Dynamic energy-saving control method and device for landscape lighting

Abstract

The invention provides a dynamic energy-saving control method and device for garden landscape lighting, and the method comprises the steps: obtaining comprehensive monitoring data, such as multi-point environment illumination, people flow distribution and time period characteristics, and generating a partition lighting demand matrix through multi-dimensional fusion processing; identifying a contradictory region between the lighting demand and the energy consumption limitation, generating a conflict level mark based on energy efficiency analysis, and reversely deducing minimum power configuration; constructing a dynamic energy-saving constraint condition, carrying out boundary testing, generating a flexible energy-saving boundary, and determining an energy consumption control range; establishing a hierarchical control architecture of a main control layer and an auxiliary control layer, and generating a basic dimming instruction sequence and a supplementary control rule; analyzing an illumination influence relation of adjacent areas to establish a linkage mechanism, and optimizing operation parameters through cooperative control; the illumination power is dynamically adjusted by adopting an envelope modulation technology, a periodic fluctuation feature is extracted to identify an optimal energy efficiency mode, a self-adaptive control strategy is constructed, and intelligent dynamic energy-saving control of garden landscape illumination is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent lighting control, and in particular to a dynamic energy-saving control method and device for garden landscape lighting. Background Art

[0002] With the acceleration of urbanization and the improvement of people's quality of life, landscape lighting has become a key means of enhancing a city's image and improving the living environment. However, traditional landscape lighting systems suffer from high energy consumption and operating costs. Many lighting facilities operate at high power levels during low-traffic hours late at night, resulting in significant energy waste. Furthermore, lighting needs vary significantly across different regions and time periods, making simple timed on / off control difficult to meet the requirements of refined management.

[0003] Existing lighting energy-saving technologies often rely on fixed dimming strategies or simple sensor-triggered control, lacking in-depth analysis and dynamic adaptability to the complexities of garden environments. These methods are unable to accurately identify conflicting lighting needs across different areas, struggle to maximize energy savings while maintaining lighting quality, and are unable to intelligently adjust to real-time environmental changes. Therefore, there is an urgent need for intelligent garden lighting energy-saving control methods. Summary of the Invention

[0004] The present invention provides a dynamic energy-saving control method and device for garden landscape lighting, which aims to generate a partitioned lighting demand matrix through multi-dimensional data fusion, identify conflicting areas between high illumination demand and low energy consumption requirements, construct dynamic energy-saving constraints and flexible energy-saving boundaries, establish a hierarchical control architecture and regional linkage mechanism, generate time-varying energy distribution curves and envelope modulation signals, and finally identify the optimal energy efficiency mode through periodic fluctuation characteristics to achieve intelligent dynamic energy-saving control of garden landscape lighting.

[0005] A first aspect of the present invention provides a dynamic energy-saving control method for garden landscape lighting, comprising the following steps: Obtaining comprehensive monitoring data of the garden landscape, the comprehensive monitoring data including multi-point environmental illumination data, crowd distribution data, and time period characteristic data, and performing multi-dimensional fusion processing on the comprehensive monitoring data to generate a zoned lighting demand matrix; Identify conflicting areas between high illumination requirements and low energy consumption requirements based on the partitioned lighting demand matrix and preset energy consumption standards, perform energy efficiency analysis on the conflicting areas to generate conflict level markers, and reversely deduce the minimum power configuration for each area based on the conflict level markers; constructing a dynamic energy-saving constraint condition according to the minimum power configuration and the time period characteristic data, performing a boundary test on the dynamic energy-saving constraint condition to generate a flexible energy-saving boundary, and determining the energy consumption control range of each area based on the flexible energy-saving boundary; generating a lighting control strategy based on the energy consumption control range, performing lighting priority analysis on the lighting control strategy to generate a main control layer and an auxiliary control layer, generating a basic dimming instruction sequence for the main control layer, and generating supplementary control rules for the auxiliary control layer based on the basic dimming instruction sequence; Analyze the mutual influence of lighting in adjacent areas according to the supplementary control rules, establish a regional linkage mechanism, coordinately control each lighting circuit through the regional linkage mechanism to obtain real-time power consumption data, and determine optimized control parameters based on the real-time power consumption data; constructing a time-varying energy distribution curve based on the optimized control parameters, generating an envelope modulation signal based on the time-varying energy distribution curve, and dynamically modulating the lighting power through the envelope modulation signal to obtain energy fluctuation state data; Periodic fluctuation characteristics are extracted from the energy fluctuation state data, an optimal energy efficiency mode is identified based on the periodic fluctuation characteristics, and an adaptive control strategy is constructed using the optimal energy efficiency mode to achieve dynamic energy-saving control of garden landscape lighting.

[0006] A second aspect of the present invention provides a dynamic energy-saving control device for garden landscape lighting, comprising: A data acquisition module is used to obtain comprehensive monitoring data of the garden landscape, which includes multi-point environmental illumination data, crowd distribution data, and time period characteristic data, and to perform multi-dimensional fusion processing on the comprehensive monitoring data to generate a zone lighting demand matrix; an energy efficiency analysis module, configured to identify conflicting areas between high illumination requirements and low energy consumption requirements based on the partitioned lighting demand matrix and preset energy consumption standards, perform energy efficiency analysis on the conflicting areas to generate conflict level markers, and reversely deduce the minimum power configuration for each area based on the conflict level markers; a boundary construction module, configured to construct a dynamic energy-saving constraint condition based on the minimum power configuration and the time period characteristic data, perform boundary testing on the dynamic energy-saving constraint condition to generate a flexible energy-saving boundary, and determine the energy consumption control range of each area based on the flexible energy-saving boundary; a hierarchical control module, configured to generate a lighting control strategy based on the energy consumption control range, perform lighting priority analysis on the lighting control strategy to generate a main control layer and an auxiliary control layer, generate a basic dimming instruction sequence for the main control layer, and generate supplementary control rules for the auxiliary control layer based on the basic dimming instruction sequence; a linkage control module, configured to analyze the mutual influence of lighting in adjacent areas according to the supplementary control rules, establish a regional linkage mechanism, coordinately control each lighting circuit through the regional linkage mechanism to obtain real-time power consumption data, and determine optimized control parameters based on the real-time power consumption data; a power modulation module, configured to construct a time-varying energy distribution curve based on the optimized control parameters, generate an envelope modulation signal based on the time-varying energy distribution curve, and dynamically modulate the lighting power using the envelope modulation signal to obtain energy fluctuation state data; An adaptive optimization module is used to extract periodic fluctuation characteristics from the energy fluctuation state data, identify the optimal energy efficiency mode based on the periodic fluctuation characteristics, and use the optimal energy efficiency mode to construct an adaptive control strategy to achieve dynamic energy-saving control of garden landscape lighting.

[0007] The beneficial effects of this invention are reflected in the following aspects: First, through multidimensional data fusion and intelligent analysis technology, it achieves accurate identification and quantitative assessment of garden landscape lighting needs. This method comprehensively considers multiple factors, such as ambient illumination, crowd distribution, and time period characteristics, to generate a zoned lighting demand matrix. This method identifies areas where high illumination requirements conflict with low energy consumption requirements. Through conflict level marking and reverse deduction techniques, it determines the minimum power configuration for each area, providing a scientific basis for the rational allocation of lighting power and avoiding energy waste caused by excessive lighting. Second, it constructs dynamic energy-saving constraints and flexible energy-saving boundaries, improving traditional fixed threshold control methods. Through disturbance testing and safety margin analysis, the control method can adaptively adjust the control boundaries, expanding energy-saving potential while ensuring lighting stability. A hierarchical control architecture prioritizes critical areas, and a regional linkage mechanism ensures lighting coordination between adjacent areas, providing technical support for the overall optimized control of garden lighting. Finally, by introducing envelope modulation technology and periodic feature analysis, this method explores the operating patterns of lighting facilities, identifies optimal energy-efficiency modes, and constructs adaptive control strategies. The data-driven intelligent control method can dynamically adjust operating parameters according to real-time environmental changes, maintain good energy efficiency performance in different seasons, weather, holidays and other scenarios, and provide a reliable technical solution for energy-saving operation and service quality assurance of garden landscape lighting.

[0008] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The accompanying drawings herein illustrate specific examples of the technical solutions described in the present invention, and together with the specific implementation methods constitute a part of the specification, and are used to explain the technical solutions, principles and effects of the present invention.

[0010] Unless otherwise specified or defined, the same reference numerals in different drawings represent the same or similar technical features, and the same or similar technical features may also be represented by different reference numerals.

[0011] Figure 1It is a flow chart of a dynamic energy-saving control method for garden landscape lighting according to the present invention.

[0012] Figure 2 This is a structural block diagram of a dynamic energy-saving control device for garden landscape lighting according to the present invention. DETAILED DESCRIPTION

[0013] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary details.

[0014] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0015] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0016] The technical solutions of the embodiments of this application are introduced below.

[0017] like Figure 1 As shown, an embodiment of the present invention provides a dynamic energy-saving control method for garden landscape lighting, including the following steps S110 to S170: Step S110: Obtain comprehensive monitoring data of the garden landscape, which includes multi-point environmental illumination data, crowd distribution data, and time period characteristic data. Perform multi-dimensional fusion processing on the comprehensive monitoring data to generate a partition lighting demand matrix.

[0018] Specifically, a distributed sensor network acquires real-time monitoring data from the garden landscape area, building a comprehensive environmental perception system. Multi-point ambient illumination data is collected using an array of light sensors arranged in a 20m x 20m grid, covering the entire garden area. Measurement accuracy reaches ±5%, and the sampling frequency is set every 10 seconds. Illuminance data includes two components: natural and artificial light. It supports real-time calculation of total illuminance, covering a range of 0-50,000 lux and supporting full-time monitoring day and night. Crowd distribution data is obtained by integrating infrared counters, video analysis, and WiFi probes. Infrared counters are deployed at key access points to count pedestrians. Video analysis uses the YOLOv5 algorithm to identify the number of people in an area in real time. WiFi probes analyze crowd concentrations by detecting mobile device MAC addresses. Crowd density is calculated using a weighted fusion formula: Density = α × Count_IR + β × Count_Video + γ × Count_WiFi, where α = 0.5, β = 0.3, and γ = 0.2 are weighting coefficients to ensure data accuracy. Time period feature data includes timestamps, season identifiers, weather conditions, and holiday markers, with time accuracy down to the minute. Seasons are divided into spring, summer, autumn, and winter. Weather conditions include sunny, cloudy, rainy, and foggy days. Holiday markers distinguish between weekdays, weekends, and statutory holidays. The data collection system utilizes an edge computing architecture, performing preliminary data preprocessing at each monitoring node, including outlier detection, data cleaning, and format standardization, to ensure data quality. Comprehensive monitoring data for the landscape is acquired through the collaborative perception of multiple heterogeneous sensors.

[0019] In some embodiments, the multi-dimensional fusion processing of the comprehensive monitoring data to generate a zoned lighting demand matrix includes: performing regional aggregation processing on the multi-point environmental illumination data to generate a basic illumination distribution map; mapping the crowd distribution data based on the basic illumination distribution map to obtain regional activity density parameters; and weighting the time period characteristic data based on the regional activity density parameters to generate a zoned lighting demand matrix.

[0020] Spatial aggregation analysis was performed on the multi-point ambient illumination data, and a continuous illumination distribution field was constructed using Kriging interpolation. Regional aggregation divided the landscape into standard 50m x 50m grid cells, each containing 4-9 monitoring points. Aggregation was performed using the area-weighted average method, calculated as Grid_Lux = Σ(Wi × Luxi) / ΣWi, where Wi represents the influence weight of each monitoring point. Wi is calculated based on the distance decay function as 1 / (1 + di²), and di represents the distance from the monitoring point to the grid center. Illumination data were interpolated using the spherical model variogram, with a nugget value set at 5%, a sill value at 95% of the total variance, and a range of 60 meters based on the monitoring point spacing. Outliers were handled using the 3σ criterion; data points exceeding ±3 standard deviations of the mean were corrected using neighborhood interpolation. A basic illumination distribution map was generated, containing four attributes for each grid cell: mean illumination, maximum illumination, minimum illumination, and illumination variance. Data were stored in GeoTIFF format with a spatial resolution of 5 meters to ensure precise representation. The distribution map is rendered using pseudo-color mapping, with low-illuminance areas displayed in blue (0-50 lux), medium-illuminance areas in green (50-200 lux), and high-illuminance areas in red (above 200 lux), for easy intuitive analysis.

[0021] Based on the constructed basic illumination distribution map, spatial registration and mapping analysis of pedestrian flow distribution data were performed to calculate the comprehensive activity intensity of each area. The mapping process first involved converting the pedestrian flow data from the sensor coordinate system to the standard landscape coordinate system, with a conversion accuracy of less than 0.5 meters. The pedestrian flow density data was spatially overlaid with the illumination grid, and the point-level pedestrian flow data was assigned to the corresponding grid cell using the nearest neighbor interpolation method. The regional activity density parameter was calculated by combining pedestrian flow density and dwell time using the formula: Activity_Density = Density × Duration_Factor. Duration_Factor is categorized by dwell time: a coefficient of 0.2 for transient transit (<30 seconds), 0.5 for short dwell times (30 seconds to 2 minutes), 0.8 for medium dwell times (2-10 minutes), and 1.0 for long dwell times (>10 minutes). Nighttime pedestrian flow correction was performed to account for the characteristics of nighttime activity. The pedestrian flow density between 22:00 and 6:00 was multiplied by a nighttime correction factor of 0.3 to reflect the specific lighting needs of nighttime. Activity hotspot identification uses kernel density estimation, using a Gaussian kernel function and a bandwidth parameter of 50 meters to identify key areas of crowd gathering. Abnormal activity detection uses time series analysis to identify sudden crowd gatherings. Areas with crowd density exceeding three standard deviations of the historical mean are marked as areas of abnormal activity. Density data normalization maps activity density to a range of 0-1, facilitating subsequent weight calculations.

[0022] Based on regional activity density parameters and combined with time period characteristic data, a zoning lighting demand matrix is ​​generated. The matrix is ​​constructed using a spatiotemporal grid model, dividing the garden landscape into M×N spatial grids, with each grid corresponding to a lighting control unit. The time period weight distribution is determined based on the lighting demand characteristics at different times of the day. The early morning period has a lower weight, the evening period has the highest weight, and the weights of other periods are between the two. The environmental correction coefficient comprehensively considers seasonal, weather, and holiday factors. The seasonal coefficient is determined based on the length of daylight hours, the weather coefficient is set according to natural lighting conditions, and the holiday coefficient is adjusted according to the increase or decrease in the flow of people. The comprehensive environmental correction coefficient is calculated by Environment_Factor=Season_Factor×Weather_Factor×Holiday_Factor to ensure that the lighting demand calculation takes into account various environmental changes. The formula for calculating the intensity of a zone's lighting demand is: Demand(i,j,t) = Activity_Density(i,j) × Time_Weight(t) × Environment_Factor, where Demand(i,j,t) represents the lighting demand intensity for the i-th row and j-th column grid during time period t, Activity_Density(i,j) is the regional activity density parameter for that grid, Time_Weight(t) is the weight coefficient for the corresponding time period, and Environment_Factor is the comprehensive environmental correction factor. The demand intensity value ranges from 0 to 10, with larger values ​​indicating greater lighting demand. The resulting zone lighting demand matrix is ​​a multidimensional array, corresponding to the spatial grids and time periods.

[0023] Step S120 , based on the partition lighting demand matrix and the preset energy consumption standard, identifies the conflicting areas between high illumination demand and low energy consumption demand, performs energy efficiency analysis on the conflicting areas to generate conflict level tags, and reversely derives the minimum power configuration of each area based on the conflict level tags.

[0024] Specifically, conflict areas are identified using a zoned lighting demand matrix combined with preset energy consumption standards for landscape lighting. The demand matrix analysis uses a threshold segmentation method to identify high-illumination demand areas (Demand[i,j,t]) with a demand intensity greater than 7. These areas are primarily located in key landscape nodes, heavily trafficked corridors, and important viewing areas. The preset energy consumption standards are based on the scale of the garden and equipment configuration, with an overall power density limit of 20W / square meter. The maximum power per 50×50 meter grid area does not exceed 500W, and the average daily power consumption is controlled within 150W. The conflict identification algorithm calculates the demand-to-energy consumption ratio for each grid using the formula: Conflict_Index = Demand_Level × Required_Power / Power_Limit, where Demand_Level is the lighting demand intensity in the demand matrix (ranging from 0 to 10), Required_Power is the theoretical power required to meet that demand intensity (in watts), Power_Limit is the power limit standard for the area (in watts), and Conflict_Index is a dimensionless conflict index. Areas identified as conflicting are those with an index greater than 1.5. Spatial connectivity analysis employed a connected domain labeling algorithm to merge adjacent conflicting grids into contiguous regions. The minimum connected area was set at 2,500 square meters (1×1 grid) to ensure that identified conflicting areas were valuable for resolution. Time series analysis analyzed the distribution of conflicting areas across time periods and found that conflicts were most concentrated between 6:00 PM and 9:00 PM, accounting for over 70% of all conflicting areas throughout the day. Regional priority was determined based on a comprehensive score of landscape importance and pedestrian density, with conflicting areas along the main landscape belt receiving the highest priority.

[0025] A quantitative energy efficiency analysis was conducted on the identified conflicting areas, and a conflict rating system was constructed. For each identified conflicting area, the specific location coordinates, demand intensity values, and degree of power overrun were extracted as inputs. The energy efficiency analysis used the Analytic Hierarchy Process (AHP) to establish an evaluation model. The primary indicators included three dimensions: lighting function importance, energy consumption constraint stringency, and technical implementation difficulty, with weights of 0.4, 0.35, and 0.25, respectively. The lighting function importance assessment was based on the actual functional attributes of the conflicting area: core landscape areas had an importance coefficient of 1.0, main corridors had a coefficient of 0.8, secondary landscape areas had a coefficient of 0.6, and marginal areas had a coefficient of 0.3. The stringency of energy consumption constraints was directly quantified using the Conflict_Index value; a higher index indicates stricter constraints. The technical implementation difficulty assessment considered the adjustability and retrofit cost of existing equipment within the conflicting area. The difficulty coefficient for LED equipment was 0.3, for dimmable traditional equipment was 0.6, and for non-dimmable equipment was 1.0. The overall score is a weighted summation method: Score = 0.4 × Importance + 0.35 × Strictness + 0.25 × Difficulty. Conflict levels are categorized into five levels: L1 (score 0-2, mild conflict), L2 (score 2-4, general conflict), L3 (score 4-6, moderate conflict), L4 (score 6-8, severe conflict), and L5 (score 8-10, extremely severe conflict). Each conflict area is assigned a corresponding level based on the overall score, with L4 and L5 areas being prioritized for resolution.

[0026] In some embodiments, the reverse derivation of the minimum power configuration of each area based on the conflict level mark includes: establishing an illuminance-power reverse mapping relationship according to the conflict level mark; performing a power reduction analysis on the conflicting area based on the illuminance-power reverse mapping relationship to obtain a power-illuminance change curve; analyzing the illuminance attenuation trend of the power-illuminance change curve to determine the critical threshold that meets the minimum illumination requirement; and determining the minimum power configuration of each area based on the critical threshold.

[0027] Based on the generated conflict level markers, a hierarchical illuminance-power reverse mapping relationship is established. Based on measured device performance data, differentiated mapping strategies are implemented for different conflict levels: L5 utilizes a maximum reduction mapping, allowing power to be reduced to 30% of the rated value; L4 utilizes a moderate reduction mapping with a power lower limit of 40%; L3 utilizes a conservative reduction mapping with a power lower limit of 60%; and L1-L2 maintain a normal power range of 80%-100%. The mapping function is corrected to account for ambient reflectance: The reflectance coefficient for hard pavement is 0.3-0.5, the reflectance coefficient for foliage is 0.1-0.2, and the reflectance coefficient for water is 0.05-0.15. The temperature compensation mechanism addresses the temperature drift characteristics of LED devices. For every 10°C increase in ambient temperature, the luminous efficacy decreases by 3%, and the mapping relationship is adjusted accordingly. The device aging factor is determined based on age: 1.0 for new devices, 0.9 for those 3-5 years old, and 0.8 for those over 5 years old.

[0028] Based on the power reduction ranges in the established reverse mapping relationship, power reduction optimization analysis was conducted for each conflicting area. Using the previously determined power lower limit parameters, L5 areas were tested at a 30% power lower limit, L4 areas at a 40% lower limit, and L3 areas at a 60% lower limit. The reduction process was implemented in stages. In the first stage, the power was reduced to 80% of the target according to the mapping relationship to evaluate the effectiveness. In the second stage, the power was adjusted to 60% of the target based on feedback. In the third stage, fine-tuning was performed to achieve the power lower limit determined by the mapping relationship. Coordination between adjacent areas took into account lighting continuity. When a reduction in one area exceeded 40% according to the mapping relationship, the power in adjacent areas was appropriately increased by 5%-10% to compensate. Reduction simulations used optical modeling software for predictions, and a three-dimensional optical model was built to verify the reduction effect and ensure that lighting quality met basic requirements. A real-time monitoring mechanism recorded power changes and illuminance responses during the reduction process, with data sampling every minute to form a continuous data sequence. By continuously recording the power setpoints and corresponding actual illuminance measurements for each reduction stage, a power-illuminance data set was established. The reduction curve fitting adopts the cubic spline interpolation method to connect the discrete power-illuminance data points into a smooth and continuous curve. The horizontal axis is the power percentage (0-100%) and the vertical axis is the corresponding illuminance value (lux), generating the power-illuminance change curve of each conflict area.

[0029] The power-illuminance curve was analyzed for attenuation patterns to determine the critical operating point of the lighting effect. A segmented fitting method was used to identify three characteristic regions of the curve: linear, nonlinear, and saturation. The linear segment corresponds to a 0-25% power reduction, with a near-linear relationship between illuminance and power, and a slope of -1.5 to -2.0 lux / W. The nonlinear segment, corresponding to a 25%-60% reduction, exhibits a gradual decrease in attenuation, with a slope varying from -2.0 to -0.8 lux / W. The saturation segment, corresponding to a reduction of more than 60%, exhibits a gradual decrease in illuminance, with a slope less than -0.5 lux / W. Minimum illumination requirements are based on national standards and garden function settings: no less than 30 lux in key landscape areas, no less than 50 lux in important corridors, no less than 15 lux in secondary areas, and no less than 8 lux in background areas. A safety margin is set to account for weather changes and equipment aging, adding a 25% safety margin to the minimum requirements. The inflection point detection algorithm identifies key turning points in the attenuation curve and uses the second-order difference method to calculate the change in the curve's curvature, locating the points where illumination attenuation accelerates and decelerates. Meteorological correction takes into account the impact of severe weather conditions such as haze and rain on lighting performance. Illumination requirements are increased by 20% in haze and by 15% in rainy days.

[0030] Based on the determined critical threshold, the minimum power allocation plan for each zone is derived. Using the critical threshold as a constraint, power consumption is minimized while meeting minimum illumination requirements. The allocation algorithm utilizes a constrained optimization approach. Constraints include that the illumination value must be greater than or equal to the critical threshold for the corresponding zone, and that the power range must be within the device's regulation capabilities. A load balancing strategy ensures the rationality of power distribution across zones, preventing excessive or insufficient power in individual zones. Device characteristic constraints consider the minimum power limits of dimming devices, with LED devices dimming to a minimum of 20% of rated power and traditional dimming devices to a minimum of 50%. A time-varying allocation plan tailors power requirements to different time periods, allowing for a further 20%-30% power reduction during nighttime and maintaining the standard configuration during daytime. A priority allocation mechanism prioritizes lighting quality in high-conflict zones, while allowing for further power reductions in low-conflict zones when necessary. Configuration verification is achieved through both field testing and simulation calculations to ensure that actual lighting results meet expectations. Contingency plans automatically restore standard power to a zone when illumination falls below the critical threshold, ensuring basic lighting safety. A maintenance adjustment mechanism regularly adjusts power configurations based on seasonal changes and equipment aging to maintain optimal operating conditions for the lighting system. Through critical threshold constraints and optimization calculations, the minimum power configuration of each area was finally determined.

[0031] Step S130 , constructing dynamic energy-saving constraints based on the minimum power configuration and time period characteristic data, performing boundary tests on the dynamic energy-saving constraints to generate flexible energy-saving boundaries, and determining the energy consumption control range of each area based on the flexible energy-saving boundaries.

[0032] Specifically, a dynamic energy-saving constraint that adapts to time-varying demand is constructed by leveraging each region's minimum power configuration and combining it with time-of-day data. Using a spatiotemporal coupling model, the minimum power configuration serves as the power lower limit, and the time-of-day data serves as a time-varying weighting factor. The constraint expression is Power_Constraint(t) = Min_Power × Time_Factor(t) × Season_Modifier × Weather_Modifier, where Min_Power is the regional minimum power configuration value derived by S120, and Time_Factor(t) is a dynamic adjustment coefficient based on time-of-day characteristics. The time-of-day adjustment coefficient is set differently for six time periods: 0.8 for early morning (12:00-6:00), 1.2 for morning (6:00-9:00), 1.0 for midday (9:00-12:00), 1.1 for afternoon (12:00-18:00), 1.5 for evening (18:00-21:00), and 0.9 for night (21:00-24:00). The seasonal modifier accounts for the effects of natural light variations, with correction coefficients of 1.0 for spring, 0.7 for summer, 1.2 for autumn, and 1.4 for winter. The weather modifier adjusts based on real-time meteorological data, with coefficients of 1.0 for sunny days, 1.1 for cloudy days, 1.3 for light rain, 1.5 for heavy rain, and 1.6 for haze. Constraint boundaries are set at both hard and soft levels. The hard constraint is the absolute lower limit for safety lighting, while the soft constraint is the ideal upper limit for energy conservation. Multi-objective constrained optimization simultaneously considers energy conservation, lighting quality, and system stability, with weights of 0.5, 0.3, and 0.2, respectively.

[0033] In some embodiments, the boundary test of the dynamic energy-saving constraint condition to generate a flexible energy-saving boundary includes: applying a disturbance signal to the dynamic energy-saving constraint condition to generate a test response; identifying a stable interval and an unstable interval based on the test response; performing a safety margin analysis on the stable interval to form an inner boundary; and constructing a flexible energy-saving boundary based on the inner boundary and the boundary of the unstable interval.

[0034] Multiple disturbance signals were applied to the constructed dynamic energy-saving constraint, and the robustness of the constraint was tested through system response analysis. The disturbances were applied using a superposition method, directly affecting the key parameters of the constraint. This resulted in the perturbed constraint expression Power_Constraint_test(t) = Power_Constraint(t) × [1 + Disturbance_Signal(t)], where Power_Constraint_test(t) is the perturbed power constraint value (in watts), Power_Constraint(t) is the constructed energy-saving constraint, and Disturbance_Signal(t) is a time-varying disturbance signal (dimensionless, ranging from -1 to 1). The power disturbance signal simulates sudden load changes and is a square wave signal with an amplitude of ±30% of the baseline power. It directly acts on the Min_Power parameter to test the stability of the minimum power configuration. The illuminance disturbance signal simulates ambient light changes and is a triangular wave signal with an amplitude of ±40% of the baseline illuminance. It indirectly affects the calculation of Time_Factor(t) by affecting the illuminance measurement results. The temperature disturbance signal simulates the impact of ambient temperature changes on equipment performance. It uses a ramp signal with a change rate of 0.5-2°C / hour to act on the Season_Modifier and Weather_Modifier parameters respectively to verify the adaptability of the seasonal and weather correction coefficients. Response monitoring records the impact of changes in constraints on system performance by measuring key parameters such as power consumption changes, illumination fluctuations, and system stability under disturbance constraints in real time. Response data acquisition uses high-precision sensors, and the data recording frequency is set to 10Hz to ensure that the transient response characteristics of the system are captured. Disturbance testing is divided into two modes: single disturbance and compound disturbance. Single disturbance tests the independent influence of each parameter, while compound disturbance tests the coupling effect of multiple factors. The test conditions cover three states: normal operation, boundary operation, and overload operation, and comprehensively evaluate the robust performance of the constraints.

[0035] Based on the acquired test response data, statistical analysis and pattern recognition methods are used to identify the system's stable and unstable intervals. Stable interval identification is based on the statistical characteristics of the response signal. Stability criteria include a mean drift rate of less than 2%, a standard deviation change of less than 5%, and a peak fluctuation of less than 10%. The response signal is preprocessed using a sliding average filter to eliminate high-frequency noise, with a window length of 30 sampling points. Stability assessment utilizes a combination of time-domain and frequency-domain analysis. Time-domain analysis calculates the first- and second-order moment statistics of the signal, while frequency-domain analysis uses FFT transforms to identify the system's natural frequency and damping characteristics. Unstable interval identification utilizes an anomaly detection algorithm, identifying instability when the response signal exhibits sustained growth, periodic oscillation, or sudden jumps. Instability types are categorized into three modes: gradual instability (slow divergence of the response), oscillatory instability (periodic oscillation of the response), and sudden instability (sudden changes in the response). A binary search algorithm is used to precisely locate the stability boundary. Refinement testing is performed near the critical points of the stable and unstable intervals, achieving boundary location accuracy of 0.5% of the parameter variation range. Interval mapping establishes a mapping between the parameter space and stability, forming a stability distribution map. Through response data analysis and statistical identification, the stable and unstable intervals were identified.

[0036] Exemplarily, the safety margin analysis of the stable interval to form an inner boundary includes: extracting boundary characteristic parameters of the stable interval; identifying environmental fluctuation factors based on the multi-point environmental illumination data; setting a safety margin for the boundary characteristic parameters according to the environmental fluctuation factors; and shrinking the boundary of the stable interval based on the safety margin to form an inner boundary.

[0037] Based on the distributed data within the stable interval, the geometric boundary features of the stable interval are extracted, and the boundary geometry is identified using a boundary tracing algorithm. Boundary feature parameters include geometric properties such as the coordinates of the boundary curve's control points, curvature radius, tangent slope, and area. Control point extraction utilizes a boundary tracing algorithm. 100 feature points are selected at equal intervals along the boundary of the stable interval and their coordinates (Pi, Ti) in parameter space are recorded, where Pi represents the power parameter and Ti represents the time parameter. Curvature calculation utilizes a three-point method, calculating the local curvature radius from three adjacent control points to identify the concave and convex characteristics of the boundary. The boundary equation is fitted using a B-spline curve, with control point spacing of 1% of the parameter range. The fitting accuracy exceeds 99% of the original boundary. Boundary gradient analysis calculates the direction and magnitude of the boundary normal vector to identify areas of most dramatic boundary variation. Feature parameter statistics include global features such as boundary length, enclosing area, center of gravity coordinates, and moment of inertia. Parameter correlation analysis assesses the degree of correlation between different boundary segments and identifies key control areas within the boundary. Boundary classification divides complex boundaries into basic elements such as straight lines, arcs, and freeform curves to facilitate subsequent processing.

[0038] Based on the multi-point environmental illuminance data obtained above, environmental fluctuation factors are identified through statistical analysis. The identification of environmental fluctuation factors uses the coefficient of variation (CV) of illuminance as the core index, and quantifies the fluctuation intensity of environmental illuminance by calculating the ratio of the standard deviation to the mean of historical illuminance data. Data processing first sorts the multi-point illuminance data in time series to establish a complete time series of illuminance changes. Statistical analysis calculates the coefficient of variation of illuminance under different time periods, different seasons, and different weather conditions to identify the basic characteristics of environmental fluctuations. Diurnal variation analysis statistically calculates the CV values of illuminance in each period to identify the time period with the most剧烈 fluctuation during the day. Seasonal variation analysis compares the differences in CV values of illuminance data in the four seasons of spring, summer, autumn, and winter to quantify seasonal fluctuation characteristics. Weather impact analysis statistically calculates the distribution of CV values under different weather conditions such as sunny, cloudy, rainy, and foggy days to establish the correlation between weather and fluctuation intensity. Abnormal fluctuation detection identifies the mutation points and outliers in illuminance data to evaluate the impact of extreme environmental events. Comprehensive evaluation calculates the overall environmental fluctuation factor CV_total as a quantitative index of environmental stability. According to the value of CV_total, the degree of environmental fluctuation is divided into three levels: high fluctuation (CV>0.3), medium fluctuation (0.1<CV≤0.3), and low fluctuation (CV≤0.1). Through the statistical analysis of illuminance data, environmental fluctuation factors are identified.

[0039] According to the identified levels of environmental fluctuation factors, combined with the extracted boundary characteristic parameters, differential safety margins are set for the boundaries. The safety margins are configured using the environmental fluctuation levels: high-fluctuation environment (CV>0.3) corresponds to a high-risk level, and the safety margin is set to 1.8; medium-fluctuation environment (0.1<CV≤0.3) corresponds to a medium-risk level, and the safety margin is set to 1.5; low-fluctuation environment (CV≤0.1) corresponds to a low-risk level, and the safety margin is set to 1.2. The influence of boundary characteristic parameters is considered in the setting of safety margins. For curved boundary segments with a small radius of curvature, due to the剧烈 change in stability, the safety margin is additionally increased by 0.2; for straight boundary segments, the stability change is gentle, and the safety margin can be appropriately reduced by 0.1. Special weather correction considers the impact of extreme weather conditions. The safety margin is increased by 0.3 for thunderstorm weather, 0.2 for foggy weather, and 0.4 for sandstorm weather. The time-varying safety margin is dynamically adjusted according to the time characteristics of environmental fluctuations. The safety margin can be appropriately reduced by 10% when the fluctuation is small at night, and increased by 15% when the fluctuation is large during the day. Equipment characteristic correction considers the sensitivity of different lighting equipment to environmental changes. The safety margin of LED equipment with strong environmental adaptability can be reduced by 0.1, and the safety margin of traditional lighting fixtures with high environmental sensitivity is increased by 0.2. Regional differential setting considers the importance of different functional areas. The safety margin of the core landscape area is appropriately increased by 20%, and that of the edge area can be reduced by 10%.

[0040] Based on a set safety margin, the stability interval boundary is indented to form a more conservative and reliable inner boundary. A proportional scaling method is used, with the shrinkage ratio calculated as: Shrink_Ratio = (Safety_Factor - 1) / Safety_Factor, where Shrink_Ratio is the boundary shrinkage ratio (ranges between 0 and 1) and Safety_Factor is the safety factor (ranges greater than 1). This formula ensures that a larger safety factor leads to a larger shrinkage ratio and a more conservative boundary, ensuring a sufficient safety margin for the indented boundary. The shrinkage process maintains the similarity of the boundary shape, achieving proportional scaling through barycentric coordinate transformation. A segmented shrinkage strategy employs differentiated shrinkage ratios for different boundary segments: a 10% increase in shrinkage for curved segments with greater curvature and a 5% reduction for straight segments. Shrinkage constraints ensure that the indented boundary still contains a reasonable parameter range, with a minimum usable range of at least 60% of the original range. Boundary smoothing smoothes the indented boundary, eliminating potential sharp corners and discontinuities. Inner boundary verification verifies the validity of the inner boundary by testing the stability of any point within the boundary, with the number of test points being no less than 200. The boundary parameter output includes complete information such as the coordinates of the inner boundary control points, the boundary equation, and the available parameter range.

[0041] Based on the generated inner boundary parameters and instability interval boundaries, a complete flexible energy-saving boundary system is constructed. A multi-level design is employed, with the inner boundary serving as the safe operating boundary and the instability boundary serving as the absolute prohibition boundary, creating a buffer zone between the two. A boundary fusion algorithm uses a weighted average method to interpolate the inner and instability boundaries to generate an intermediate transition boundary. Weights are assigned based on the distance from the boundary point to the inner boundary, with closer distances receiving greater weights to ensure a smooth transition. The flexible nature of the boundary is reflected in its adaptive adjustment capability, dynamically adjusting boundary parameters based on real-time operating conditions. Boundary zoning management divides the flexible boundary into three levels: a green safety zone, a yellow caution zone, and a red warning zone, with differentiated control strategies applied to each zone. A real-time monitoring mechanism continuously monitors the relationship between system operating status and boundaries, promptly detecting boundary violations. An early warning mechanism sets multiple warning thresholds, triggering warning signals step by step as the system approaches the instability boundary. The automatic protection function automatically initiates protective measures, including power limiting and load shifting, when it detects the system is approaching the instability boundary. By integrating the inner and instability boundaries, a flexible energy-saving boundary is constructed.

[0042] Based on the generated flexible energy-saving boundaries, the actual energy consumption control range for each zone is determined. This control range is determined using a safety margin design principle, with the actual operating boundary indented 10%-15% from the flexible boundary to ensure stable system operation under disturbances. The control range consists of three tiers: the green zone represents the recommended operating range, occupying 60%-80% of the boundary; the yellow zone represents the acceptable operating range, occupying 80%-95% of the boundary; and the red zone represents the restricted operating range, occupying 95%-100% of the boundary. Regional differentiation is implemented based on the importance and load characteristics of each area. The control range for core landscape areas is relatively loose, allowing operation within the yellow zone; while the control range for peripheral areas is more stringent, primarily operating within the green zone. A dynamic adjustment mechanism adaptively adjusts the control range based on real-time monitoring data, promptly updating parameters when continuous monitoring detects boundary drift. Linked control considers the mutual influence of adjacent zones to prevent adjustments in a single zone from affecting the overall lighting performance. The emergency control plan specifies responses when a zone exceeds the control range, including power limiting, load transfer, and system degradation. Based on flexible boundary constraints and safety margin design, the energy consumption control range of each area is finally determined.

[0043] Step S140, generate a lighting control strategy based on the energy consumption control range, perform lighting priority analysis on the lighting control strategy to generate a main control layer and an auxiliary control layer, generate a basic dimming instruction sequence for the main control layer, and generate supplementary control rules for the auxiliary control layer based on the basic dimming instruction sequence.

[0044] Specifically, a hierarchical lighting control strategy system is constructed using the energy consumption control range of each zone, employing a three-tiered structure: the green zone (60%-80% boundary range) adopts a standard control strategy, allowing normal power adjustment; the yellow zone (80%-95% boundary range) adopts a throttling control strategy, limiting power increases; and the red zone (95%-100% boundary range) adopts a limiting control strategy, forcing power reduction. The control strategy algorithm uses piecewise linear control. Within the green zone, the power adjustment response coefficient is set to 1.0, reduced to 0.6 in the yellow zone, and capped at 0.3 in the red zone. Time-varying control accounts for the varying energy consumption control ranges at different times of day. The control range can be relaxed by 15% during nighttime and tightened by 10% during peak daytime hours. A regional linkage strategy ensures coordinated control between adjacent zones. When a zone enters the yellow or red control range, adjacent zones automatically compensate with power by 5%-15%. An emergency control mechanism implements rapid response strategies when regional energy consumption exceeds the control range, including immediate power restriction, load transfer, and tiered power outages. The parameterized control strategy design supports adaptive adjustment in different seasons and weather conditions, with the parameter adjustment range within ±20% of the baseline value. The safety protection strategy ensures that the lighting does not fall below the most basic safety requirements under any control state.

[0045] Based on the lighting control strategy, a primary control layer and a secondary control layer are divided. Control strategy complexity analysis identifies areas requiring precise control: Standard control strategy areas, which allow for normal power regulation and require precise management, are prioritized for inclusion in the primary control layer. Moderation and restriction control strategy areas can employ relatively simple control logic and are suitable for management in the secondary control layer. Regional importance is assessed using the Analytic Hierarchy Process (AHP). Evaluation indicators include regional importance (weight 0.4), crowd density (weight 0.3), safety requirements (weight 0.2), and landscape value (weight 0.1). Regional importance is based on functional classification: primary importance is assigned to main landscape axes and core viewing areas; secondary importance is assigned to main corridors and secondary landscape areas; and tertiary importance is assigned to marginal and background areas. Crowd density is assessed using S110 crowd distribution data. Areas with a density greater than 5 people / 100 m2 are considered high-density, 2-5 people / 100 m2 are considered medium-density, and less than 2 people / 100 m2 are considered low-density. The comprehensive score is a weighted summation: Score = 0.4 × Importance + 0.3 × Density + 0.2 × Safety + 0.1 × Landscape. The primary control layer is determined based on the control strategy type and the comprehensive score: areas with a standard control strategy and a score greater than 7 are included in the primary control layer, requiring precise control and priority protection; areas with a restrained control strategy and a score of 5-7 are the main body of the secondary control layer; and areas with a restricted control strategy or a score less than 5 are included in the secondary control layer. The primary control layer typically covers 15-25% of the critical areas, the secondary control layer covers 60-70% of the general areas, and backup control covers the remaining areas. The architecture of the primary and secondary control layers is established through a dual analysis of control strategy complexity and regional importance.

[0046] Refined control design is performed for the master control layer zones. Combined with the control strategy type from the first step, a standardized basic dimming command sequence is generated. This command sequence construction fully considers the characteristics of different control strategies: Standard control strategy zones generate a full range of dimming commands, throttling control strategy zones limit power increase commands, and restricted control strategy zones only generate power decrease commands. The command sequence consists of four basic elements: zone ID, target power value, adjustment time, and execution priority. The target power value calculation incorporates the control strategy response factor using the formula: Target_Power = Base_Power × Demand_Factor × Strategy_Factor, where Base_Power is the zone's baseline power, Demand_Factor is the demand adjustment factor (0.5-1.5), and Strategy_Factor is the control strategy response factor (1.0 for standard control, 0.6 for throttling control, and 0.3 for restricted control). The dimming time is set based on device response characteristics and user experience: 2-5 seconds for LED devices and 5-10 seconds for traditional fixtures. Execution priority is set based on the control strategy type: Standard control strategy areas default to P2, and moderate control strategy areas to P3. This is automatically raised to P1 when an anomaly is detected. Sequential control ensures that instructions are executed in the predetermined order and time, avoiding instruction conflicts and system disruptions.

[0047] In some embodiments, the generation of supplementary control rules for the auxiliary control layer based on the basic dimming instruction sequence includes: analyzing the timing characteristics of the basic dimming instruction sequence to extract the main control rhythm; dividing the auxiliary control layer into time windows based on the main control rhythm; mapping the basic dimming instruction sequence based on the time windows to generate auxiliary control layer adjustment parameters; and constructing supplementary control rules coordinated with the main control layer based on the auxiliary control layer adjustment parameters.

[0048] In-depth time series analysis of the basic dimming command sequence is performed to extract the operational rhythm characteristics of the main control system. Time series feature analysis uses a time series decomposition method to decompose the command sequence into three components: trend term, periodic term, and random term. The trend term reflects the overall trend of power regulation at the main control layer, and a moving average filter is used to extract long-term variation patterns. The periodic term identifies periodic patterns in the command sequence, and Fourier transform is used to analyze the main frequency components to identify typical rhythms such as daily and hourly cycles. The random term represents irregular dimming actions, and residual analysis is used to identify sudden dimming demands. Main control rhythm quantification includes three key parameters: average dimming frequency (times / hour), dimming amplitude standard deviation, and dimming response time. Dimming frequency analyzes the command issuance density during different time periods to identify high-frequency and low-frequency dimming periods. Dimming amplitude analysis calculates the statistical distribution of power changes and identifies the proportional relationship between large dimming and fine dimming. Response time analysis analyzes the time distribution from command issuance to completion to determine the system response characteristics. Rhythm stability assessment evaluates the stability of the main control rhythm by calculating the coefficient of variation of rhythm parameters.

[0049] Based on the extracted main control rhythm parameters, the auxiliary control layer's control system is divided into time windows to establish a time base for synchronization with the main control layer. This time window division utilizes an adaptive segmentation algorithm, determining the window size based on the frequency characteristics of the main control rhythm. The basic time window is set at 1.5 times the average dimming interval of the main control layer, ensuring sufficient time for the auxiliary control layer to respond to the main control layer's dimming actions. The window size is dynamically adjusted from 50% to 200% of the baseline value, shortening the window during high-frequency dimming periods and extending it during low-frequency dimming periods. Four time window types are available: synchronous window (fully synchronized with the main control layer), delayed window (2-5 seconds behind the main control layer), predictive window (1-3 seconds ahead of the main control layer), and independent window (operating on a fixed cycle). The window selection strategy is based on regional correlation: synchronous windows are used for closely correlated areas, delayed windows for moderately correlated areas, and independent windows for weakly correlated areas. Window boundary processing ensures smooth transitions between window switches and avoids sudden changes in control commands.

[0050] Based on the divided time windows, the master layer's basic dimming command sequence is fully mapped onto the auxiliary layer's timeline, generating auxiliary layer-specific adjustment parameters. The mapping algorithm leverages four key elements of the command sequence: region identification for determining the corresponding auxiliary layer, target power value for calculating auxiliary layer power, adjustment time for time window alignment, and execution priority for determining the response strategy. Time mapping uses linear interpolation to convert the master layer command execution time into relative time within the auxiliary layer window. Power mapping is calculated based on the master layer's target power value and the auxiliary layer's regional characteristics: auxiliary layer power = master layer power × regional scaling factor. The regional scaling factor for auxiliary layers is typically 0.6-0.8, while for marginal auxiliary layers it's 0.3-0.5. Priority mapping maps the master layer's P1 / P2 / P3 levels to the auxiliary layer's fast / normal / delayed response modes. Adjustment parameters consist of five core elements: dimming time (determined by time mapping), dimming amplitude (calculated based on power mapping), dimming speed (set to 80%-120% of the master control layer based on priority), hold time (proportional to the master control layer's adjustment time), and response mode (determined by priority mapping). By fully mapping all information in the master control layer's command sequence, the auxiliary control layer's adjustment parameters are generated.

[0051] Based on the adjustment parameters, supplementary control rules for the auxiliary control layer are constructed, highly coordinated with the main control layer. The specific values ​​of the adjustment parameters are used: the dimming time parameter is used to determine the triggering time of the rule and establish a time-based rule schedule; the dimming amplitude parameter is converted into a power adjustment rule, defining the power adjustment range and step size for each time period; the dimming speed parameter is used to set the gradual change control rule, with fast dimming using a direct jump and slow dimming using a multi-step gradual change; the hold time parameter forms the steady-state control rule; a short hold time allows frequent adjustments, while a long hold time limits the adjustment frequency; and the response mode parameter determines the rule execution strategy: fast responses are executed immediately, normal responses are queued, and delayed responses are executed on a waiting basis. The rule system includes six basic rule categories: timing coordination rules (based on dimming time), amplitude control rules (based on dimming amplitude), speed matching rules (based on dimming speed), steady-state hold rules (based on hold time), mode switching rules (based on response mode), and exception handling rules. The rule execution engine dynamically adjusts rule parameters based on the real-time adjustment parameters to achieve adaptive control, ultimately generating supplementary control rules coordinated with the main control layer.

[0052] Step S150: Analyze the mutual influence relationship of lighting in adjacent areas based on the supplementary control rules, establish a regional linkage mechanism, coordinately control each lighting circuit through the regional linkage mechanism to obtain real-time power consumption data, and determine the optimized control parameters based on the real-time power consumption data.

[0053] In some embodiments, analyzing the mutual influence relationship of lighting in adjacent areas according to the supplementary control rules and establishing a regional linkage mechanism includes: identifying the lighting coverage relationship of adjacent areas based on the supplementary control rules; constructing an inter-regional influence matrix based on the lighting coverage relationship; generating linkage trigger conditions for each area based on the influence matrix; and establishing a regional linkage mechanism based on the linkage trigger conditions.

[0054] Based on the supplementary control rules, the lighting coverage relationships between adjacent areas are identified. First, the six basic rule categories in the supplementary control rules are extracted, focusing on analyzing the regional timing relationships defined by the timing coordination rules and the power correlations determined by the amplitude control rules. Physical coverage analysis uses the lighting equipment's light distribution curve and installation parameters to calculate the effective illumination range of each luminaire. The illumination radius is determined by luminaire type: 15-30 meters for floodlights, 5-10 meters for garden lights, and 3-5 meters for underground lights. Coverage overlap calculates the intersection of the illumination ranges of adjacent areas. The overlap formula is Overlap_Ratio = Area_intersection / Area_total, where Area_intersection is the overlapping area and Area_total is the total coverage area. An overlap greater than 30% is considered strong coverage. Functional coverage assessment, based on the mode switching rules in the supplementary control rules, determines the coverage requirements of the lighting function. Entrance areas require coverage of adjacent passageways, and landscape nodes require coverage of viewing paths. Temporal coverage analysis utilizes the response delay parameter in the supplementary control rules to identify coverage relationships formed by the lighting activation sequence. Coverage direction determines the relationship between the primary coverage area and the covered area, creating a directed coverage graph. Coverage intensity quantifies the physical and functional coverage, with a value range of 0 to 1. By analyzing supplementary control rules and coverage analysis, the lighting coverage relationship between adjacent areas is identified.

[0055] Based on the recognized lighting coverage relationship, an influence matrix reflecting the degree of mutual influence between regions is constructed. It adopts the form of an N×N square matrix, where N is the total number of control regions, and the matrix element M[i,j] represents the influence intensity of region i on region j. The calculation of matrix elements directly uses the quantization result of the coverage relationship, M[i,j]=Overlap_Ratio[i,j]×Distance_Factor×Function_Factor, where Distance_Factor = 1 / (1 + 0.1×d[i,j]) is the distance attenuation factor, d[i,j] is the distance between the region centers, and Function_Factor is determined according to the main control layer / auxiliary control layer division (1.0 for the main control layer to the main control layer, 0.8 for the main control layer to the auxiliary control layer, 0.5 for the auxiliary control layer to the auxiliary control layer). Matrix normalization ensures that the row sum is 1, indicating that the total influence of a single region is constant. Sparsification processing sets elements less than 0.1 to zero to reduce the computational complexity. Symmetry testing evaluates the degree of symmetry of the matrix, and asymmetric elements reflect one-way influence relationships. Eigenvalue analysis evaluates the system stability by calculating the matrix eigenvalues, and the maximum eigenvalue less than 1 ensures system convergence. Matrix partitioning decomposes the large matrix into multiple sub-matrices according to the region functions.

[0056] Based on the constructed influence matrix, personalized linkage trigger conditions are generated for each region. Trigger condition generation analyzes each row of the influence matrix and extracts the source regions with an influence intensity M[i,j] greater than 0.3 as potential trigger sources. Trigger event definition combines the execution modes of the supplementary control rules in S140, including power change events (change amplitude exceeding 30%), state switching events (switch state change), and mode switching events (rule mode change). Trigger strength calculation is based on the matrix element values, Trigger_Strength = M[i,j]×Event_Magnitude, where Event_Magnitude is the event amplitude (normalized 0 - 1). Composite trigger condition design supports the joint trigger of multiple source regions, and the linkage is triggered when Σ(M[i,j]×Event[i])>0.5. The time window constraint inherits the time window setting of the auxiliary control layer in S140, and the minimum trigger interval is the same as the time window size. Priority setting is divided according to the influence intensity: M[i,j]>0.6 is the P1 level for immediate trigger, 0.3<M[i,j]≤0.6 is the P2 level for a 5-second delay, and M[i,j]≤0.3 is the P3 level for a 15-second delay. Trigger parameter encoding includes the source region ID list, event type, trigger strength value, response requirements, etc. Condition verification verifies the logical completeness by traversing all possible trigger combinations.

[0057] Based on the generated linkage trigger conditions, a complete regional linkage mechanism is constructed to achieve coordinated control of the lighting system. Using the aforementioned trigger conditions as control logic, a linkage response is initiated when the trigger conditions are detected. The response action is calculated using the influence matrix and trigger strength. The linkage power adjustment is calculated as ΔP[j] = P[i] × M[i,j] × Trigger_Strength × 0.6, where ΔP[j] is the power adjustment for target zone j (in watts), P[i] is the current power of trigger source zone i (in watts), M[i,j] is the influence of zone i on zone j (ranging from 0 to 1), Trigger_Strength is the trigger strength calculated in step 3, and 0.6 is the response attenuation coefficient to ensure a moderate response amplitude and avoid overshoot. The linkage propagation path is determined based on the connectivity of the influence matrix and a breadth-first search is used to limit the search to a two-hop neighborhood. The response delay is set using the priority parameter in the trigger condition: 0-1 seconds for P1, 5 seconds for P2, and 15 seconds for P3. When multiple linkage commands affect the same region, a weighted average method is used to calculate the final adjustment: ΔP_final = Σ(ΔP[k] × M[k,j]) / Σ(M[k,j]), where ΔP_final is the final power adjustment for region j, ΔP[k] is the power adjustment from each trigger source region k, and M[k,j] is the corresponding impact strength. The sum is calculated across all source regions that have linkage effects on region j. Stability protection monitors power changes after linkage and immediately terminates linkage if oscillation is detected (power variation frequency > 0.2Hz). The linkage log records the complete chain of information for each linkage: trigger source → trigger condition → impact path → response action → actual effect, ultimately completing the establishment of the regional linkage mechanism.

[0058] Through an established regional linkage mechanism, coordinated control is implemented across each lighting circuit in the landscape, collecting real-time power consumption data. This coordinated control mechanism translates the linkage mechanism into specific control instructions, which are distributed to each lighting circuit via the intelligent power distribution system. Circuit identification determines the control unit based on the electrical topology. Each circuit contains 3-8 lighting nodes with a total power of 2-5 kW. Control sequencing takes into account the grid's carrying capacity, with high-power adjustments executed in batches, and single adjustments not exceeding 20% ​​of the total capacity. Command encoding utilizes priority queue management, and coordinated control instructions are inserted into the normal control sequence to ensure timely response. Real-time power consumption monitoring is performed using smart meters installed in each circuit, with a sampling accuracy of 0.1 kW and a sampling frequency of 10 Hz. Data is uploaded via the RS485 bus. Data preprocessing includes outlier removal, moving average filtering, and power factor correction to ensure data quality. Power consumption data is correlated with regional lighting status, establishing a complete data chain from power consumption to illumination control. A time-series database is used for data storage, supporting high-frequency writes and fast queries, storing 7 days of detailed data and 1 year of statistical data.

[0059] Based on collected real-time power consumption data, statistical analysis and optimization algorithms are used to determine the optimal control parameters. First, time series decomposition is performed to identify trend, cyclical, and random terms. The cyclical terms reflect the patterns of daily lighting demand. Energy efficiency metrics, including power consumption per unit area (W / m2), power consumption per unit illuminance (W / lux), and power utilization, are calculated, establishing a multi-dimensional evaluation system. Parameter sensitivity analysis uses perturbation testing to assess the impact of various control parameters on power consumption and identify key optimization parameters. The optimization objective function is defined as minimizing total power consumption while satisfying illuminance constraints, and is solved using the Lagrange multiplier method. Parameter optimization utilizes a particle swarm algorithm, with the search space encompassing power baseline values ​​(±30%), response coefficients (0.3-1.0), and delay times (1-10 seconds). The iterative optimization process is set to 100 generations of evolution with a population size of 50. The convergence criterion is an improvement in the objective function of less than 0.1%. Constraints include lighting standards, equipment capacity limitations, and user comfort requirements. The optimal parameter set includes a complete configuration of power setpoints for each zone, linkage response coefficients, and time-of-day adjustment factors. Through data-driven optimization analysis, the optimal control parameters were finally determined.

[0060] Step S160: construct a time-varying energy distribution curve based on the optimized control parameters, generate an envelope modulation signal based on the time-varying energy distribution curve, and dynamically modulate the lighting power through the envelope modulation signal to obtain energy fluctuation state data.

[0061] Specifically, a time-varying curve reflecting the spatiotemporal energy distribution characteristics of the garden lighting system is constructed using optimized control parameters. This time-varying energy distribution curve is based on three key parameters within the optimized control parameters: the power setpoint for each zone, the linkage response factor, and the time period adjustment factor. The curve function is defined as E(t,x,y) = P_opt(x,y) × T_factor(t) × R_coef(x,y), where E(t,x,y) is the energy distribution value at location (x,y) at time t (unit: W / m2), P_opt(x,y) is the regional power setpoint obtained from S150 optimization, T_factor(t) is the time period adjustment factor (ranging from 0.5 to 1.5), and R_coef(x,y) is the linkage response factor (ranging from 0.3 to 1.0). Time discretization uses a 5-minute sampling interval, with a day divided into 288 time slices. Spatial gridding divides the garden area into 10×10 meter grid cells, with each grid cell corresponding to an energy value. Boundary conditions ensured continuity of energy variations between adjacent grids, with gradients limited to 20% per grid. Daily periodic pattern extraction used Fourier analysis to identify the primary periodic components, retaining the fundamental frequency and the first three harmonics. Data smoothing employed a two-dimensional Gaussian filter with a kernel standard deviation of 1.5 grid cells. Outlier processing eliminated energy values ​​exceeding three times the standard deviation and repaired them using neighborhood interpolation. Time-varying energy distribution curves were constructed through parameter mapping and spatiotemporal modeling.

[0062] In some embodiments, generating an envelope modulation signal based on the time-varying energy distribution curve includes: performing envelope extraction on the time-varying energy distribution curve to obtain an energy profile; performing peak-valley feature analysis on the energy profile to generate a modulation depth parameter; generating a carrier signal based on the modulation depth parameter; and generating an envelope modulation signal based on superposition of the carrier signal and the optimized control parameter.

[0063] Envelope extraction is performed on the time-varying energy distribution curve to obtain a profile reflecting the energy trend. Envelope extraction first spatially integrates the three-dimensional energy distribution E(t,x,y) to obtain the total energy time series E_total(t) = ∬E(t,x,y)dxdy, reducing the three-dimensional data to a one-dimensional time signal. The Hilbert transform method is used to convert the one-dimensional signal E_total(t) from a real-valued energy signal into an analytical signal, and its amplitude envelope is extracted. Preprocessing steps include detrending and bandpass filtering to remove DC components and high-frequency noise, while retaining the significant frequency components between 0.01 and 0.5 Hz. Local extrema detection identifies all local maxima points in the energy curve, with a 15-minute extrema detection window. The upper envelope is constructed by connecting all local maxima points using cubic spline interpolation to form a smooth upper boundary curve. The lower envelope is processed similarly for local minima, ensuring that both the upper and lower envelopes completely encompass the original curve. Envelope smoothing uses a Savitzky-Golay filter with a 25-point window and a polynomial order of 3. Envelope verification checks whether the extracted envelope accurately reflects the energy variation trend, with deviations controlled within 5%. Multiscale envelope analysis considers energy variations at different time scales, extracting hourly and daily envelopes. Envelope feature parameters calculate descriptive metrics such as average width, rate of change, and asymmetry. Energy profiles are obtained through signal processing techniques and curve analysis.

[0064] The extracted energy profile undergoes depth feature analysis to identify peak-valley distribution patterns and generate modulation depth parameters. Peak and valley types are determined based on the first-order derivative's zero crossing point and the sign of the second-order derivative. Peak feature extraction includes parameters such as peak magnitude, peak time, peak width, and peak shape factor. Valley analysis similarly extracts features such as valley depth, valley time, and valley width. Peak-valley pairing pairs adjacent peaks and valleys based on their timing relationships to form a complete fluctuation cycle. The modulation depth is calculated using the formula Mod_Depth = (Peak_Value - Valley_Value) / (Peak_Value + Valley_Value), where Mod_Depth is the modulation depth (ranging from 0 to 1), Peak_Value is the peak energy, and Valley_Value is the valley energy. The time-varying modulation depth takes into account varying modulation requirements during different time periods, with a modulation depth of 0.2-0.4 during daytime and 0.4-0.7 at night. A depth limit is set to prevent overmodulation, with the maximum modulation depth limited to 0.8. Depth smoothing eliminates sudden changes in modulation depth using a moving average method with a window of 5 sampling points. Modulation depth mapping establishes a time-depth lookup table, supporting real-time query of the modulation depth value at any time.

[0065] Based on the generated modulation depth parameters, a carrier signal tailored to the characteristics of the lighting system is constructed. The carrier frequency is selected taking into account the persistence of vision and device response characteristics of the human eye, with a base frequency set at 0.05Hz to ensure no visible flicker. A sine wave is used as the carrier waveform, which exhibits better harmonic characteristics than a square wave. The carrier amplitude is dynamically adjusted based on the modulation depth parameter. The amplitude calculation formula is A(t) = A_base × Mod_Depth(t), where A(t) is the time-varying amplitude, A_base is the base amplitude (ranging from 0.5 to 1.0), and Mod_Depth(t) is the time-varying modulation depth generated in step 4. Phase control ensures the phase relationship between the carriers in different regions, with the main control region leading and the auxiliary control regions lagging by π / 8, respectively. Frequency modulation introduces slow frequency variations within a ±20% modulation range to enhance the naturalness of the signal. Carrier distortion control ensures THD (total harmonic distortion) is less than 5% through harmonic analysis. Digital processing discretizes the continuous carrier, with the sampling rate set to at least 20 times the carrier frequency. The carrier buffer pregenerates a complete cycle of carrier data to improve real-time generation efficiency. The carrier synchronization mechanism ensures that the carrier phases of multiple modulation channels are locked to avoid beat frequency phenomena.

[0066] Based on the generated carrier signal and the optimized control parameters described above, the final envelope modulation signal is generated through signal superposition. The superposition algorithm uses multiplication modulation. The envelope modulation signal calculation formula is S_env(t)=S_carrier(t)×(1+m×P_ctrl(t)), where S_env(t) is the envelope modulation signal, S_carrier(t) is the generated carrier signal, m is the modulation coefficient (range: 0.3-0.8), and P_ctrl(t) is the spatially averaged normalized value of the S150 power setting. P_ctrl(t) is calculated as (∬P_opt(x,y)dxdy / Area-P_min) / (P_max-P_min), ensuring that the value is between 0 and 1. The modulation coefficient is adaptively adjusted based on the real-time energy status, decreasing the modulation coefficient when energy is sufficient and increasing it when energy is limited. Signal limiting ensures that the modulation signal does not exceed the device's tolerance range, with an upper limit set at 95% of the rated value and a lower limit set at 20%. Phase compensation takes into account signal transmission and device response delays, pre-compensating for 10-50ms of phase advance. Signal smoothing provides a gradual transition at the modulation parameter switching point, with a transition time set to 2-5 seconds. Multi-channel synthesis supports simultaneous generation of multiple modulated signals, with independent parameter configuration for each channel. Signal verification verifies the frequency content of the modulated signal through spectrum analysis to ensure that no harmful frequencies are introduced. The output interface adapter converts the digital modulation signal into a standard control interface format, generating an envelope modulation signal.

[0067] Dynamic power modulation is implemented on the garden lighting system using a generated envelope modulation signal to monitor energy fluctuations in real time. Power modulation is implemented using PWM (pulse width modulation) or a 0-10V analog dimming interface to convert the digital envelope signal into a control signal recognizable by the device. Modulation mapping establishes a linear relationship between the envelope signal amplitude and actual power, with the mapping coefficient determined based on the device's rated power. A zoned modulation strategy divides the garden into eight modulation zones, each with independent modulation to prevent grid shock caused by large-scale synchronization. Modulation timing control ensures that the modulation phases of each zone are staggered, with adjacent zones differing by π / 4, resulting in smooth power fluctuations. The real-time monitoring system deploys power analyzers and illuminance sensors, with a density of at least one monitoring point per 100 square meters. Data collection parameters include instantaneous power, average power, power factor, harmonic content, and illuminance values, with a sampling rate of 10Hz. Fluctuation feature extraction calculates statistical indicators such as power mean, variance, crest factor, and fluctuation rate. Abnormal fluctuation detection identifies abnormal patterns such as sudden power changes, sustained deviations, and periodic oscillations. Data correlation analysis evaluates the correlation between the modulation signal and the actual power response, with a correlation coefficient greater than 0.9. Fluctuation state classification categorizes energy fluctuations into three types: stable, transitional, and oscillatory. Energy fluctuation state data is obtained through dynamic modulation implementation and comprehensive monitoring.

[0068] Step S170 , extracting periodic fluctuation characteristics from the energy fluctuation state data, identifying the optimal energy efficiency mode based on the periodic fluctuation characteristics, and constructing an adaptive control strategy using the optimal energy efficiency mode to achieve dynamic energy-saving control of garden landscape lighting.

[0069] Specifically, the autocorrelation function method is used to extract periodic fluctuation characteristics from energy fluctuation state data. The autocorrelation function R(τ)=Σ[P(t)×P(t+τ)] / Σ[P(t)²] identifies recurring periodic patterns in the data by calculating the correlation coefficient under different time delays τ. During the calculation, the original data is first smoothed with a 5-minute sliding window to eliminate transient disturbances and retain the main trend. The scanning range of the time delay τ is set to 0-48 hours, covering various possible period scales. The local maximum of R(τ) directly indicates the inherent period of garden lighting. When R(24h) exceeds 0.8, it indicates the presence of a strong diurnal period feature, reflecting the basic law of day and night alternation; R(12h) greater than 0.5 corresponds to a semi-diurnal period, which coincides with the peak electricity consumption period in the morning and evening; R(4-6h) in the range of 0.3-0.5 indicates the presence of short-term fluctuations, which are usually related to human activity patterns. Power spectral density analysis uses the Welch method to verify periodicity identified in the time domain. Long-term data is segmented and then spectrally averaged, effectively suppressing the influence of random noise. The peak position of the spectrum directly corresponds to the periodic frequency, and the peak height quantifies the energy contribution of the periodic component. Phase stability is statistically evaluated by phase drift over multiple consecutive days. A stable period should have a phase drift of less than π / 6. The extracted periodic fluctuation features contain complete temporal information for daily, semi-daily, and short-term cycles.

[0070] In some embodiments, the identifying of the optimal energy efficiency mode based on the periodic fluctuation characteristics includes: performing frequency domain analysis on the periodic fluctuation characteristics to extract the main frequency components; constructing an energy efficiency feature vector based on the main frequency components; generating multiple candidate operating modes based on the energy efficiency feature vector; and performing energy efficiency ratio analysis on the candidate operating modes to determine the optimal energy efficiency mode.

[0071] The extracted periodic fluctuation features were subjected to in-depth frequency domain analysis, converting the identified diurnal (24-hour), semi-diurnal (12-hour), and short-term (4-6-hour) cycles into corresponding frequency components for precise analysis. The fundamental frequency corresponding to the diurnal cycle is 1 / 86400 Hz, the semi-diurnal cycle to 2 / 86400 Hz, and the short-term cycle to the 4-6 / 86400 Hz range. This a priori frequency information guides subsequent spectral analysis. A 4096-point FFT was used for frequency domain transformation, focusing on analyzing the frequencies corresponding to these cycles and their harmonic components. Preprocessing used the least squares method to remove linear trends from the data and prevent DC components from interfering with the target frequency analysis. A Kaiser window was used for windowing to suppress spectral leakage and ensure accurate frequency localization of the periodic components. Power spectral density calculations verified the accuracy of the time domain analysis, demonstrating clear spectral peaks at the expected frequency locations. Peak detection used a threshold of three times the noise standard deviation. The spectral peaks for the diurnal and semi-diurnal cycles significantly exceeded the threshold, with energy contributions reaching 45% and 30%, respectively. Short-cycle spectral peaks within a range of 4-6 times the fundamental frequency reflect the complexity of human activity. When the main frequency components are sorted by energy contribution, the first 6-8 frequencies account for over 90% of the cumulative energy. These frequency components fully characterize the periodic energy consumption characteristics of garden lighting.

[0072] Based on the extracted primary frequency components, energy efficiency feature vectors are constructed to mathematically represent different operating states. Feature design leverages identified diurnal and semi-diurnal frequency components, extracting normalized amplitude and phase angle for each primary frequency as frequency domain features. The high amplitude of the diurnal frequency indicates strong diurnal variations in garden lighting, while its phase angle indicates the timing of peak energy consumption. Energy features are calculated from time domain data, including average power consumption reflecting baseline energy consumption, peak-to-valley ratio reflecting energy consumption fluctuations, and standard deviation characterizing stability. Feature fusion is achieved through weighted concatenation: V = [w_f × V_f, w_e × V_e], where V is the complete fused feature vector, V_f is the frequency domain feature vector, and V_e is the energy feature vector. w_f = 0.7 and w_e = 0.3 are corresponding weight coefficients, emphasizing the dominant role of frequency domain features. Vector normalization uses the L2 norm to eliminate the influence of different dimensions. Principal component analysis projects high-dimensional features into a low-dimensional space, with the first few principal components explaining over 95% of the variance. The eigenvector after dimensionality reduction retains the essence of the original periodic information, and different energy efficiency states form a clearly separable clustering structure in the feature space.

[0073] Energy efficiency feature vectors are used to generate candidate operating modes through cluster analysis, discretizing the continuous energy efficiency state into manageable operating modes. The K-means++ algorithm leverages the periodicity in the feature vectors, and the selection of initial cluster centers takes into account the distribution of diurnal characteristics. Feature vectors with similar diurnal patterns are clustered into the same mode, forming operating modes that correspond to actual application scenarios. The characteristics of the cluster center for the ultra-energy-saving mode show extremely low average power consumption and gentle diurnal fluctuations; the energy-saving mode retains a clear diurnal characteristic but reduces overall power consumption; the feature vector for the standard mode is close to the average level; and the comfort and enhanced modes exhibit a higher power consumption baseline. The clustering process is continuously adjusted through iterative optimization until the feature consistency within each mode meets the required level. For special scenario modes, such as the night mode, the feature vector has its frequency components enhanced during the nighttime period and its daytime components suppressed. Each generated operating mode inherits the periodicity of the original data and is configured with a corresponding set of control parameters.

[0074] An energy efficiency ratio analysis is conducted on candidate operating modes, comprehensively evaluating energy savings and service quality to determine the optimal mode. The energy efficiency ratio is calculated using the formula EER = Σ(wᵢ×Sᵢ) / E_norm, where EER is the energy efficiency ratio, Sᵢ is the score of the i-th service quality indicator (ranging from 0 to 1), wᵢ is the corresponding weight coefficient, and E_norm is the normalized energy consumption value. Among the four dimensions of service quality evaluation, lighting performance requires an illumination satisfaction rate exceeding 90% and a uniformity greater than 0.4; visual comfort requires a flicker index within 5%; response speed ensures that the main dimming process completes within 2 seconds; and reliability is assessed through failure rate statistics. The weights of 0.4, 0.3, 0.2, and 0.1 for each dimension reflect their varying importance in actual applications. The energy consumption assessment integrates the 24-hour power curve for each mode and normalizes it against the standard mode. The energy-saving mode achieved an energy efficiency ratio of 1.42, demonstrating a 42% reduction in energy consumption while maintaining 85% service quality. This outstanding performance stems from its rational utilization of the cyclical nature of garden lighting. The standard mode achieved an energy efficiency ratio of 1.15, serving as a benchmark. However, the enhanced mode, due to its high energy consumption, achieved an energy efficiency ratio of only 0.82. Sensitivity analysis verified the stability of the energy-saving mode under parameter perturbations, with its energy efficiency ratio fluctuating within ±8%, demonstrating optimal overall performance.

[0075] An adaptive control strategy is constructed based on the identified optimal energy-saving mode, achieving intelligent and dynamic energy conservation for garden lighting. The control architecture employs a layered design: the perception layer deploys light sensors, pedestrian detectors, and other devices to collect real-time environmental data. The decision layer runs an adaptive control algorithm, generating control commands based on optimal mode parameters and real-time feedback. The execution layer precisely adjusts the lighting power in each zone through a PWM dimming driver. The adaptive control core employs a model reference approach, using the ideal power curve in energy-saving mode as a reference trajectory and achieving tracking control through feedback adjustment. Parameter adjustment utilizes the MIT adaptive law, with the adjustment rate proportional to the tracking error and error sensitivity. The adaptive gain is set to 0.01 to ensure stable convergence. Real-time performance monitoring calculates the current energy efficiency ratio every 10 minutes and compares it with the energy-saving mode baseline value of 1.42 to assess operating status. If the energy efficiency ratio falls below 85% of the baseline value for 20 consecutive minutes, a mode evaluation program is triggered to analyze the cause and decide whether to switch operating modes. A 5-minute soft transition period is used for mode switching, and power parameters are linearly interpolated to prevent sudden changes in lighting that may impact the user experience. The adaptive parameter adjustment range is limited to ±20% of the baseline value and is dynamically optimized based on factors such as pedestrian density, ambient illumination, and weather conditions. Abnormal protection features multiple safety mechanisms. If power fluctuations exceed 30% or illumination deviations exceed 40%, the system automatically switches to safe mode to maintain basic lighting. By implementing adaptive control strategies, dynamic energy-saving control of landscape lighting is ultimately achieved.

[0076] In order to implement the garden landscape lighting dynamic energy-saving control method corresponding to the above method embodiment, to achieve the corresponding functions and technical effects. Figure 2 , Figure 2 The following is a block diagram of a dynamic energy-saving control device 200 for garden landscape lighting provided by an embodiment of the present application. For ease of explanation, only the parts related to this embodiment are shown. The dynamic energy-saving control device 200 for garden landscape lighting provided by an embodiment of the present application includes: The data acquisition module 201 is used to obtain comprehensive monitoring data of the garden landscape, which includes multi-point environmental illumination data, crowd distribution data, and time period characteristic data, and to perform multi-dimensional fusion processing on the comprehensive monitoring data to generate a zone lighting demand matrix; An energy efficiency analysis module 202 is configured to identify conflicting areas between high illumination requirements and low energy consumption requirements based on the zoned lighting demand matrix and preset energy consumption standards, perform energy efficiency analysis on the conflicting areas to generate conflict level markers, and reversely deduce the minimum power configuration for each area based on the conflict level markers; Boundary construction module 203, configured to construct dynamic energy-saving constraint conditions based on the minimum power configuration and the time period characteristic data, perform boundary testing on the dynamic energy-saving constraint conditions to generate flexible energy-saving boundaries, and determine the energy consumption control range of each area based on the flexible energy-saving boundaries; A hierarchical control module 204 is configured to generate a lighting control strategy based on the energy consumption control range, perform lighting priority analysis on the lighting control strategy to generate a main control layer and a secondary control layer, generate a basic dimming instruction sequence for the main control layer, and generate supplementary control rules for the secondary control layer based on the basic dimming instruction sequence; A linkage control module 205 is configured to analyze the mutual influence of lighting in adjacent areas according to the supplementary control rules, establish a regional linkage mechanism, coordinately control each lighting circuit through the regional linkage mechanism, obtain real-time power consumption data, and determine optimized control parameters based on the real-time power consumption data; A power modulation module 206 is configured to construct a time-varying energy distribution curve based on the optimized control parameters, generate an envelope modulation signal based on the time-varying energy distribution curve, and dynamically modulate the lighting power using the envelope modulation signal to obtain energy fluctuation state data; The adaptive optimization module 207 is used to extract periodic fluctuation characteristics from the energy fluctuation state data, identify the optimal energy efficiency mode based on the periodic fluctuation characteristics, and use the optimal energy efficiency mode to construct an adaptive control strategy to achieve dynamic energy-saving control of garden landscape lighting.

[0077] The above-mentioned garden landscape lighting dynamic energy-saving control device 200 can implement the garden landscape lighting dynamic energy-saving control method of the above-mentioned method embodiment. The optional options in the above-mentioned method embodiment also apply to this embodiment and will not be described in detail here. The remaining contents of the embodiment of this application can be referred to the contents of the above-mentioned method embodiment and will not be repeated in this embodiment.

[0078] The above embodiments are not exhaustive and may include many other embodiments not listed above. Any replacements and improvements made without violating the concept of the present invention are within the scope of protection of the present invention.

Claims

1. A garden landscape lighting energy-saving control optimization method, characterized in that: include: Obtaining comprehensive monitoring data of the garden landscape, the comprehensive monitoring data including multi-point environmental illumination data, crowd distribution data, and time period characteristic data, and performing multi-dimensional fusion processing on the comprehensive monitoring data to generate a zoned lighting demand matrix; Identify conflicting areas between high illumination requirements and low energy consumption requirements based on the partitioned lighting demand matrix and preset energy consumption standards, perform energy efficiency analysis on the conflicting areas to generate conflict level markers, and reversely deduce the minimum power configuration for each area based on the conflict level markers; constructing a dynamic energy-saving constraint condition according to the minimum power configuration and the time period characteristic data, performing a boundary test on the dynamic energy-saving constraint condition to generate a flexible energy-saving boundary, and determining the energy consumption control range of each area based on the flexible energy-saving boundary; generating a lighting control strategy based on the energy consumption control range, performing lighting priority analysis on the lighting control strategy to generate a main control layer and an auxiliary control layer, generating a basic dimming instruction sequence for the main control layer, and generating supplementary control rules for the auxiliary control layer based on the basic dimming instruction sequence; Analyze the mutual influence of lighting in adjacent areas according to the supplementary control rules, establish a regional linkage mechanism, coordinately control each lighting circuit through the regional linkage mechanism to obtain real-time power consumption data, and determine optimized control parameters based on the real-time power consumption data; constructing a time-varying energy distribution curve based on the optimized control parameters, generating an envelope modulation signal based on the time-varying energy distribution curve, and dynamically modulating the lighting power through the envelope modulation signal to obtain energy fluctuation state data; Periodic fluctuation characteristics are extracted from the energy fluctuation state data, an optimal energy efficiency mode is identified based on the periodic fluctuation characteristics, and an adaptive control strategy is constructed using the optimal energy efficiency mode to achieve dynamic energy-saving control of garden landscape lighting.

2. The method according to claim 1, characterized in that The multi-dimensional fusion processing of the comprehensive monitoring data to generate a partition lighting demand matrix includes: Performing regional aggregation processing on the multi-point environmental illumination data to generate a basic illumination distribution map; Mapping the crowd flow distribution data based on the basic illumination distribution map to obtain a regional activity density parameter; The time period characteristic data is weighted based on the regional activity density parameter to generate a partition lighting demand matrix.

3. The method according to claim 1, characterized in that The reverse derivation of the minimum power configuration of each area based on the conflict level mark includes: Establishing an illumination-power reverse mapping relationship according to the conflict level mark; Performing power reduction analysis on the conflicting area based on the illumination-power reverse mapping relationship to obtain a power-illuminance change curve; Analyzing the illumination attenuation trend of the power-illuminance variation curve to determine a critical threshold that meets the minimum illumination requirement; The minimum power configuration of each area is determined based on the critical threshold.

4. The method according to claim 1, wherein The performing boundary testing on the dynamic energy-saving constraint condition to generate a flexible energy-saving boundary includes: applying a disturbance signal to the dynamic energy-saving constraint condition to generate a test response; identifying a stable interval and an unstable interval based on the test response; Performing a safety margin analysis on the stability interval to form an inner boundary; A flexible energy-saving boundary is constructed based on the inner boundary and the unstable interval boundary.

5. The method according to claim 1, wherein The generating of a supplementary control rule for the auxiliary control layer based on the basic dimming instruction sequence includes: Analyze the timing characteristics of the basic dimming instruction sequence to extract the main control rhythm; Dividing the auxiliary control layer into time windows based on the main control rhythm; Mapping the basic dimming instruction sequence based on the time window to generate auxiliary control layer adjustment parameters; A supplementary control rule coordinated with the main control layer is constructed based on the auxiliary control layer adjustment parameters.

6. The method according to claim 1, characterized in that Analyzing the mutual influence relationship of lighting in adjacent areas according to the supplementary control rules and establishing a regional linkage mechanism includes: identifying lighting coverage relationships between adjacent areas based on the supplementary control rules; Constructing an inter-regional impact matrix based on the lighting coverage relationship; generating linkage trigger conditions for each region based on the impact matrix; A regional linkage mechanism is established based on the linkage triggering conditions.

7. The method according to claim 1, characterized in that Generating an envelope modulation signal based on the time-varying energy distribution curve includes: Performing envelope extraction on the time-varying energy distribution curve to obtain an energy profile; performing peak-valley characteristic analysis on the energy profile to generate a modulation depth parameter; generating a carrier signal based on the modulation depth parameter; An envelope modulation signal is generated based on the superposition of the carrier signal and the optimized control parameter.

8. The method according to claim 1, characterized in that The identifying the optimal energy efficiency mode based on the periodic fluctuation characteristics includes: Performing frequency domain analysis on the periodic fluctuation characteristics to extract main frequency components; constructing an energy efficiency feature vector based on the main frequency components; generating a plurality of candidate operating modes based on the energy efficiency feature vector; An energy efficiency ratio analysis is performed on the candidate operating modes to determine the optimal energy efficiency mode.

9. The method according to claim 4, characterized in that The performing of a safety margin analysis on the stability interval to form an inner boundary includes: extracting boundary characteristic parameters of the stable interval; Identifying environmental fluctuation factors based on the multi-point environmental illumination data; Setting a safety margin for the boundary characteristic parameter according to the environmental fluctuation factor; The boundary of the stability interval is contracted based on the safety margin to form an inner boundary.

10. A garden landscape lighting energy-saving control optimization device, characterized in that: include: A data acquisition module is used to obtain comprehensive monitoring data of the garden landscape, which includes multi-point environmental illumination data, crowd distribution data, and time period characteristic data, and to perform multi-dimensional fusion processing on the comprehensive monitoring data to generate a zone lighting demand matrix; an energy efficiency analysis module, configured to identify conflicting areas between high illumination requirements and low energy consumption requirements based on the partitioned lighting demand matrix and preset energy consumption standards, perform energy efficiency analysis on the conflicting areas to generate conflict level markers, and reversely deduce the minimum power configuration for each area based on the conflict level markers; a boundary construction module, configured to construct a dynamic energy-saving constraint condition based on the minimum power configuration and the time period characteristic data, perform boundary testing on the dynamic energy-saving constraint condition to generate a flexible energy-saving boundary, and determine the energy consumption control range of each area based on the flexible energy-saving boundary; a hierarchical control module, configured to generate a lighting control strategy based on the energy consumption control range, perform lighting priority analysis on the lighting control strategy to generate a main control layer and an auxiliary control layer, generate a basic dimming instruction sequence for the main control layer, and generate supplementary control rules for the auxiliary control layer based on the basic dimming instruction sequence; a linkage control module, configured to analyze the mutual influence of lighting in adjacent areas according to the supplementary control rules, establish a regional linkage mechanism, coordinately control each lighting circuit through the regional linkage mechanism to obtain real-time power consumption data, and determine optimized control parameters based on the real-time power consumption data; a power modulation module, configured to construct a time-varying energy distribution curve based on the optimized control parameters, generate an envelope modulation signal based on the time-varying energy distribution curve, and dynamically modulate the lighting power using the envelope modulation signal to obtain energy fluctuation state data; An adaptive optimization module is used to extract periodic fluctuation characteristics from the energy fluctuation state data, identify the optimal energy efficiency mode based on the periodic fluctuation characteristics, and use the optimal energy efficiency mode to construct an adaptive control strategy to achieve dynamic energy-saving control of garden landscape lighting.

Images