter values ​​of each lighting node with the current environmental data to generate an incremental parameter package, which contains the updated control strategy content for each lighting node.

[0026] In this embodiment, the adaptive fusion coefficient includes environmental factor coefficient, control stability coefficient, energy efficiency optimization coefficient, and node load coefficient.

[0027] In this embodiment, the step of periodically evaluating the lighting control effect through the feedback mechanism of edge computing nodes, dynamically adjusting the improved artificial fish swarm algorithm parameters and inter-layer mapping weights according to the operating status, adaptively optimizing the allocation of computing resources and control strategies, and updating historical data to obtain globally coordinated intelligent lighting and optimal energy efficiency includes: Edge computing nodes periodically collect real-time operating data of lighting equipment, compare it with preset target states, calculate control deviations, and evaluate the effectiveness of the current control strategy. Based on the evaluation results, the edge computing nodes determine whether the control strategy needs to be adjusted. If the deviation exceeds the set threshold, the nodes adjust the relevant parameters in the improved artificial fish swarm algorithm through the feedback mechanism. Edge computing nodes dynamically adjust their local control strategies based on the adjusted artificial fish swarm algorithm and inter-layer mapping weights; Edge computing nodes record and store the control strategies, optimization parameters, and device performance data after each adjustment. They update the control strategy changes and optimization results in historical data and periodically upload historical data to the cloud for global optimization.

[0028] Example 1: To verify the feasibility of this invention in practice, it was applied to the intelligent lighting system of a large commercial office building. The building has 2000 lighting nodes, each equipped with sensors for light intensity, temperature and humidity, occupant activity, and energy consumption for real-time data collection and transmission. Previously, the building used a centralized intelligent lighting control system, with a central control unit managing and adjusting all lighting nodes. However, during peak hours, when there are many people in the building, the response time of the centralized system increases significantly, especially under heavy data transmission and computational burdens, often resulting in response delays and even lighting control failures at some nodes. Because traditional systems rely on static rules and cannot adjust control parameters in real time according to environmental and user needs, the lighting system is not only inefficient but also frequently wastes energy, and user comfort is not optimally guaranteed.

[0029] In May 2025, this invention was first implemented in an intelligent lighting system within an office building. Each lighting node's edge computing unit collects and processes environmental data in real time, including light intensity, temperature, humidity, and human activity, using an improved artificial fish swarm algorithm for local optimization decisions. The edge computing unit can adaptively adjust based on the actual feedback from each node, optimizing lighting effects and avoiding over-illumination, thus reducing unnecessary energy consumption. In this way, each lighting node can make decisions locally, reducing the computational burden on the central control unit and improving the system's real-time responsiveness. Leveraging a collaborative attention mechanism, the cloud can weightedly aggregate the state summaries of all lighting nodes to generate a global optimization control strategy, ensuring coordinated operation of the entire building's lighting system. The system adaptively adjusts the control parameters of each node based on real-time environmental data changes, thereby optimizing energy efficiency and improving user comfort.

[0030] Table 1 Performance Comparison of Intelligent Lighting Systems Data metrics Before implementation (traditional control system) After implementation (edge ​​computing-based control system) System response time (seconds) 3.2 0.8 Lighting energy consumption (kWh / month) 2500 1800 User satisfaction (%) 72 94 Lighting brightness deviation (Lux) 15 3 Equipment failure rate (%) 5.2 1.5 Number of times the lighting automatically adjusts (per day) 50 250 Light environment adaptability (%) 65 95 System data transmission latency (milliseconds) 800 150 Number of times lighting brightness is adjusted (per month) 30 180 As can be seen from the comparative data in Table 1, the present invention significantly improves system response time compared to traditional control systems. Traditional centralized control systems, relying on a central control unit for data processing, result in a long system response time, averaging 3.2 seconds. In contrast, the edge computing-based control system, by moving data processing and decision-making calculations to the vicinity of each lighting node, greatly reduces communication latency and computational burden, shortening the response time to 0.8 seconds and improving the system's real-time performance and flexibility.

[0031] This invention also demonstrates significant advantages in energy efficiency and user satisfaction. Traditional systems, lacking real-time adjustment mechanisms, consume a lot of energy, often resulting in over-illumination and energy waste. In contrast, this invention dynamically adjusts parameters such as brightness and color temperature of lighting equipment based on real-time environmental data and user needs, reducing unnecessary energy consumption by 24%. User satisfaction increased from 72% to 94%, primarily due to the system's ability to provide a more precise and personalized lighting experience, meeting users' higher demands for comfort.

[0032] Significant improvements have been made in lighting brightness deviation, equipment failure rate, and the number of lighting adjustments required. Traditional systems exhibit large lighting brightness deviations and cannot quickly respond to environmental changes, resulting in unstable lighting effects with an average deviation of 15 Lux. This invention, however, can adjust brightness in real time, reducing the deviation to 3 Lux, thus improving lighting comfort and stability. The equipment failure rate has decreased from 5.2% to 1.5%, reflecting improved system stability and reduced system failures and maintenance costs. Furthermore, this invention enhances lighting accuracy and flexibility through more frequent automatic adjustments (180 times per month), enabling it to better adapt to different lighting environments in practical applications and improving the system's adaptability and efficiency.

[0033] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A collaborative control method for intelligent lighting integrated circuits based on edge computing, characterized in that, include: Edge computing units are deployed at each lighting node to collect environmental data in real time and construct local environmental status information for the node. Based on the collected environmental data, an improved artificial fish swarm algorithm is used to make local optimization decisions for each lighting node, generating a locally optimal control result that adapts to the current environment. Within the edge computing unit, lighting control commands are generated based on local optimal control results, the parameters of the lighting equipment are adjusted, the lighting control commands are executed, and feedback on the actual working status of the local lighting equipment is obtained. A structured state space model is established based on feedback from actual working status. The structured state space model includes an edge layer, a cloud layer, and a data storage and management layer. An adaptive collaborative mapping function is introduced between the edge layer and the cloud layer to facilitate information exchange, generate a global optimization control strategy and form an incremental parameter package, and store historical data in the data storage and management layer. Incremental parameter packages are distributed to each edge node. Each node dynamically balances the old and new parameters based on the adaptive fusion coefficient. When edge-cloud communication is interrupted, the edge node enters autonomous operation mode to maintain short-term state prediction and independent control. By leveraging the feedback mechanism of edge computing nodes, the lighting control effect is evaluated periodically. Based on the operating status, the parameters of the improved artificial fish swarm algorithm and the inter-layer mapping weights are dynamically adjusted to adaptively optimize the allocation of computing resources and control strategies. Historical data is then updated to achieve globally coordinated intelligent lighting and optimal energy efficiency.

2. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The edge computing unit includes a sensor module, a data processing unit, and a control execution unit.

3. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The environmental data includes light intensity, temperature and humidity, human activity, and energy consumption data.

4. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The local environment status information of the construction node includes: An edge computing node is deployed at each lighting node. The edge computing unit connects to multiple sensor modules and collects environmental data in real time through the sensors, including light sensors, temperature and humidity sensors, human activity sensors, and energy consumption sensors. Each sensor collects data periodically or continuously and transmits the collected raw data to the data processing unit of the edge computing unit. The collected environmental data is input into the data processing unit for noise reduction, filtering, and normalization. The processed environmental data is integrated into the local environmental status information of the node, which serves as the real-time status input of the lighting node. By fusing data from multiple sensors, a multi-dimensional dataset that comprehensively reflects the current environmental status of the lighting node is generated and transmitted to the control execution unit to generate lighting control commands adapted to the current environment.

5. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The generation of locally optimal control results adapted to the current environment includes: Initialize the artificial fish swarm, setting the position and speed of each fish. The position represents the control parameter of the lighting node, and the speed represents the change of the control parameter. The initial control parameters include lighting brightness, on / off status, color temperature, and lighting area. The position of each fish represents the combination of control parameters of the lighting node, and the speed represents the change of the control parameter. During initialization, the control parameter position is the initial value of brightness, on / off status, color temperature, and lighting area, and the speed is the initial change of the control parameter. A local search is performed on each artificial fish, specifically: Calculate the distance between the current fish position and the target position, and calculate the speed of the search direction based on the difference between the current local optimum and the global optimum. The search speed is determined by the inertia weight, the difference between the local optimal position and the current fish position, and the difference between the global optimal position and the current fish position. The search speed is equal to the speed of the previous moment multiplied by the inertia weight, plus the difference between the local optimal position and the current fish position multiplied by the local learning factor, plus the difference between the global optimal position and the current fish position multiplied by the global learning factor. The inertia weight controls the influence of the previous velocity, while the local and global learning factors control the influence of the local and global optimal positions on the search direction; A fuzzy dynamic decision factor is introduced to dynamically adjust the weighting coefficients in the speed calculation. The fuzzy dynamic decision factor is adjusted in real time based on environmental data, user needs and energy efficiency requirements. The fuzzy dynamic decision factor adjusts the weighting coefficients in the speed calculation, optimizes the position update process of each fish, and limits the position and speed update range of each fish. The position and speed of the fish swarm are updated through local and global searches. The fitness of each fish is calculated based on the fitness function, which includes lighting effect and energy efficiency assessment. The weight coefficients are adjusted according to the needs. The fitness function comprehensively considers the current lighting brightness and energy consumption data, adjusts the weight of lighting brightness to optimize the lighting effect, and controls energy efficiency to reduce energy consumption. Based on the fitness calculation results, the optimal control parameters are selected, the global optimal position and the local optimal position are updated, and the iterative search continues until the stopping conditions are met. The stopping conditions include the fitness reaching a preset threshold or reaching the maximum number of iterations. When the stopping conditions are met, the optimal control parameters are output as the final control strategy for lighting control.

6. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The establishment of a structured state-space model based on actual working status feedback includes: The edge layer takes real-time collected environmental data as input and uses local state space equations to estimate the state of the current lighting node in real time, generating local state information of the node. The local state information generated by each edge node is uploaded to the cloud layer, and a global state space equation is constructed in the cloud layer. By aggregating the local state information from each edge node, a global optimization control strategy is formed. The data storage and management layer stores historical data, control policies, status information, and device logs for each edge node; Based on the calculation results of the local state space equation and the global state space equation, the control strategy of each node is dynamically adjusted, and the nodes are optimized and controlled in real time. Based on the real-time estimation results of the edge layer and the global optimization strategy of the cloud layer, the control parameters of each lighting node are adjusted.

7. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The process of introducing an adaptive collaborative mapping function between the edge layer and the cloud layer for information interaction, generating a global optimization control strategy, and forming an incremental parameter package includes: The adaptive cooperative mapping function receives local state information from the edge layer and dynamically adjusts it based on real-time input data to generate a simplified state summary. The adaptive cooperative mapping function consists of three main processing parts: an input processing part, a dynamic mapping part, and an output generation part, specifically: The input processing section receives local state information from the edge layer and adjusts the data preprocessing strategy according to the current load, real-time requirements, and environmental changes. The dynamic mapping part maps the input data into a simplified state summary using a weighted averaging method based on a real-time adjusted processing strategy. The output generation section converts the simplified state summary into a format suitable for uploading and sends it to the cloud layer; The cloud layer uses a collaborative attention mechanism to automatically adjust the weights of the state summaries of each node. The collaborative attention mechanism adaptively adjusts the weights based on the current environmental state, real-time feedback and needs of each node, generating a global optimization control strategy. The cloud layer calculates and updates the control parameters of each lighting node based on the global optimization control strategy. It then combines the updated control parameter values ​​of each lighting node with the current environmental data to generate an incremental parameter package, which contains the updated control strategy content for each lighting node.

8. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The adaptive fusion coefficients include environmental factor coefficients, control stability coefficients, energy efficiency optimization coefficients, and node load coefficients.

9. The collaborative control method for intelligent lighting integrated circuits based on edge computing according to claim 1, characterized in that, The aforementioned feedback mechanism through edge computing nodes periodically evaluates the lighting control effect, dynamically adjusts the improved artificial fish swarm algorithm parameters and inter-layer mapping weights based on the operating status, adaptively optimizes computing resource allocation and control strategies, and updates historical data to obtain globally coordinated intelligent lighting and optimal energy efficiency, including: Edge computing nodes periodically collect real-time operating data of lighting equipment, compare it with preset target states, calculate control deviations, and evaluate the effectiveness of the current control strategy. Based on the evaluation results, the edge computing nodes determine whether the control strategy needs to be adjusted. If the deviation exceeds the set threshold, the nodes adjust the relevant parameters in the improved artificial fish swarm algorithm through the feedback mechanism. Edge computing nodes dynamically adjust their local control strategies based on the adjusted artificial fish swarm algorithm and inter-layer mapping weights; Edge computing nodes record and store the control strategies, optimization parameters, and device performance data after each adjustment. They update the control strategy changes and optimization results in historical data and periodically upload historical data to the cloud for global optimization.

Images