Method and system for LED intelligent lighting management based on cloud-edge collaboration
By adopting a cloud-edge collaborative LED smart lighting management method, the cloud generates control intentions and combines them with real-time edge perception, which solves the communication load and control complexity problems of large-scale LED smart lighting systems and achieves stable and secure lighting management and energy consumption optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHONGSHAN XINCHUANGMING ELECTRONIC TECH CO LTD
- Filing Date
- 2026-03-09
- Publication Date
- 2026-06-02
AI Technical Summary
Existing LED intelligent lighting systems suffer from high communication loads and high control complexity in large-scale and complex environments. Cloud prediction bias leads to inconsistent lighting effects, lacks security and controllability, and is difficult to meet the requirements of continuity and stability for road lighting and other applications.
By adopting a cloud-edge collaborative management approach, the cloud generates control intentions with a global optimization tendency and issues them in the form of strategies. The edge controller generates specific control commands by combining real-time perception. Safety verification is carried out by introducing control constraint strategies and smoothing coefficients to ensure lighting quality and operational safety.
It reduces communication burden and policy execution risks, avoids lighting instability caused by cloud prediction bias or sudden environmental changes, achieves better energy management and operating efficiency, and meets safety and comfort requirements.
Smart Images
Figure CN122138296A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of LED intelligent lighting management, and particularly relates to an LED intelligent lighting management method and system based on cloud-edge collaboration. Background Technology
[0002] With the continuous expansion of urban lighting, LED streetlights, landscape lights, and public area lighting systems have gradually evolved from traditional timed switching and individual lamp dimming modes to intelligent lighting systems managed uniformly through communication networks and centralized platforms. Existing similar technical solutions typically employ a centralized or quasi-centralized control architecture. A cloud-based or centralized control platform generates lighting control strategies based on time, regional divisions, and some environmental parameters, and then distributes specific dimming commands to individual lamps or regional control nodes via communication networks for execution. While these systems improve the automation level of lighting management to some extent and offer energy savings, they have gradually revealed significant shortcomings in large-scale, complex operating environments. On the one hand, existing technologies often rely on the cloud to directly generate and distribute specific brightness values or time-brightness curves. The highly specific strategy expression leads to a sharp increase in communication load and control complexity when dealing with a large number of lamps and significant regional differences. Furthermore, they are highly sensitive to network stability; delays or packet loss can easily cause inconsistent lighting effects or even safety hazards. On the other hand, cloud-based strategy generation is typically based on macro-level statistical data and forecasts, making it difficult to reflect real-time operational conditions at the edge, such as pedestrian and vehicle traffic, partial obstruction, and aging or malfunctioning lighting fixtures. When forecast deviations are significant, directly executing rigid control commands issued from the cloud can easily lead to insufficient illuminance, sudden brightness changes, or abnormal energy consumption. Furthermore, existing solutions, in pursuing energy conservation or participating in demand response, generally lack structural safeguards for the safety and controllability of lighting systems, often relying on rollback or manual intervention, which is insufficient for applications such as road lighting that demand extremely high continuity and stability.
[0003] The root cause of the above problems is that existing technologies have failed to establish a reasonable division of labor and constraint mechanism between the global optimization capabilities in the cloud and the security execution capabilities on the edge. Cloud-edge collaboration is more at the level of "data upload - command issuance" and has not yet formed an effective control mode that adapts to the scale, uncertainty and rigid security constraints of lighting systems. Summary of the Invention
[0004] The purpose of this invention is to propose a cloud-edge collaborative LED intelligent lighting management method and system to solve the above-mentioned problems.
[0005] To achieve the above objectives, a cloud-edge collaborative LED intelligent lighting management method is provided in a first aspect of the present invention, the method comprising the following steps: The cloud-based lighting management platform calculates and generates system-level lighting control targets based on the current baseline lighting level of the target area and the management expectation coefficient of the current time window; the system-level lighting control targets are used to describe the target brightness ratio that the area should achieve as a whole within the current time window; The cloud-based lighting management platform generates a control constraint strategy based on the system-level lighting control target, combined with the regional operation stability coefficient and the preset safety belt center target value, and distributes the control constraint strategy to the corresponding edge controller; wherein, the control constraint strategy is the lower boundary and upper boundary of the control constraint strategy. The edge controller receives the control constraint strategy and, in conjunction with the real-time acquisition of the lamp operation level in the current target area and the historical execution instructions of the previous cycle, generates a regional LED lighting control instruction under the constraints of the lower and upper boundaries. The edge controller receives the regional LED lighting control command and, in conjunction with the calibration relationship of each lamp in the current target area, calculates the target dimming duty cycle corresponding to each lamp and sends it to each lamp driver for execution, thus completing the intelligent lighting management for this cycle.
[0006] Preferably, the current reference lighting level of the area is obtained by the edge controller periodically collecting the operating status of all online LED lights in the target area, and the value is a proportional value between 0 and 1; the management expectation coefficient of the current time window is the operating tendency coefficient corresponding to the current time automatically matched according to the preset time period division.
[0007] Preferably, the current reference lighting level of the area is obtained by the edge controller periodically collecting the operating status of all online LED lights in the target area, specifically: Each LED luminaire driver power supply has a dimming duty cycle output interface. The edge controller reads the current duty cycle value of each luminaire through the communication bus and performs an arithmetic average of the duty cycles of all online luminaires in the target area to obtain the regional brightness ratio. The regional brightness ratio is then processed by time-weighted averaging according to a set time window and uploaded to the cloud database. The cloud uses this information to obtain the current reference lighting level of the target area.
[0008] Preferably, the cloud-based lighting management platform generates a control constraint strategy based on the system-level lighting control target, combined with the regional operational stability coefficient and the preset safety belt center target value, specifically: Based on the system-level lighting control target, the basic offset is calculated in combination with the regional operation stability coefficient, and a penalty term is added to reflect the degree of deviation between the system-level lighting control target and the preset target value of the seat belt center, so as to obtain the allowable offset range. An initial interval is constructed with the system-level lighting control target as the center and the allowable offset range as the radius. The initial interval is then trimmed with the preset safety belt boundary to obtain the lower and upper boundaries of the control constraint policy, so as to ensure that the final interval falls within the safety belt.
[0009] Preferably, the regional operational stability coefficient is obtained by performing window difference statistics and mapping on the historical sequence of the regional average duty cycle reported by the edge in the cloud.
[0010] Preferably, the steps for generating the regional LED lighting control command are as follows: The real-time operating level and the system-level lighting control target are weighted and interpolated using a smoothing coefficient to obtain candidate values that converge toward the target. Based on the candidate values, a penalty term proportional to the change in real-time operating level relative to the historical executed instructions of the previous cycle is introduced to correct the candidate values, resulting in the corrected result. The corrected result is projected into the control constraint strategy to obtain the regional LED lighting control command.
[0011] Preferably, the smoothing coefficient is derived from a preset item in the edge-side region parameter table; the region-level LED lighting control command represents the target duty cycle ratio that is uniformly executed within the region during this control cycle.
[0012] Preferably, the calibration relationship of each lamp includes the calibration parameters and target duty cycle of each lamp; the calibration parameters include the proportional mapping coefficient and offset correction amount of the corresponding lamp; the calibration parameters are obtained by multi-point calibration and linear fitting of the lamps during the system debugging phase.
[0013] Preferably, the target duty cycle is obtained by inverse algebraic solution of the regional LED lighting control command into the linear calibration relationship of the luminaire.
[0014] A second aspect of the present invention provides an LED intelligent lighting management system based on cloud-edge collaboration, the system comprising: The control target generation unit is used to calculate and generate system-level lighting control targets based on the current baseline lighting level of the target area and the management expectation coefficient of the current time window; the system-level lighting control targets are used to describe the target brightness ratio that the area as a whole should achieve within the current time window; A boundary generation unit is used to generate a control constraint strategy based on the system-level lighting control target, combined with the regional operation stability coefficient and the preset safety belt center target value, and to send the control constraint strategy to the corresponding edge controller; wherein, the control constraint strategy is the lower boundary and the upper boundary of the control constraint strategy. The instruction generation unit is used to receive the control constraint strategy and, in combination with the real-time acquisition of the lamp operation level of the current target area and the historical execution instructions of the previous cycle, generate regional LED lighting control instructions under the constraints of the lower and upper boundaries. The instruction execution unit is used to receive the regional LED lighting control instruction, and calculate the target dimming duty cycle corresponding to each lamp in combination with the calibration relationship of each lamp in the current target area, and send it to each lamp driver for execution to complete the intelligent lighting management of this cycle.
[0015] The beneficial technical effects of the present invention are at least as follows: This invention proposes a cloud-edge collaborative management approach for large-scale LED intelligent lighting systems. By introducing a more rational division of responsibilities and information expression method between cloud-based strategy generation and edge-side execution, it enables the global optimization goal to be implemented safely and stably in complex and uncertain real-world operating environments. The core innovation of this invention lies in moving away from rigid, specific dimming values directly from the cloud to the edge or luminaires. Instead, the cloud, based on information such as astronomical clocks, environmental changes, operating history, and power grid signals, formulates a globally optimized control intention and issues it in a policy expression form with constraint characteristics. This ensures consistency in policy direction while reserving necessary adjustment space for the edge. The edge control unit, based on this policy expression and combined with locally sensed real-time operating status, concretizes and performs safety checks on lighting control, ensuring that dimming behavior always meets the basic requirements of lighting quality and operational safety. In this way, this invention effectively reduces communication burden and policy execution risks without increasing system complexity, avoiding lighting instability caused by cloud prediction bias or sudden environmental changes. Simultaneously, it enables the lighting system to achieve better energy management and operational efficiency while meeting safety and comfort requirements. This technical approach aligns perfectly with the application characteristics of cloud-edge collaborative architecture in city-level lighting management, providing a new implementation path for intelligent lighting systems to move from simple centralized control to system-level management with adaptive capabilities and engineering feasibility. Attached Figure Description
[0016] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.
[0017] Figure 1 This is a flowchart of the LED intelligent lighting management method based on cloud-edge collaboration of the present invention. Detailed Implementation
[0018] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.
[0019] like Figure 1 As shown in the embodiment of the present invention, the LED smart lighting management method based on cloud-edge collaboration includes: S1. The cloud-based lighting management platform calculates and generates a system-level lighting control target based on the current baseline lighting level of the target area and the management expectation coefficient of the current time window; the system-level lighting control target is used to describe the target brightness ratio that the area should achieve as a whole within the current time window.
[0020] Specifically, under the cloud-edge collaborative architecture, the cloud-based lighting management platform first formulates system-level lighting control targets. This objective characterizes the overall lighting level that the target area should achieve within the current time window. Based on the current operational status of the area and the management intent for the time period, this process generates a region-level target quantity through a unified calculation method, providing a clear basis for the construction of subsequent constraint strategies. The formation of system-level objectives relies on the integration of existing operational data and preset management strategies, rather than directly generating dimming instructions for specific lighting fixtures, thereby achieving structural separation between the management and execution layers.
[0021] Furthermore, the current benchmark lighting level in the region The data is derived from the periodic data collection by the edge controller of the operating status of all online LED lights within the area. Specifically, each LED light driver has a dimming duty cycle output interface. The edge controller reads the current duty cycle value of each light through the communication bus and performs an arithmetic average of the duty cycles of all online lights in the area to obtain the area-level brightness ratio. This average value is then processed by time-weighted averaging according to a set time window and uploaded to the cloud database. The cloud uses this data to obtain the current area's baseline lighting level. Its value ranges from 0 to 1, representing a proportional value. Time window management expectation coefficient. This is generated by the cloud-based time strategy configuration module, which is a rule engine structure that automatically matches the operational tendency coefficient corresponding to the current time according to preset time period divisions. For example, setting a high-protection period... Set during stable operation period Set during energy-saving operation. The system automatically reads the corresponding coefficient value when the current time enters the corresponding interval.
[0022] In obtaining and Subsequently, the system-level lighting control target is formed through the following calculations: ; in, This indicates the current baseline lighting level for the area, obtained by edge data collection and cloud-based statistical processing. This represents the management expectation coefficient corresponding to the current time window, which is automatically generated by the time strategy configuration module. This represents the system-level lighting control target, used to describe the target brightness ratio that the area should achieve as a whole within the current time window. This calculation method keeps the target value continuously correlated with the current actual operating state. For example, when the current reference lighting level of the area is... And the management expectation coefficient for the current time window is At that time, the calculation yielded This indicates that the overall target brightness ratio of the area is 60%. (This is from the generation process.) Then, the system limits its boundaries according to a preset global security range, for example, limiting... The value should be between 0.3 and 1.0 to ensure that the operation of the area lighting meets the established safety requirements.
[0023] S2. The cloud-based lighting management platform generates a control constraint strategy based on the system-level lighting control target, combined with the regional operation stability coefficient and the preset safety belt center target value, and sends the control constraint strategy to the corresponding edge controller; wherein, the control constraint strategy is the lower boundary and upper boundary of the control constraint strategy.
[0024] Specifically, this step uses the system-level lighting control target output from step one. As numerical input, the "single-point target" is transformed into an "executable constraint interval" in the cloud, enabling the edge side to generate regional-level LED lighting control commands within a clearly defined and legal range. The mathematical foundation used here comes from classic approaches to interval analysis and constraint optimization: first, constructing a symmetrical interval around the target value. Then trim the interval to the safety belt configured during deployment. The internal margin is designed to meet engineering boundaries. Unlike common fixed margins, this solution incorporates a margin... The design is a combination of "basic offset term + center deviation penalty term": the basic offset term reflects the impact of regional operational stability on the adjustable space; the center deviation penalty term, derived from regularization (which can be viewed as a Tikhonov regularization form with quadratic penalty), is used to convergently adjust the margin when the target value deviates from the long-term recommended center, making the constraint boundary more closely match the stability requirements of road / park lighting. Because , , , , All quantities are expressed in proportional form and have been normalized. The quantities in the above calculations are all dimensionless proportional quantities. Multiplication, addition, and squaring operations are mathematically self-consistent, and the constructed... upper and lower limits of the interval and Maintaining units of the same type (all are proportional quantities) is in accordance with common sense.
[0025] Furthermore, to obtain the margin related to the regional operational characteristics, the cloud first obtains the regional operational stability coefficient from the operational database. . The raw data source is the "regional average duty cycle sequence" periodically reported by the edge controller. This sequence is obtained by reading the duty cycle driven by online lights within the region on the edge side and averaging it. The cloud performs adjacent window differencing on this sequence according to time windows, takes the average of the absolute values of the differences as the fluctuation intensity, and then compresses it through the mapping relationship fixed during deployment. get Then construct the allowable offset range. This construction starts from the classic empirical linear term "target ratio is positively correlated with margin", and... As a scaling factor, This expresses the idea that "the more unstable the situation, the more lenient the execution space needs to be"; based on this, a secondary penalty is introduced. Its prototype comes from quadratic regularization (squared deviation penalty), used to address deviations from the target in the long-term recommender center. The allowance is then controllably adjusted. The allowable offset is obtained by superimposing the two parts. : ; in, The system-level lighting control target output in step one is generated by the cloud according to a time window and provided to this step for use. The regional operational stability coefficient is derived from the window difference statistics and mapping of the historical sequence of the regional average duty cycle reported by the edge in the cloud. The convergence coefficient is derived from the cloud-based regional configuration table (pre-configured and stored according to road grade, park type, or operation and maintenance zone). The target value for the seat belt center is derived from the same configuration table (used to express the long-term recommended target center position for this area). To allow for the offset magnitude, it serves as the width parameter for subsequently constructing the constraint interval. The logical relationship of this formula is: first, using... and Calculate the base offset, then use and The squared deviation plus the penalty term yields the final offset magnitude, thus... It simultaneously reflects both the "current target scale" and the "regional stability characteristics".
[0026] Furthermore, after obtaining Subsequently, the cloud-based classic interval pruning operation merges the symmetrical interval with the seatbelt boundary to obtain the control constraint strategy. upper and lower boundaries and The derivation logic is as follows: first construct the original boundary. and Then, take values not less than the lower boundary. The value, with respect to the upper boundary, is not greater than The value of is used to ensure that the final interval falls within the safety belt: ; in, and The seatbelt boundary parameters are derived from the cloud-based regional configuration table and are fixed during system deployment. The allowable offset range is calculated using the above formula; and These are the lower and upper boundaries of the control constraint strategy, respectively, calculated by the numerical pruning logic of the cloud strategy service; The core content is , , This is composed of elements that are directly used as boundary constraints by the edge side when generating control commands in the next step. The relationship between the two equations above is as follows: the first equation is first composed of... Push The second form is then from and Push , This forms a continuous derivation chain from the "target" to the "constraint interval".
[0027] A set of computational examples corresponding to lighting scenarios is provided to demonstrate a workable substitution process. Let the system-level lighting control target for a certain park road zone, formed in step one, be... The cloud-based system obtains the operational stability coefficient of the partition from historical window difference statistics. The configuration table provides a recommendation center. With convergence coefficient Meanwhile, the seat belt configuration is , Substitute into the first equation to calculate. First calculate the basic offset. ; then calculate the penalty items Adding the two together yields Substitute into the second equation to calculate the boundary: the original lower boundary. Original upper boundary ; after the seat belt is cut, the lower boundary upper boundary .
[0028] Therefore, the control and constraint strategy generated by this time window The effective interval is And retain the central objective Align the target direction when materializing the edge side.
[0029] S3. The edge controller receives the control constraint strategy and, in conjunction with the real-time acquisition of the lamp operation level in the current target area and the historical execution instructions of the previous cycle, generates a regional LED lighting control instruction under the constraints of the lower and upper boundaries.
[0030] Specifically, this step involves distributing control and constraint strategies via the cloud. Generate zone-level LED lighting control instructions under specified conditions The previous stage already set the system-level lighting control targets. Convert to allowable control range And as The core content is sent to the edge controller, therefore this step is based on... As the convergence center, with As a hard boundary, the real-time regional operating status obtained from the edge side forms control commands that can be directly issued to the lighting fixture drivers. The mathematical foundation of this algorithm comes from first-order inertial / exponential smoothing updates in discrete control (equivalent to linear interpolation between the current state and the target state) and interval projection in constrained optimization (clipping candidate values to a closed interval). Based on this, this step introduces a "variation penalty term" to suppress frequent fine-tuning caused by communication jitter, lighting fixture response differences, or short-period fluctuations, forming a control command generation method more suitable for large-scale regional synchronous dimming. Because in this scheme… , , , , All are proportions between 0 and 1. , It is also a proportionality coefficient. When participating in linear combination, difference and trimming operations, its dimensions are consistent, and the calculation result is still a proportional quantity, which satisfies common sense.
[0031] Furthermore, the edge controller acquires the current real-time operating level of the region within a control cycle. . The raw data comes from the dimming duty cycle interface of the lamp driver power supply: the edge controller polls the online lamp drivers in the area via the fieldbus, reads the current duty cycle value of each lamp, removes offline or communication failure lamps, and then calculates the arithmetic average of the online lamp duty cycles to obtain the value. The edge controller simultaneously reads the region-level control commands executed in the previous time window from local non-volatile memory. This value is written after the previous window is successfully issued, and is used to characterize the execution state of the previous round. Based on the classic first-order smooth update, the candidate control quantity can be given by the interpolated form of "convergence of the current state towards the target", that is... ,in The parameter table for the edge region contains smoothing coefficients used to adjust the transition speed; a change penalty term is constructed based on this. The prototype of this term originates from the idea of differential penalty / damping (using the difference between the current state and the previous instruction as the suppression factor), and is used to reduce the impact of short-cycle jitter on the control output. Also derived from the parameter table of the edge region. The interpolation term and penalty term are combined to obtain candidate values, which are then constrained to a certain range using interval projection (equivalently implemented by the projection operator in constrained optimization). Internally, this forms the final regional-level control commands. : ; in, , , All from input strategies ; The real-time operating level is obtained by the edge controller through polling the lamp driver duty cycle interface and taking the regional average. Execute instructions for the previous window read from local storage by the edge controller; The smoothing coefficient is derived from a preset item in the edge-side region parameter table; The change penalty coefficient is derived from the same parameter table; The regional LED lighting control command output in this step represents the target duty cycle ratio for unified execution within the region during this control cycle. The derivation of this formula is as follows: first, interpolation terms are used to achieve... Towards The continuous transition is then suppressed by a difference penalty term. relatively The deviation caused by the output disturbance is finally addressed by interval projection to ensure that the output satisfies the hard boundary conditions given in the previous step. This allows the "target center + allowed interval" to be implemented as an edge-executable control quantity.
[0032] Furthermore, a set of computational examples illustrates the substitution and computation process. A cloud-based distribution strategy is assumed. Give , , The edge controller polls the region's real-time operating level during the current control cycle. ;Read instructions from the previous window from local storage ;Regional parameter table configuration , First, calculate the interpolation term by substituting it into the formula: ; Then calculate the penalty item: Combine to obtain candidate values Perform interval projection: first compare with the lower bound. Then compare with the upper bound. Finally obtained The edge controller will Write the target duty cycle field into the area control message and generate a dimming command frame according to the lighting drive protocol: for loops that support broadcast addresses, send the command frame synchronously; for loops that do not support broadcast, send the command frame in batches according to the lighting address. This ensures that area coverage is completed within the same control cycle, so that multiple LED lights in the area can be coordinated and adjusted with a consistent target duty cycle.
[0033] S4. The edge controller receives the regional LED lighting control command, and calculates the target dimming duty cycle corresponding to each lamp based on the calibration relationship of each lamp in the current target area, and sends it to each lamp driver for execution to complete the intelligent lighting management of this cycle.
[0034] Specifically, this step uses the area-level LED lighting control commands generated by S3. As input, the regional proportional control quantity is translated into executable dimming parameters for the luminaire driver layer, and the coordinated execution of multiple luminaires within the region is completed. The mathematical basis for this step comes from linear calibration models and inverse algebraic solutions: In LED dimming engineering, the ratio between the driver input (e.g., PWM duty cycle or digital dimming code) and the output luminous flux is often calibrated to form an approximately linear relationship, typically in the form of... ,in Indicates the proportion of target luminous flux, Indicates the first The dimming duty cycle ratio of each lamp, and These are the calibration parameters for the luminaire. Based on this classic calibration relationship, this step uses algebraic transformation to obtain a directly executable dimming duty cycle calculation formula, and then uses grouped synchronous distribution in the execution link to ensure that the area luminaires complete consistent dimming within the same control cycle. Because... , , , Both are expressed using proportional quantities between 0 and 1. The dimensions of the same type are consistent on both sides of the linear relationship, and the inverse calculation result is still a proportional quantity, which is consistent with the common sense of driver dimming input.
[0035] Furthermore, the edge controller receives Then, the calibration parameters for each lamp are read from the local lamp characteristic table. and The source of this characteristic table is the calibration record from the system commissioning phase: during commissioning, each lamp was driven at several preset duty cycle points, and the duty cycle read back by the driver was recorded as the ratio of the relative brightness measured on-site or by a lux meter. Subsequently, linear fitting was performed on these points in the edge controller to obtain the characteristic table. and The data is then written to non-volatile memory in the form of "lamp address - parameter pairs". During operation, the edge controller reads the corresponding parameters according to the lamp address index, forming a mapping specific to the individual lamp differences. This is based on the calibration relationship. The inverse algebraic solution yields the first... The target duty cycle that each lighting fixture should achieve : ; in, The regional LED lighting control command is generated and issued in step three. For the first The proportional mapping coefficients for each lamp come from the lamp calibration parameter table stored in the edge controller; For the first The offset correction for each lamp is also derived from the calibration parameter table; This is the target duty cycle ratio that the luminaire driver should perform. This formula is derived from... Transpose to obtain Divide both sides by get The derivation chain is clear and consistent with the classic linear model.
[0036] get Subsequently, the edge controller writes the data into the dimming field of the lighting driver protocol and executes the distribution. The execution link is organized according to the fieldbus capabilities: for protocols that support multicast or broadcast, the edge controller constructs frames for distribution within the same period by loop or logical group; for lamp-by-lamp addressing protocols, the edge controller sends dimming frames in the order of lamp addresses within the same control cycle, and carries a timestamp or same-cycle sequence number in the frame, so that the driver updates the PWM output according to the same cycle boundary after receiving the frame, thereby realizing regional synchronous dimming. The PWM module or constant current dimming module on the driver side writes the received target duty cycle into the register, outputs the PWM signal corresponding to the duty cycle, drives the LED chip to work at that duty cycle, and completes the physical dimming action.
[0037] A set of quantitative calculation examples is given to demonstrate the substitution process. Assume that step three outputs a region-level control command. A certain lighting fixture corresponds to the edge controller calibration parameter table. , Substituting into the inverse formula, we get: First calculate , then calculate The edge controller will The dimming command frame corresponding to the lamp address is written and sent within this control cycle; after receiving it, the driver writes the duty cycle into the PWM register and updates the output. The duty cycle is approximately 0.665263, and the lamp's output luminous flux reaches the target ratio according to its calibration relationship. Repeat the above "read parameters - calculate" process for other lights in the area. The process of "frame creation and distribution" involves completing the coordinated dimming of multiple LED lights, ultimately forming a regional-level control command. The corresponding actual lighting operating status.
[0038] This invention also provides an LED intelligent lighting management system based on cloud-edge collaboration, the system comprising: The control target generation unit is used to calculate and generate system-level lighting control targets based on the current baseline lighting level of the target area and the management expectation coefficient of the current time window; the system-level lighting control targets are used to describe the target brightness ratio that the area as a whole should achieve within the current time window; A boundary generation unit is used to generate a control constraint strategy based on the system-level lighting control target, combined with the regional operation stability coefficient and the preset safety belt center target value, and to send the control constraint strategy to the corresponding edge controller; wherein, the control constraint strategy is the lower boundary and the upper boundary of the control constraint strategy. The instruction generation unit is used to receive the control constraint strategy and, in combination with the real-time acquisition of the lamp operation level of the current target area and the historical execution instructions of the previous cycle, generate regional LED lighting control instructions under the constraints of the lower and upper boundaries. The instruction execution unit is used to receive the regional LED lighting control instruction, and calculate the target dimming duty cycle corresponding to each lamp in combination with the calibration relationship of each lamp in the current target area, and send it to each lamp driver for execution to complete the intelligent lighting management of this cycle.
[0039] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0040] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or units may be electrical, mechanical, or other forms.
[0041] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0042] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
Claims
1. A cloud-edge collaborative LED intelligent lighting management method, characterized in that, The method includes: The cloud-based lighting management platform calculates and generates system-level lighting control targets based on the current baseline lighting level of the target area and the management expectation coefficient of the current time window; the system-level lighting control targets are used to describe the target brightness ratio that the area should achieve as a whole within the current time window; The cloud-based lighting management platform generates a control constraint strategy based on the system-level lighting control target, combined with the regional operation stability coefficient and the preset safety belt center target value, and distributes the control constraint strategy to the corresponding edge controller; wherein, the control constraint strategy is the lower boundary and upper boundary of the control constraint strategy. The edge controller receives the control constraint strategy and, in conjunction with the real-time acquisition of the lamp operation level in the current target area and the historical execution instructions of the previous cycle, generates a regional LED lighting control instruction under the constraints of the lower and upper boundaries. The edge controller receives the regional LED lighting control command and, in conjunction with the calibration relationship of each lamp in the current target area, calculates the target dimming duty cycle corresponding to each lamp and sends it to each lamp driver for execution, thus completing the intelligent lighting management for this cycle.
2. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 1, characterized in that, The current baseline lighting level of the area is obtained by the edge controller periodically collecting the operating status of all online LED lights in the target area, and the value is a proportional value between 0 and 1; the management expectation coefficient of the current time window is the operating tendency coefficient that is automatically matched to the current time according to the preset time period division.
3. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 2, characterized in that, The current baseline lighting level of the area is obtained by the edge controller periodically collecting data on the operating status of all online LED lights within the target area, specifically: Each LED luminaire driver power supply has a dimming duty cycle output interface. The edge controller reads the current duty cycle value of each luminaire through the communication bus and performs an arithmetic average of the duty cycles of all online luminaires in the target area to obtain the regional brightness ratio. The regional brightness ratio is then processed by time-weighted averaging according to a set time window and uploaded to the cloud database. The cloud uses this information to obtain the current reference lighting level of the target area.
4. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 1, characterized in that, The cloud-based lighting management platform generates a control constraint strategy based on the system-level lighting control target, combined with the regional operational stability coefficient and the preset safety belt center target value, specifically: Based on the system-level lighting control target, the basic offset is calculated in combination with the regional operation stability coefficient, and a penalty term is added to reflect the degree of deviation between the system-level lighting control target and the preset target value of the seat belt center, so as to obtain the allowable offset range. An initial interval is constructed with the system-level lighting control target as the center and the allowable offset range as the radius. The initial interval is then trimmed with the preset safety belt boundary to obtain the lower and upper boundaries of the control constraint policy, so as to ensure that the final interval falls within the safety belt.
5. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 1, characterized in that, The regional operational stability coefficient is derived from the window difference statistics and mapping of the historical sequence of the average duty cycle of the region reported by the edge in the cloud.
6. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 1, characterized in that, The steps for generating the regional LED lighting control command are as follows: The real-time operating level and the system-level lighting control target are weighted and interpolated using a smoothing coefficient to obtain candidate values that converge toward the target. Based on the candidate values, a penalty term proportional to the change in real-time operating level relative to the historical executed instructions of the previous cycle is introduced to correct the candidate values, resulting in the corrected result. The corrected result is projected into the control constraint strategy to obtain the regional LED lighting control command.
7. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 6, characterized in that, The smoothing coefficient is derived from a preset item in the edge-side region parameter table; the region-level LED lighting control command represents the target duty cycle ratio that is uniformly executed within the region during this control cycle.
8. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 1, characterized in that, The calibration relationship of each lamp includes the calibration parameters and target duty cycle of each lamp; the calibration parameters include the proportional mapping coefficient and offset correction amount of the corresponding lamp; the calibration parameters are obtained by multi-point calibration and linear fitting of the lamps during the system debugging phase.
9. The LED intelligent lighting management method based on cloud-edge collaboration according to claim 8, characterized in that, The target duty cycle is obtained by inverse algebraic solution of the regional LED lighting control command into the linear calibration relationship of the luminaire.
10. A cloud-edge collaborative LED intelligent lighting management system, characterized in that, The system includes: The control target generation unit is used to calculate and generate system-level lighting control targets based on the current baseline lighting level of the target area and the management expectation coefficient of the current time window; the system-level lighting control targets are used to describe the target brightness ratio that the area as a whole should achieve within the current time window; A boundary generation unit is used to generate a control constraint strategy based on the system-level lighting control target, combined with the regional operation stability coefficient and the preset safety belt center target value, and to send the control constraint strategy to the corresponding edge controller; wherein, the control constraint strategy is the lower boundary and the upper boundary of the control constraint strategy. The instruction generation unit is used to receive the control constraint strategy and, in combination with the real-time acquisition of the lamp operation level of the current target area and the historical execution instructions of the previous cycle, generate regional LED lighting control instructions under the constraints of the lower and upper boundaries. The instruction execution unit is used to receive the regional LED lighting control instruction, and calculate the target dimming duty cycle corresponding to each lamp in combination with the calibration relationship of each lamp in the current target area, and send it to each lamp driver for execution to complete the intelligent lighting management of this cycle.