An ambient light detection type light intensity adaptive power control method

Abstract

The application discloses an ambient light detection type light intensity adaptive power supply control method and relates to the technical field of intelligent lighting control. The locally cooperative game model is constructed to change each lamp node from an independent execution unit into an intelligent agent with local decision-making capability. Each lamp node acquires the state information of its optical neighbor node through periodic data sharing and constructs a locally cooperative game model with its brightness as a decision variable and the maximization of a local utility function as a target. When all nodes execute the local optimization in parallel and asynchronously, the decisions of the nodes influence and restrict each other through optical channels and communication channels. Through iteration, the method can dynamically converge to a Nash equilibrium point without the intervention of a central dispatcher. In the equilibrium state, the difference between the illumination of each node and the expected value is minimized, illumination uniformity and energy efficiency optimization are achieved, and efficient mapping from local intelligence to global optimization is realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent lighting control technology, specifically to an ambient light detection-based adaptive power supply control method for light intensity. Background Technology

[0002] Ambient light detection adaptive dimming technology is one of the core research directions in the field of intelligent lighting and energy-saving control. Its basic concept is to monitor the ambient light intensity in real time through light sensors and automatically adjust the output of artificial light sources accordingly, so as to achieve energy saving while meeting the illuminance requirements of specific spaces.

[0003] However, existing technologies have significant shortcomings in achieving self-adaptation and intelligence. When the system scales up and multiple lamps with ambient light sensing capabilities coexist, the single-lamp single-control mode is used. That is, each lamp makes independent decisions based on its own sensor data. When a lamp brightens itself due to the dimming of ambient light, its emitted light becomes part of the ambient light read by the sensors of its neighboring lamps. This causes the neighboring lamps to misjudge that the ambient light is sufficient and dim themselves, which in turn triggers the first lamp to brighten again, forming a negative feedback loop. This results in unstable lighting throughout the space. Furthermore, due to the lack of coordination, it is impossible to carry out differentiated and refined lighting distribution based on the actual degree of influence of natural light on each point in the space, which wastes energy and makes it difficult to ensure uniform illuminance.

[0004] In summary, existing ambient light detection adaptive control methods lack a distributed collaborative decision-making mechanism, making it difficult to balance the relationship between illumination uniformity, energy consumption control, and dynamic adaptability. Therefore, there is an urgent need for a control method that enables information sharing and collaborative game among luminaires. By linking local decision-making with global optimization, this method can overcome the inherent defects of traditional independent control modes and meet the comprehensive needs of intelligent lighting systems for high efficiency, comfort, and energy saving. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide an ambient light detection-based adaptive power supply control method for lighting intensity. This method transforms each lighting node from an independent execution unit into an intelligent agent with local decision-making capabilities through a constructed cooperative game model. Each lighting node acquires the state information of its optical neighbors through periodic data sharing and constructs a local cooperative game model with its own brightness as the decision variable and maximizing the local utility function as the objective. When all nodes execute this local optimization in parallel and asynchronously, their decisions influence and constrain each other through optical and communication channels. Through iteration, without the intervention of a central scheduler, it can dynamically converge to a Nash equilibrium point. In this equilibrium state, the difference between the illuminance and the expected value at each point is minimized, achieving uniform illumination and energy efficiency optimization, and demonstrating an efficient mapping from local intelligence to global optimization.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an ambient light detection-based adaptive power supply control method for light intensity, the specific steps of which are as follows:

[0007] S100. Configure independent intelligent agent attributes for each lamp, establish a wireless communication network between lamps, enable all lamp intelligent agents to conduct broadcast communication, and construct a list of neighbor nodes containing all other lamps within their communication range.

[0008] S200. Each lighting intelligent agent detects the local ambient light intensity at its location in real time through its integrated ambient light sensor. Through the wireless communication network, each node generates a shared data packet containing its own identity, real-time ambient light intensity value, and its current brightness, and broadcasts the data packet to all nodes in its neighbor node list.

[0009] S300. Each lighting agent constructs a local cooperative game model based on its own shared data packets and the shared data received from neighboring nodes. The model uses the current brightness of the lighting agent as the decision variable and aims to maximize its local utility function to solve the problem, thereby calculating a candidate target brightness that reaches a local Nash equilibrium.

[0010] S400: Each lamp agent compares the candidate target brightness calculated by S300 with the initial brightness value, performs a brightness decision based on the comparison result, and adjusts its own brightness in real time.

[0011] S500, repeats S200-S400 according to a preset cycle, dynamically updates ambient light intensity and neighborhood information, optimizes brightness adjustment parameters in real time, and maintains dynamic balance of lighting throughout the space.

[0012] Furthermore, in S100, the agent attributes include assigning a unique identifier to each lamp, setting an initial brightness value, and the position information of each lamp in space.

[0013] Furthermore, in S100, the specific process of constructing the neighbor node list includes:

[0014] During the initialization phase, each lighting agent sends a probe beacon containing its own identity and location information, and listens for probe beacons from other lighting agents.

[0015] When a lighting agent receives a detection beacon from another lighting agent, it calculates the relative distance to the signal source.

[0016] When the relative distance is less than the preset communication cooperation radius, the identity of the signal source is recorded in its own list of neighboring nodes, thereby forming a network topology focused on local area cooperation.

[0017] This list of neighboring nodes serves as the scope for data broadcasting in S200 and the set of participants for constructing a local cooperative game model in S300.

[0018] Furthermore, in S300, the construction process of the local cooperative game model is as follows:

[0019] Game element definition: Each lighting agent, along with its own list of neighboring nodes, constitutes the set of participating lighting nodes in the current local collaborative game. Each lighting node... The strategy variable is its own current brightness. The set of all strategy variables constitutes the strategy space of this local cooperative game model;

[0020] Local environmental state perception: lighting node Based on the shared data obtained by S200, the total perceived illuminance at its location is calculated. ;

[0021] Construction of multi-objective utility function: for lamp node Construct a quantified local utility function The function will adjust its brightness. It is mapped to a scalar value that represents its local benefit;

[0022] Nash equilibrium solution: In each control cycle, the lamp node Get the current brightness of other neighboring nodes Treating it as a fixed policy, the feasible range of the policy space is determined by numerical optimization methods. , The search is performed within the [function name] to find a local utility function that satisfies its own local utility function. Strategies to maximize This strategy That is, in the current round of the game, the lamp node. Candidate target brightness ,in, This represents the minimum luminous intensity of the light fixture. This represents the maximum luminous intensity of the lamp.

[0023] Decision information update and output: Calculate the brightness values ​​of candidate targets. The internal state register of the lighting agent is updated and prepared for brightness adjustment in S400. At the same time, this brightness value will be shared with neighboring nodes as its current brightness in the next cycle S200.

[0024] Furthermore, the total perceived illuminance ,in, It is a lighting node The ambient light illuminance value is directly detected by its own ambient light sensor. It is a lighting node The set of neighboring nodes, Neighboring nodes The current brightness, Neighboring nodes Lighting fixture nodes The illuminance contribution weighting coefficient at the location.

[0025] Furthermore, the local utility function ,in, It is a lighting node The utility value is the highest value, indicating better overall performance. It is at the node Total illuminance perceived at the location It is the preset global illuminance value. It is a node Current brightness The power consumption function and , These are weighting coefficients, which respectively adjust the relative importance of illuminance uniformity and energy consumption in the optimization objective.

[0026] Furthermore, the specific steps of the numerical optimization method are as follows:

[0027] Calculate the local utility function Regarding one's own decision variables The first-order partial derivative of is expressed as: ,in, Neighboring nodes Lighting fixture nodes The illuminance contribution weighting coefficient at the location, These are weighting coefficients. Power consumption function The derivative, It is at the node Total illuminance perceived at the location It is the preset global illuminance value;

[0028] With nodes The current brightness value is used as the initial point for iteration. In each iteration, along the gradient... In the positive direction, according to the preset step size factor Update brightness value: ;

[0029] Repeat the iterative process until any preset convergence condition is met, at which point the brightness value is obtained. This refers to the nodes in the current game situation. The candidate target brightness within this decision-making cycle, wherein the convergence condition is:

[0030] The magnitude of the first partial derivative is less than the threshold. ,Right now ;

[0031] The change in brightness between adjacent iterations is less than the threshold. ,Right now .

[0032] Furthermore, in S400, the brightness decision-making process is as follows:

[0033] Lighting Node Read the calculated candidate target brightness And obtain the initial brightness value from its agent properties. Calculate the absolute difference between the two. ,Right now ;

[0034] The calculated brightness difference Compared with the preset stability threshold Compare and make a decision based on the comparison results:

[0035] when If the current lighting environment changes, the brightness of the candidate target will be adjusted accordingly. Adopted as the target brightness for this round;

[0036] when If the brightness is not adjusted, the system will maintain its current state.

[0037] Compared with existing technologies, this ambient light detection-based adaptive power supply control method for light intensity has the following advantages:

[0038] I. This invention transforms each lighting node from an independent execution unit into an intelligent agent with local decision-making capabilities through a constructed collaborative game model. Each lighting node acquires the state information of its optical neighbor nodes through periodic data sharing, and constructs a local collaborative game model with its own brightness as the decision variable and maximizing the local utility function as the objective. When all nodes execute this local optimization in parallel and asynchronously, their decisions will influence and constrain each other through the optical and communication channels. After iteration, without the intervention of a central scheduler, it can dynamically converge to a Nash equilibrium point. In this equilibrium state, the difference between the illuminance and the expected value of each point is minimized, realizing uniform illumination and energy efficiency optimization, and demonstrating an efficient mapping from local intelligence to global optimization.

[0039] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0041] Figure 1 A flowchart of an ambient light detection-based adaptive power supply control method for lighting intensity.

[0042] Figure 2 This is a flowchart illustrating the steps of an ambient light detection-based adaptive power supply control method for lighting intensity. Detailed Implementation

[0043] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0044] Example 1

[0045] This embodiment provides an ambient light detection-based adaptive power supply control method for light intensity, aiming to solve problems such as unstable illumination, energy waste, and poor illuminance uniformity that easily occur in existing intelligent lighting systems with independent control of multiple lamps. Figure 2As shown, in S100, each lamp is configured with independent intelligent agent attributes, a wireless communication network is established, broadcast communication is performed, and a list of neighboring nodes containing all other lamps within its communication range is constructed. In S200, the local ambient light intensity is detected through an ambient light sensor, and each node generates a shared data packet, which is broadcast to all nodes in its neighboring node list. In S300, the lamp intelligent agent constructs a local collaborative game model based on its own shared data packet and the shared data of neighboring nodes, and calculates the candidate target brightness that reaches the local Nash equilibrium. In S400, each lamp intelligent agent compares the calculated candidate target brightness with the initial brightness value, performs brightness decision based on the comparison result, and adjusts its own emission brightness in real time. In S500, S200-S400 are repeated according to a preset cycle to dynamically update the ambient light intensity and neighborhood information, optimize the brightness adjustment parameters in real time, and maintain the dynamic balance of the entire space illumination. This process does not require a central scheduler, and can achieve energy-saving goals while meeting the space illumination requirements through distributed collaborative decision-making. It is suitable for large-scale lighting scenarios where multiple lamps coexist, such as shopping malls, office buildings, and underground parking garages.

[0046] First, the system enters the initialization phase (S100). In this phase, each lamp is given intelligent agent characteristics and a local communication topology is established. The intelligent agent characteristics include: assigning each lamp a unique identifier, which allows each lamp to accurately distinguish itself from other nodes in subsequent data sharing and neighbor identification, avoiding data confusion; and setting the initial brightness value. This value ensures basic lighting at startup; it also records the coordinates of each lamp in the actual space. This information is the core basis for subsequent calculations of the relative distance between lamps and determination of the range of neighboring nodes. These attributes together constitute the basic state of the agent. The wireless communication network adopts a low-power protocol, enabling all lamps to broadcast communication. Specifically, each lamp sends a probe beacon during initialization, containing its own ID and location coordinates, and simultaneously listens for other beacons. When a beacon is received, the Euclidean distance between the lamp and the signal source is calculated. If the distance is less than the preset communication cooperation radius, the signal source ID is recorded in the neighboring node list. This process forms a locally cooperative network topology, in which each lamp only interacts with its neighboring nodes, reducing communication overhead. The neighboring node list serves as the set of participants in subsequent data broadcasting and game model, ensuring the locality and scalability of control. This stage defines the cooperation range through spatial proximity, avoiding the complexity of global communication, and laying the foundation for distributed optimization.

[0047] Then, the data sharing phase (S200) begins. This phase enables real-time sensing and dissemination of environmental and lighting status. Each lighting fixture uses an integrated high-precision ambient light sensor to detect the local ambient light intensity at its installation location, denoted as... The sensor measures at a fixed sampling rate to ensure the timeliness of the data. Based on the measurement results, each luminaire generates a shared data packet containing three elements: its own ID, real-time ambient light intensity value, etc. and the current brightness Data packets are broadcast wirelessly to all nodes in the neighbor node list, while simultaneously receiving similar data packets from neighbors. This broadcasting mechanism allows each light fixture to obtain complete status information of its local area, including the brightness of its neighbors. The sharing of ambient light data and the periodicity of the shared data are synchronized with the control cycle to ensure that decisions are based on the latest status. The operation of this stage relies on a reliable wireless transmission protocol, including data verification and retransmission mechanisms to cope with network packet loss. By sharing information, the information silos of traditional independent control are broken down, enabling each luminaire to sense the behavior of its neighbors and changes in the environment, providing a data foundation for collaborative game.

[0048] Next, the local collaborative game phase (S300) begins. In this phase, a game theory model transforms the local state into an optimal decision. Each light fixture, based on its own and its neighbors' shared data, constructs a local collaborative game model with itself as the focus. The participants in the game include itself and all its neighboring nodes. Each participating light fixture node... The strategy variable is its own brightness. The strategy space is the feasible brightness range. , Lighting fixture nodes Calculate total perceived illuminance Its formula is: ,in, It is a lighting node The ambient light illuminance value is directly detected by its own ambient light sensor. It is a lighting node The set of neighboring nodes, Neighboring nodes The current brightness, Neighboring nodes Lighting fixture nodes The illuminance contribution weighting coefficient of the location, total perceived illuminance Quantified nodes Construct a local utility function based on the actual illumination level at the location. Its expression is: ,in, It is a utility value; the larger the value, the better the overall performance. It is the preset global desired illuminance, set by the user's needs. and It is a weighting coefficient that is greater than zero. The importance of adjusting illuminance uniformity, i.e. catch priority, The importance of regulating energy consumption control It is the power consumption function, defined as This study simulates the characteristic of power consumption increasing with the square of luminance in real-world lighting fixtures. The utility function balances illuminance tracking and energy minimization. The first term penalizes illuminance deviation, driving the system towards the desired illuminance; the second term penalizes high energy consumption, promoting energy saving. To solve for Nash equilibrium—where each node, given its neighbor policy, cannot unilaterally change its luminance to improve utility—gradient ascent is used for numerical optimization. Brightness of neighbor Treat it as fixed, calculate about First-order partial derivatives: ,in, Neighboring nodes Lighting fixture nodes The illuminance contribution weighting coefficient at the location, These are weighting coefficients. Power consumption function The derivative, It is at the node Total illuminance perceived at the location It is a preset global illuminance value; the optimization process is based on the current brightness. Using the initial point as the step size factor, Update along the gradient direction: Repeat the iteration until the convergence condition is met: gradient magnitude Or changes in brightness The obtained brightness The candidate target brightness is updated in the internal status register. This provides data support for the brightness decision in S400, and also allows for further updates in the next control cycle, S200. Its current brightness will be included in the shared data packet and broadcast to neighboring nodes, providing the latest input for the neighboring nodes' game calculations. This process realizes the dynamic interaction of strategies between lamps and ensures the coordination of the entire network.

[0049] Secondly, the brightness decision-making stage (S400) is entered. This stage ensures the stable execution of the decision, avoiding over-adjustment and ensuring the stability of the luminaire nodes. Read candidate target brightness and initial brightness Calculate the absolute difference ΔL quantifies the magnitude of brightness adjustment. Compared with the preset stability threshold Comparison: If This indicates significant environmental changes and adoption. To achieve the target brightness, the luminous intensity is adjusted via a dimming circuit; if This indicates minimal change, maintaining the original brightness, avoiding oscillations, and a threshold value. Based on the system tolerance setting, the principle is to introduce a hysteresis mechanism to filter noise and small fluctuations, thereby improving the system robustness. The operation process in this stage combines the dead zone strategy in control theory to ensure that decisions are triggered only when there are significant changes, thereby reducing hardware wear and energy waste.

[0050] Finally, the system enters the cyclic execution phase (S500), which executes S200 to S400 in a fixed control cycle to form a closed-loop control. In each cycle, the luminaire updates the environmental state through S200, calculates new candidate brightness through S300, and makes adjustments through S400. This iteration enables the network to dynamically respond to changes in ambient light and network topology. Distributed parallel computing allows the decisions of each node to be coupled with each other through optical and communication channels. After multiple iterations, the system stabilizes at the global Nash equilibrium point. At this point, no node can improve its utility by making unilateral changes, thus achieving uniform illuminance and minimized energy consumption.

[0051] In summary, this embodiment achieves autonomous and collaborative control of the lighting network through a distributed intelligent agent architecture and a local collaborative game model. This method transforms traditional independent control into swarm intelligence, overcomes the problems of negative feedback oscillation and uneven illuminance, and enables the system to converge to the global Nash equilibrium point without a central scheduler. It effectively solves the shortcomings of existing technologies and provides an efficient, energy-saving, and stable control scheme for intelligent lighting systems, which is suitable for large-scale applications such as intelligent buildings and street lighting.

[0052] Example 2

[0053] Based on Embodiment 1, this embodiment provides a method for adaptive power supply control of ambient light intensity using ambient light detection, including the specific steps involved in such control. Figure 1 As shown, the specific steps are as follows:

[0054] (1) Initialize the lighting fixture properties and communication network

[0055] Assign a unique identifier to each light fixture as a distinguishing feature.

[0056] Set an initial brightness value for each luminaire to determine the baseline brightness when the luminaire is started.

[0057] Record the specific location information of each light fixture in the space to provide a basis for subsequent neighbor node judgment.

[0058] Establish a wireless communication network between the lamps to ensure that all lamp agents can broadcast communication.

[0059] During the initialization phase, each lighting agent sends a probe beacon containing its own identity and location information.

[0060] Each lighting agent listens to the probe beacons sent by other lighting agents.

[0061] When a certain lighting agent receives a detection beacon from another lighting device, it calculates the relative distance between itself and the signal source lighting device.

[0062] If the calculated relative distance is less than the preset communication cooperation radius, the identity of the signal source lamp is recorded in its own list of neighboring nodes.

[0063] Complete the construction of a list of all lamp neighbor nodes. This list will be used for subsequent data broadcasting scope definition and participant determination in the local collaborative game model.

[0064] (2) Ambient light detection and data sharing

[0065] Each intelligent luminaire uses its integrated ambient light sensor to detect the local ambient light intensity at its location in real time.

[0066] Each lighting agent generates a shared data packet, which includes its own identity, the real-time detected ambient light intensity value, and its current brightness value.

[0067] Each lighting agent broadcasts the generated shared data packets to all nodes in its neighbor node list through the established wireless communication network.

[0068] Each lighting agent receives shared data packets broadcast by neighboring nodes and collects relevant information about the neighboring nodes.

[0069] (3) Constructing a local collaborative game model and calculating the brightness of candidate targets

[0070] Determine the participants in the local collaborative game: The participants in the local collaborative game are the lamp agent itself and the lamps in its neighbor node list.

[0071] Define the game strategy variables: take the current brightness of each participant as the strategy variable of the local game and determine the range of values ​​for the strategy variables.

[0072] Perceive local environmental conditions: Calculate the total perceived illuminance at the current location of the lamp based on the ambient light intensity detected by itself and the shared data received from neighboring nodes.

[0073] Construct a local utility function: Construct a local utility function for the current luminaire node. This function can convert the luminaire brightness into a value that represents local benefits and is used to measure the overall effect of the brightness setting.

[0074] Fixed neighbor node strategy: In each control cycle, the current brightness of neighbor nodes is treated as a fixed value.

[0075] Search for optimal brightness: Using a numerical optimization method, a search is conducted within the feasible range of brightness values ​​to find the brightness value that maximizes the local utility function value of the current luminaire node.

[0076] Determine candidate target brightness: The brightness value that maximizes the local utility function obtained from the search is determined as the candidate target brightness of the lamp node in the current game round.

[0077] Update decision information: Update the calculated candidate target brightness value to the internal state register of the lighting agent to prepare for subsequent brightness adjustment. At the same time, this brightness value will be shared with neighboring nodes as its current brightness in the next cycle.

[0078] (4) Brightness decision and adjustment

[0079] Each luminaire agent reads the calculated candidate target brightness from its internal status register.

[0080] Each lighting agent obtains its initial brightness value from its own agent attributes.

[0081] Calculate the absolute difference between the candidate target brightness and the initial brightness value.

[0082] The calculated absolute difference in brightness is compared with a preset stability threshold.

[0083] If the brightness difference is greater than the stability threshold, it is determined that the current lighting environment has changed, the candidate target brightness is determined as the target brightness for this round, and the luminous brightness of the lamp itself is adjusted accordingly.

[0084] If the brightness difference is less than the stability threshold, the current lighting environment is determined to be stable, the original brightness is maintained, and no brightness adjustment is made.

[0085] (5) Dynamic cyclic optimization

[0086] After completing a round of brightness decision and adjustment, determine whether the current lighting environment is approaching the Nash equilibrium point. If not, return to step "(2) Ambient light detection and data sharing".

[0087] It continuously responds to changes in ambient light and network topology, enabling the entire lighting network to dynamically respond to these changes and eventually stabilize at a globally optimized Nash equilibrium point.

[0088] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An ambient light detection-based adaptive power supply control method for lighting intensity, characterized in that, The specific steps of this method are as follows: S100. Configure independent intelligent agent attributes for each lamp, establish a wireless communication network between lamps, enable all lamp intelligent agents to conduct broadcast communication, and construct a list of neighbor nodes containing all other lamps within their communication range. S200. Each lighting intelligent agent detects the local ambient light intensity at its location in real time through its integrated ambient light sensor. Through the wireless communication network, each node generates a shared data packet containing its own identity, real-time ambient light intensity value, and its current brightness, and broadcasts the data packet to all nodes in its neighbor node list. S300. Each lighting agent constructs a local cooperative game model based on its own shared data packets and the shared data received from neighboring nodes. The model uses the current brightness of the lighting agent as the decision variable and aims to maximize its local utility function to solve the problem, thereby calculating a candidate target brightness that reaches a local Nash equilibrium. In S300, the construction process of the local cooperative game model is as follows: Game element definition: Each lighting agent, along with its own list of neighboring nodes, constitutes the set of participating lighting nodes in the current local collaborative game. Each lighting node... The strategy variable is its own current brightness value. The set of all strategy variables constitutes the strategy space of this local cooperative game model; Local environmental state perception: lighting node Based on the shared data obtained by S200, the total perceived illuminance at its location is calculated. ; Construction of multi-objective utility function: for lamp node Construct a quantified local utility function The function will adjust its brightness. It is mapped to a scalar value that represents its local benefit; Nash equilibrium solution: In each control cycle, the lamp node Get the current brightness of other neighboring nodes Treating it as a fixed policy, the feasible range of the policy space is determined by numerical optimization methods. , The search is performed within the [function name] to find a local utility function that satisfies its own local utility function. Strategies to maximize This strategy That is, in the current round of the game, the lamp node. Candidate target brightness ,in, This represents the minimum luminous intensity of the light fixture. This represents the maximum luminous intensity of the lamp. Decision information update and output: Calculate the brightness values ​​of candidate targets. The internal state register of the lighting agent is updated and prepared for brightness adjustment in S400. At the same time, this brightness value will be shared with neighboring nodes as its current brightness in the next cycle S200. S400: Each lamp agent compares the candidate target brightness calculated by S300 with the initial brightness value, performs a brightness decision based on the comparison result, and adjusts its own brightness in real time. S500, repeats S200-S400 according to a preset cycle, dynamically updates ambient light intensity and neighborhood information, optimizes brightness adjustment parameters in real time, and maintains dynamic balance of lighting throughout the space.

2. The ambient light detection-based adaptive power supply control method for lighting intensity according to claim 1, characterized in that, In S100, the agent attributes include assigning a unique identifier to each lamp, setting an initial brightness value, and the position information of each lamp in space.

3. The ambient light detection-based adaptive power supply control method for lighting intensity according to claim 1, characterized in that, In step S100, the specific process of constructing the neighbor node list includes: During the initialization phase, each lighting agent sends a probe beacon containing its own identity and location information, and listens for probe beacons from other lighting agents. When a lighting agent receives a detection beacon from another lighting agent, it calculates the relative distance to the signal source. When the relative distance is less than the preset communication cooperation radius, the identity of the signal source is recorded in its own list of neighboring nodes, thereby forming a network topology focused on local area cooperation. This list of neighboring nodes serves as the scope for data broadcasting in S200 and the set of participants for constructing a local cooperative game model in S300.

4. The ambient light detection-based adaptive power supply control method for lighting intensity according to claim 1, characterized in that, The total perceived illuminance ,in, It is a lighting node The ambient light illuminance value is directly detected by its own ambient light sensor. It is a lighting node The set of neighboring nodes, Neighboring nodes The current brightness, Neighboring nodes Lighting fixture nodes The illuminance contribution weighting coefficient at the location.

5. The ambient light detection-based adaptive power supply control method for lighting intensity according to claim 1, characterized in that, The local utility function ,in, It is a lighting node The utility value, It is at the node Total illuminance perceived at the location It is the preset global illuminance value. It is a node Current brightness The power consumption function and , It is the weighting coefficient.

6. The ambient light detection-based adaptive power supply control method for lighting intensity according to claim 1, characterized in that, The specific steps of the numerical optimization method are as follows: Calculate the local utility function Regarding one's own decision variables The first-order partial derivative of is expressed as: ,in, Neighboring nodes Lighting fixture nodes The illuminance contribution weighting coefficient at the location, These are weighting coefficients. Power consumption function The derivative, It is at the node Total illuminance perceived at the location It is the preset global illuminance value; With nodes The current brightness value is used as the initial point for iteration. In each iteration, along the gradient... In the positive direction, according to the preset step size factor Update brightness value: ,in, Current brightness The updated brightness; Repeat the iterative process until any preset convergence condition is met, at which point the brightness value is obtained. This refers to the nodes in the current game situation. The candidate target brightness within this decision-making cycle, wherein the convergence condition is: The magnitude of the first partial derivative is less than the threshold. ,Right now ; The change in brightness between adjacent iterations is less than the threshold. ,Right now .

7. The ambient light detection-based adaptive power supply control method for lighting intensity according to claim 1, characterized in that, In S400, the brightness decision-making process is as follows: Lighting Node Read the calculated candidate target brightness And obtain the initial brightness value from its agent properties. Calculate the absolute difference between the two. ,Right now ; The calculated brightness difference Compared with the preset stability threshold Compare and make a decision based on the comparison results: when If the current lighting environment changes, the brightness of the candidate target will be adjusted accordingly. Adopted as the target brightness for this round; when If the brightness is not adjusted, the system will maintain its current state.

Images