Metro intelligent lighting energy-saving control method

Abstract

The invention relates to the technical field of program control systems, in particular to a metro intelligent lighting energy-saving control method, which comprises the following steps of: judging a train approaching state of a platform area based on a preloaded train running route and a real-time train arrival detection signal, and determining a train approaching state according to a personnel flow rate; and setting a lighting activation starting time point and duration of the station area, and generating a station activation time window. Based on the dynamic fusion of the train running route and the real-time pull-in detection signal, the train approaching state of the platform area can be identified, the activation opportunity is judged through the participation of the personnel flow, and the time error of lighting start and stop is avoided; illuminance sensor data are introduced into an activation time window, and the brightness demand is calculated by combining the real-time personnel number, so that the illumination adjustment process is converted from static fixed value control to dynamic sensing regulation; illumination parameters are determined in a brightness percentage mapping mode, and responsiveness and environment adaptability of the dimming process are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of program control systems, and particularly to a subway intelligent lighting energy-saving control method. Background Art

[0002] The technical field of program control systems refers to a technical system that uses logical operations, state recognition, data collection, and execution mechanisms to precisely control devices, processes, or systems in a programmed manner. This field widely covers technical contents such as automation control, embedded systems, industrial process scheduling, human-computer interaction logic, and remote communication interfaces, relying on the cooperation of sensors, actuators, microprocessing units, communication protocols, and control algorithms to achieve dynamic monitoring of the state of target objects and control responses.

[0003] In the prior art, the logic setting and response mechanism of the control system mainly rely on fixed logic chains, lacking the fine-grained response ability to environmental changes, and prone to problems such as control lag or energy waste in the rail transit environment with strong time dynamics and changing scenarios. At the operation and maintenance level, the maintenance strategy is mainly based on periodic manual inspections and discrete alarm processing, lacking the linkage analysis of various state comprehensive factors, resulting in mismatches between maintenance resource allocation and response priorities, and actual problems such as increased maintenance costs and untimely responses. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of the present invention is to solve the deficiencies existing in the prior art and propose a subway intelligent lighting energy-saving control method.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions. A subway intelligent lighting energy-saving control method includes the following steps:

[0006] Based on the pre-loaded train operation route and the real-time train arrival detection signal, judge the approaching state of the train in the platform area, and combine with the passenger flow to set the starting time point and duration of the lighting activation in the platform area, and generate a platform activation time window;

[0007] Based on the platform activation time window, read the environmental illuminance sensor value in the platform area, establish a current environmental light reference value, and based on the current environmental light reference value, compare with the number of people to determine the target brightness level percentage and obtain the target lighting brightness parameter;

[0008] Based on the target lighting brightness parameter, select the lighting unit control protocol code, establish a preliminary control code sequence, and based on the preliminary control code sequence, combine to form a complete dimming instruction for the lighting unit and generate a standardized lighting control instruction;

[0009] Obtain the lamp status data of the subway lighting equipment, identify the lighting unit numbers that do not match the standardized lighting control instruction status, establish a list of abnormal lamps, and based on the abnormal lamp list, match the faulty lamp numbers in the list with preset maintenance trigger conditions to generate a designated lamp maintenance activation signal.

[0010] Preferably, the step of acquiring the station activation time window is:

[0011] According to the train route, the track number and train direction of the target platform are retrieved, the train entry timestamp in the current scheduling cycle is extracted from the real-time train entry detection signal, and the corresponding departure time stamp of the train is found from the scheduling plan, and the difference between the entry and departure is calculated to obtain the planned stay time of the train in the platform section;

[0012] According to the planned stop time of the train in the platform section, collect the personnel flow per unit area and the personnel flow velocity data of the corresponding channel in each sampling unit time period in the platform area during this period, normalize the flow value with the standard reference density, normalize the flow velocity value with the reference flow velocity value, and calculate the lighting duration;

[0013] The train entry detection time is set as the lighting activation start time point, and the lighting activation end time point is set according to the lighting duration to generate a platform activation time window.

[0014] Preferably, the step of obtaining the current ambient light reference value is:

[0015] Based on the platform activation time window, locate the start time and end time boundaries of the corresponding activation period in the current platform area, extract the original numerical sequence of illuminance in the activation time period, and generate a set of original values ​​of ambient illuminance in the platform area;

[0016] Based on the original value set of ambient light illuminance in the platform area, the original value sequence of illuminance is classified and aggregated according to the sensor number, and the sampling period consistency check is performed on each type of original value sequence of illuminance, and the data segments with discontinuity, lag or drift are eliminated to obtain the effective value set of ambient light illuminance in the platform area;

[0017] Based on the effective value set of ambient light illumination in the platform area, priority screening is performed according to the spatial arrangement position and stability of each group of sensors, and the illumination value representing the overall brightness level of the current platform is extracted to generate the current ambient light reference value.

[0018] Preferably, the step of acquiring the target lighting brightness parameter is:

[0019] Based on the current ambient light reference value, retrieve all the turnstile statistical data in the platform area at the current time point, accumulate all the entry and exit records according to the lighting control time segments, count the total number of people present in each time segment, and divide the total number of people in each segment by the effective usable area of the platform area to obtain the sequence of the number of people present per unit area corresponding to each time period;

[0020] According to the sequence of the number of people present per unit area and the current ambient light reference value, perform normalization processing on each pair of data items in the order of time segment indexing, and calculate the target brightness level percentage;

[0021] Based on the target brightness level percentage, calculate the difference in brightness components with the current ambient light reference value, and obtain the required lighting supplementary light output through linear compensation to form the target lighting brightness parameter.

[0022] Preferably, the step of obtaining the preliminary control code sequence is as follows:

[0023] Based on the target lighting brightness parameter, match the lighting unit control protocol mapping table preset in the lighting system, retrieve the lighting unit control protocol number corresponding to the level according to the level interval to which the current lighting brightness parameter belongs, and perform encoding structure analysis on the number to obtain the lighting unit control protocol code;

[0024] Based on the lighting unit control protocol code, extract the lighting unit address structure, instruction field format, and function bit identification method defined in the control protocol code. According to the physical layout order and response priority order of the lighting units in the platform area, sort the same protocol codes according to the address structure to generate a structured lighting unit instruction set;

[0025] Based on the structured lighting unit instruction set, merge and encode the protocol codes and corresponding control fields of all lighting units according to the operation timing to generate a preliminary control code sequence.

[0026] Preferably, the step of obtaining the standardized lighting control instruction is as follows:

[0027] Based on the preliminary control code sequence, analyze the lighting unit control protocol field and operation instruction field in each control frame, and sequentially extract the dimming level, execution delay, and switch status coding content according to the physical position number of the lighting unit to generate a dimming control field set for the lighting unit;

[0028] Based on the dimming control field set for the lighting unit, splice each control field with the corresponding lighting unit address instruction, unify the dimming level coding format, and supplement the protocol header identifier, synchronization bit, and check bit to form a dimming instruction sequence with a complete dimming logic structure;

[0029] Based on the dimming instruction sequence with a complete dimming logic structure, perform instruction encapsulation format conversion, integrate all lighting unit dimming instructions into the same control data frame structure, and generate a standardized lighting control instruction.

[0030] Preferably, the step of obtaining the abnormal lamp list is as follows:

[0031] Obtain the lamp status data of the subway lighting equipment, read the current brightness level, electrical status, working response flag and number information of each lighting unit through the protocol port, and form a set of current status data of the lighting equipment;

[0032] Based on the set of current status data of the lighting equipment, make a one-by-one comparison according to the matching rule between the lighting unit number and the standardized lighting control instruction, judge whether the current status of each lighting unit meets the brightness level and control response logic required by the standardized lighting control instruction, screen out the lighting unit numbers with inconsistent status, mark the non-conformity type as brightness anomaly, and summarize and store them uniformly as the abnormal lamp list.

[0033] Preferably, the step of obtaining the specified lamp maintenance activation signal is as follows:

[0034] Based on the abnormal lamp list, retrieve the failure type, cumulative failure duration, failure response times and status switching frequency of each lighting unit, extract the number of each lamp and the corresponding status parameters, and generate a list of abnormal status data corresponding to each lamp number;

[0035] According to the list of abnormal status data corresponding to each lamp number, calculate the maintenance trigger factor of each lamp;

[0036] Based on the maintenance trigger factor, judge one by one whether all lamps exceed the maintenance trigger threshold, screen out all lamp numbers that meet the conditions, and generate a specified lamp maintenance activation signal.

[0037] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0038] Based on the dynamic fusion of the train operation route and the real-time train approach detection signal, the present invention can identify the approaching state of the train in the platform area, participate in judging the activation timing through the passenger flow, and avoid the time error of lighting start and stop; introduce the illuminance sensor data within the activation time window, and calculate the brightness requirement in combination with the real-time number of people, so that the lighting adjustment process changes from static fixed-value control to dynamic perception control; determine the lighting parameters through the brightness percentage mapping method, improve the responsiveness and environmental adaptability of the dimming process; in the selection of the control protocol code and the preliminary sequence construction steps, bind the control command to the physical lighting unit, and establish an executable structure that can be arranged, sorted, and verified in the coding rule, so that the control process has consistency; compare the lighting instruction status with the lamp feedback status item by item, identify the abnormal state lamps and build a list, and perform item-by-item matching through the maintenance trigger factor to achieve the quantitative evaluation of the fault level, improve the timeliness of maintenance response and the rationality of priority judgment. It enhances the energy-saving ability of the subway lighting system in a multi-variable dynamic environment and the intelligent level of equipment management. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0041] Please refer to Figure 1 , the present invention provides a technical solution, a subway intelligent lighting energy-saving control method, including the following steps:

[0042] Based on the pre-loaded train operation route and the real-time train approach detection signal, judge the approaching state of the train in the platform area, combine the passenger flow, set the starting time point and duration of the lighting activation in the platform area, and generate the platform activation time window;

[0043] Based on the platform activation time window, read the environmental illuminance sensor value in the platform area, establish the current environmental light reference value, compare based on the current environmental light reference value and combine the number of people to determine the target brightness level percentage, and obtain the target lighting brightness parameter;

[0044] Based on the target lighting brightness parameter, select the lighting unit control protocol code, establish the preliminary control code sequence, and based on the preliminary control code sequence, combine to form a complete dimming instruction for the lighting unit, and generate a standardized lighting control instruction;

[0045] Obtain the lamp status data of subway lighting equipment, identify the lighting unit numbers that do not match the standardized lighting control instruction status, establish a list of abnormal lamps, and based on the list of abnormal lamps, match the fault lamp numbers in the list with the preset maintenance trigger conditions to generate a maintenance activation signal for the specified lamps.

[0046] The steps to obtain the platform activation time window are as follows:

[0047] According to the train operation route, retrieve the track number and train driving direction of the target platform, extract the train arrival timestamp within the current dispatching cycle from the real-time train arrival detection signal, and find the corresponding departure planned timestamp of the train from the dispatching plan, calculate the difference between arrival and departure to obtain the planned residence time of the train in the platform section;

[0048] According to the planned residence time of the train in the platform section, collect the personnel flow per unit area and the personnel flow velocity data of the corresponding channels in each sampling unit time period in the platform area during this period, normalize the flow value with the standard reference density, and normalize the flow velocity value with the reference flow velocity value, and calculate the lighting duration. The calculation formula is:

[0049]

[0050] Among them, T z is the lighting duration, D e is the train departure planned time, D s is the train arrival detection time, P i ′ = P i / P ref is the ratio of the personnel flow per unit area value in the i-th time period to the reference personnel density, V i ′ = V i / V ref is the ratio of the channel flow velocity value in the i-th time period to the reference flow velocity, g is the number of sampling time periods during the train residence time period, and k is the compensation factor of the time dimension, with the unit of seconds, used to convert the normalized adjustment term into a time unit;

[0051] Set the train arrival detection time as the lighting activation start time point, and set the lighting activation end time point according to the lighting duration to generate the platform activation time window.

[0052] Specifically, according to the train operation route data, which details the stopping stations of each train line, the physical track numbers corresponding to each platform, and the standard running direction of the train in each section. First, through the received current train identifier, such as the train number "G7053", and the name of the target stopping platform, such as "Central Hub Station", a query operation is performed in the train operation route database to retrieve and precisely match the specific track number used when train "G7053" stops at "Central Hub Station", such as "Track 3 Upward", and at the same time obtain its preset driving direction, such as "From South to North". Then, the system continuously monitors the train real-time detection system interlocked with the platform entrance, which consists of axle counters or induction coils installed beside the track. When the train wheels pass over a specific detection point, a signal containing precise time information is generated. From this series of signals, for the current concerned scheduling period, such as the entire time span from the start of operation on the current day (05:00:00) to the end of operation on the next day (01:00:00), the accurate timestamp when train "G7053" triggers the inbound detection point (such as the detector 50 meters in front of the starting end of the platform) is extracted and recorded as "May 17, 2025, 10:15:08", which is the train inbound detection time. Subsequently, based on the train identifier "G7053" and the target platform "Central Hub Station", the system accesses the train scheduling plan database, which stores the planned arrival and departure timetables of all trains at each station, and retrieves and extracts the planned departure timestamp of train "G7053" at "Central Hub Station", such as "May 17, 2025, 10:20:00". Finally, by converting this planned departure timestamp to the number of seconds since a common reference point (such as zero o'clock on the current day) (such as 37200 seconds) and subtracting the number of seconds after converting the previously obtained train inbound detection timestamp (such as 36908 seconds) (37200 - 36908 = 292 seconds), the time difference between the two is calculated, thereby obtaining the planned residence time of the train in the platform section, which is 292 seconds.

[0053] Formula: The advantage of this formula is that it can dynamically adjust the duration of lighting according to the planned stop time of the train and the real-time personnel flow and velocity within the platform area, ensure sufficient lighting during the train's stay and passenger evacuation, and at the same time avoid unnecessary energy waste when the platform is idle or there are few people. By introducing the normalized personnel density index P′ i and the personnel velocity index V′ i, combined with the logarithmic function to handle the density impact and the inverse ratio to adjust the flow rate impact, achieving refined perception and response to complex passenger flow conditions. The setting of the compensation factor k enables the system to be flexibly adjusted according to the specific characteristics and energy-saving goals of different stations, and finally achieves the overall goal of improving the energy efficiency of the subway lighting system while ensuring safety and comfort;

[0054] D e is the planned departure time of the train, and the acquisition steps are as follows: D e Obtained by querying the train operation dispatching plan database of the subway operation control center. This database stores the scheduled arrival and departure times of each train at each station. For a specific train schedule (e.g., identified by the train number) and the target platform, retrieve and read its planned departure timestamp, which is pre-arranged based on the train operation diagram. For example, for the train numbered "SN001" at the "Square Station", its planned departure time queried from the dispatching database is "2025-05-17 14:35:00".

[0055] D s is the train arrival detection time, and the acquisition steps are as follows: D s Obtained in real-time through the train automatic identification and positioning system installed at the entrance of the platform track. This system usually uses axle counters, track circuits, or wireless communication-based train speed measurement and positioning devices. When the train head reaches the predetermined detection point, the system records the current timestamp. For example, when the front axle of the "SN001" train triggers the axle counter sensor at the entrance of Platform 2 of the "Square Station", the recorded arrival detection time is "2025-05-17 14:30:05".

[0056] P′ i is the ratio of the passenger flow value per unit area in the i-th time period to the reference passenger density, and the acquisition steps are as follows: First, obtain the passenger flow value per unit area P in the i-th time period i , then obtain the reference passenger density P ref , and then calculate P′ i =P i / P ref .

[0057] P iThe acquisition steps of P are as follows: By using the video surveillance cameras deployed in the platform area in combination with intelligent video analysis algorithms, or through devices such as infrared sensors and pressure-sensing floors, the instantaneous number of people in the designated area of the platform is counted within a preset sampling time interval (e.g., every 5 seconds), and this number is divided by the effective area of the region (e.g., the effective area of the platform is 500 square meters) to obtain the number of people per unit area. For example, in the 3rd 5-second time period, the video analysis system identifies 150 people in the platform area, then P3 = 150 people / 500m 2 = 0.3 people / m 2 .

[0058] P ref The acquisition steps of P are as follows: Referring to the personnel density P ref , if the upper limit of the passenger density on the platform is 0.5 people / m 2 , then set P ref = 0.5 people / m 2 , and this value is configured in the system parameters by the subway operation management department according to the specifications and the characteristics of this station.

[0059] Therefore, if P3 = 0.3 people / m 2 and P ref = 0.5 people / m 2 , then P′3 = 0.3 / 0.5 = 0.6.

[0060] V′ i is the ratio of the channel flow velocity value in the i-th time period to the reference flow velocity. The acquisition steps are as follows: First, obtain the channel flow velocity value V i in the i-th time period, then obtain the reference flow velocity V ref , and then calculate V′ i = V i / V ref .

[0061] V i The acquisition steps are as follows: By using the platform video surveillance system, the average moving speed of passengers in the main passageways (e.g., near the stairways and escalator entrances) is analyzed through target tracking. In each sampling time interval, the moving trajectories of several representative passengers are selected, and their average speed is calculated as the channel flow velocity value for this time period. For example, in the 3rd 5-second time period, the average moving speed of passengers in the main passageway area of the platform obtained through video analysis is 0.8 m / s, then V3 = 0.8 m / s.

[0062] V ref The acquisition steps are as follows: For example, according to pedestrian flow research, the normal walking speed is about 1.2 m / s. Considering the possible short-term congestion and luggage carrying in the subway station, the reference flow velocity is set as V ref = 1.0 m / s.

[0063] Therefore, if V3 = 0.8 m / s and V ref = 1.0 m / s, then V'3 = 0.8 / 1.0 = 0.8.

[0064] g is the number of sampling time periods within the train's stop time period, and the steps to obtain it are as follows: According to the train's departure plan time D e and the train's arrival detection time D s calculate the total number of seconds T that the train is planned to stay at the platform dwell = D e - D s , and combined with the preset duration T of a single sampling unit time period sample (for example, 5 seconds), through the formula calculate and obtain. For example, if D e is 14:35:00, D s is 14:30:05, then T dwell = 295 seconds. If T sample = 5 seconds, then

[0065] k is the compensation factor for the time dimension, with the unit of seconds. The steps to obtain it are as follows: The determination of the k value is based on the statistical analysis and optimization objectives of the lighting delay requirements under different passenger flow conditions at a specific subway platform. First, collect historical data and record the average value of the normalized adjustment items at different time periods and the corresponding ideal lighting delay T determined by manual evaluation or passenger feedback ideal_ext , then, determine k by establishing the relationship between T ideal_ext and X avg . For example, if the analysis shows that when X avg is 0.15, an additional 60 seconds of lighting time is usually required to ensure that all passengers can leave comfortably, then k can be set as k = T ideal_ext / X avg = 60 s / 0.15 = 400 s.

[0066] Calculation process:

[0067] The train's departure plan time D e = 14:35:00 (converted to the number of seconds of the day is 14×3600 + 35×60 = 50400 + 2100 = 52500 seconds).

[0068] The train's arrival detection time D s = 14:30:05 (converted to the number of seconds of the day is 14×3600 + 30×60 + 5 = 50400 + 1800 + 5 = 52205 seconds).

[0069] The planned stay time of the train in the platform section (D e-D s ) = 52500 s - 52205 s = 295 s.

[0070] Duration T of a single sampling unit time period sample = 5 s.

[0071] Number of sampling time periods within the train's stop time period

[0072] Compensation factor k for the time dimension = 400 s.

[0073] For example, within g = 59 sampling time periods, P' i and V' i data for each time period were obtained. To simplify the calculation process, here, for example, the average value of all sampling points is 0.12, that is

[0074] Then the lighting duration T z is calculated as follows:

[0075]

[0076] T z = 295 s + 400 s·0.12;

[0077] T z = 295 s + 48 s;

[0078] T z = 343 s;

[0079] This result indicates that: After comprehensively considering the planned stop time of the train and the dynamic real-time personnel flow and flow velocity in the platform area, the calculated lighting duration is 343 seconds. This duration is 48 seconds longer than the simple planned stop time of the train, which is 295 seconds. This additional 48 seconds is dynamically adjusted based on the current personnel activities on the platform to ensure that the platform area remains sufficiently illuminated until all passengers are fully evacuated. If the calculated value of T z is significantly greater than (D e - D s ), it indicates that the personnel density on the platform is high or the evacuation speed is slow, and longer lighting time is required. Conversely, if this value is close to (D e - D s ), it means that there are few people on the platform or the evacuation is rapid. This value of T z will be used as the direct basis for setting the activation time window of the platform lighting to determine the termination time point of lighting activation.

[0080] The train arrival detection time obtained from the real-time train arrival detection signal, for example, this timestamp is "May 17, 2025, 14:30:05", is directly set as the starting time point for the activation of the platform area lighting without any delay or adjustment, ensuring that as soon as the train enters the platform area, the lighting in the relevant area is activated to the predetermined or calculated brightness level. Then, the system calls the lighting duration T calculated in the previous step z , and its value is, for example, 343 seconds. This T z is a comprehensive value that combines the planned train stop time and the total lighting demand time dynamically adjusted according to the real-time passenger density and flow rate. Based on the previously set lighting activation starting time point "May 17, 2025, 14:30:05", adding this lighting duration of 343 seconds for a time accumulation operation, that is, 14:30:05 + 343 seconds. First, convert 343 seconds into minutes and seconds, which is 5 minutes and 43 seconds, and then perform the time addition: 14:30:05 + 5:43 = 14:35:48. This calculation result "May 17, 2025, 14:35:48" is set as the lighting activation termination time point related to this train service. Finally, combining the determined lighting activation starting time point and the calculated lighting activation termination time point together defines a specific time interval, that is, from "May 17, 2025, 14:30:05" to "May 17, 2025, 14:35:48". This time interval with clear start and end moments constitutes the platform activation time window corresponding to this train's stop service

[0081] The steps for obtaining the current ambient light reference value are as follows:

[0082] Based on the platform activation time window, locate the start and end time boundaries of the corresponding activation period within the current platform area, extract the original sequence of illuminance values during the activation period, and generate a set of original ambient illuminance values for the platform area

[0083] Based on the set of original ambient illuminance values for the platform area, classify and aggregate the original sequence of illuminance values according to the sensor number, and perform a consistency check on the sampling period for each type of original sequence of illuminance values, excluding data segments with interrupted sampling, lag, or drift to obtain a set of effective ambient illuminance values for the platform area

[0084] Based on the set of effective ambient illuminance values for the platform area, perform priority screening according to the spatial layout position and stability of each group of sensors, extract the illuminance value representing the overall brightness level of the current platform, and generate the current ambient light reference value

[0085] Specifically, based on the platform activation time window determined in the foregoing steps, which defines the specific time period that the lighting control system needs to focus on. For example, the start time is "May 17, 2025, 14:30:05" and the end time is "May 17, 2025, 14:35:48". The system first uses the start and end time boundaries as the start and end constraints for data query, accesses the time series database storing the historical data of each illuminance sensor, and targets all activated illuminance sensors deployed in the current target platform area (such as the entire "No. 2 Down Platform" of "Square Station", specifically including the platform waiting board, the inner side of the platform screen door, and the area within the boundary of the stairway escalator entrance). These sensors are digital illuminance sensors of the same model, and each sensor is configured with a unique device identification code (such as "LXS-P2S-001" to "LXS-P2S-015"), and continuously monitors and records the ambient light intensity at its location at a fixed sampling frequency (for example, the system is set to collect data every 2 seconds), and the data unit is uniformly in lux. The system will traverse the log records of all sensors associated with the target platform in the database, and filter out all illuminance reading entries whose timestamp attribute values are greater than or equal to the start time of the platform activation time window ("May 17, 2025, 14:30:05") and less than or equal to the end time ("May 17, 2025, 14:35:48"). For example, sensor "LXS-P2S-001" may have recorded 170 readings during this time period ((343 seconds / 2 seconds per time) is approximately equal to 171 sampling points, with slight differences depending on boundary handling), such as "Time: 14:30:06, Illuminance: 125 lux", "Time: 14:30:08, Illuminance: 126 lux", until "Time: 14:35:48, Illuminance: 130 lux". Gather the illuminance readings, their corresponding timestamps, and sensor numbers collected by all sensors during this precise time period to form a raw data set containing multiple time series. This set is the original value set of the ambient illuminance in the platform area.

[0086] Based on the set of original values of the ambient light intensity in the platform area generated from the previous paragraph, which is a list of all the original readings collected by all relevant light intensity sensors on the platform within the specified activation time window. First, the system preprocesses this data. According to the sensor number information carried by each data point, for example, with "LXS-P2S-001", "LXS-P2S-002", etc. as the key values, the sequence of original light intensity values belonging to the same sensor is classified and aggregated to form multiple independent light intensity time series indexed by the sensor number. Subsequently, the system performs a sampling period consistency check on each independent light intensity time series. The specific check steps are as follows: First, detect the situation of data interruption. The system checks the difference between the timestamps of two adjacent data points in the sequence. This difference should theoretically be equal to the preset sampling period, such as 2 seconds. If the actual difference is greater than N miss times the preset sampling period. For example, N miss is set to 2.5 times (i.e., the time interval exceeds 2.5×2 seconds = 5 seconds), it is considered that data interruption occurs here, and the data segment between this sampling point and the subsequent normal sampling point until the next one is marked as unreliable. Second, detect the situation of data lag. The system compares the timestamp carried by the sensor data with the timestamp when the data is received and recorded by the central system. If the difference between the two exceeds the preset maximum network transmission and processing delay threshold T lag , for example, T lag is set to 1 second (this value is determined based on the network delay data at the 99.5% percentile statistically obtained from long-term monitoring of the subway internal network environment), then this data point is marked as lagged. Third, detect the situation of data drift. Calculate the short-term mean and standard deviation of each data point and several of its adjacent data points (for example, 2 before and after each, forming a 5-point sliding window). If the value of a certain data point deviates from its short-term mean by more than M dev times the short-term standard deviation. For example, M dev is set to 3 times, and this deviation is not caused by a drastic change in the real ambient light reflected by adjacent other sensors together (judged by comparing the data change trends of surrounding sensors), then this data point may have drift. Or, if the readings of a sensor remain completely unchanged or change very little (for example, the fluctuation is less than 1 lux) for a long period of time (for example, continuously exceeding 30 sampling points, that is, 60 seconds), while other sensors show normal fluctuations, it is also judged as drift. The system removes all the data points or data segments marked as data interruption, lag, or drift from the corresponding original light intensity value sequence. The remaining data points that pass the check form the set of effective values of the ambient light intensity in the platform area.

[0087] Based on the set of valid ambient light intensity values of the platform area obtained in the previous paragraph, this set contains a reliable sequence of light intensity readings from each sensor that has passed quality verification. The system further filters and fuses this data according to preset rules. First, the system accesses a pre-configured sensor metadata database, which stores the static attributes of each light intensity sensor, including its precise three-dimensional spatial coordinates (x, y, z), a detailed physical location description of the installation (such as "the middle ceiling of Platform 2, serial number 005"), and the importance level of this location for the assessment of platform lighting uniformity and safety (for example, divided into three levels: level 1 represents critical areas such as the concentrated passenger boarding and alighting areas and stairwells, level 2 represents general waiting areas, and level 3 represents edge or auxiliary areas). This importance level is predefined by lighting design experts based on the station layout and passenger flow characteristics and entered into the system. For example, a location weight of 3 is assigned to level 1 areas, 2 to level 2, and 1 to level 3. At the same time, the system will evaluate the working stability of each sensor based on its historical operation records over a certain period in the past (such as the most recent 7 days), mainly examining its data efficiency (i.e., the percentage of valid readings in the total readings that should be collected). For example, a data efficiency higher than 99% is rated as high stability (stability weight is 1.0), 95% - 99% is medium stability (stability weight is 0.8), and lower than 95% is low stability (stability weight is 0.6). These weight thresholds are set by analyzing historical sensor failure rates and data quality distributions. Subsequently, for each valid reading within the current platform activation time window, the system calculates a comprehensive priority score by combining the location importance weight and stability weight of its sensor. The scoring rule can be: Comprehensive Priority Score P priority,j = Location Weight j × 0.7 + Stability Weight j × 0.3. The coefficients 0.7 and 0.3 reflect the relative emphasis on location factors. The system selects the N sensors with the highest comprehensive priority scores at the current time point (or the most recent sampling period) (for example, for a medium-sized platform, N may be set to 5 to 7, and this quantity comprehensively considers computational efficiency and result representativeness to ensure coverage of at least the main area types), and fuses their latest valid light intensity readings L j using the weighted average method. The calculation formula is: This calculation result, a single lux value, is used as the current ambient light reference value representing the overall ambient light level of the current platform.

[0088] The steps to obtain the target lighting brightness parameters are as follows:

[0089] Based on the current environmental light reference value, retrieve all the turnstile statistical data in the platform area at the current time point. Accumulate all the entry and exit records according to the lighting control time segments, count the total number of people present in each time segment, and divide the total number of people in each segment by the effective usable area of the platform area to obtain the sequence of the number of people present per unit area corresponding to each time segment;

[0090] According to the sequence of the number of people present per unit area and the current environmental light reference value, perform normalization processing on each pair of data items in the order of time segment indexing, and calculate the target brightness level percentage. The calculation formula is:

[0091]

[0092] where L p is the target brightness level percentage, N i ′ = N i / N ref is the ratio of the number of people present per unit area in the i-th time segment to the reference personnel density, E i ′ = E i / E ref is the ratio of the environmental light reference value in the i-th time segment to the reference light intensity, g is the total number of sampling time segments, M is the maximum number of brightness levels supported by the lighting system, and the value is a dimensionless integer;

[0093] Based on the target brightness level percentage, calculate the difference in brightness components with the current environmental light reference value, and obtain the required lighting supplementary light output through linear compensation to form the target lighting brightness parameter.

[0094] Specifically, based on the current ambient light reference value obtained in the previous step, this single lux value, such as 120 lux, represents the current background light level in the platform area. First, the system sends a data request to the subway automatic fare collection system or the passenger information system to retrieve all the original entry and exit records of all turnstiles (for example, all 10 turnstiles numbered from AFC-G01 to AFC-G10) associated with the current platform service area within the most recent complete "lighting control time segment" cycle. The cycle of a "lighting control time segment" is preset to 5 minutes. For example, if the current time is 14:36:00, then the data within the time period from 14:30:00 to 14:35:00 is retrieved. These data usually include fields such as record ID, turnstile ID, transaction type (entry / exit), and transaction timestamp. Next, the system classifies all the retrieved entry and exit records into the corresponding 5-minute "lighting control time segments" according to their transaction timestamps. For each segment, the system independently counts the total number of inbound passengers and the total number of outbound passengers within that segment, and obtains the instantaneous total number of people N in the platform area at the end of each 5-minute time period through iterative calculation (the number of people present in the current segment = the number of people present in the previous segment + the number of inbound passengers in the current segment - the number of outbound passengers in the current segment, and the number of people present in the previous segment of the first segment can be based on historical data or set to 0) or by analyzing the snapshot number of people at a specific time point (such as the end of each segment). total,i , Subsequently, the total number of people N present in each time period total,i is divided by the preset effective use area A of this platform area platform (for example, the effective use area of Platform 2 of Square Station is measured and entered into the system as 650 square meters). The calculation formula is N i = N total,i / A platform , thereby obtaining a series of the number of people per unit area N corresponding to each 5-minute time period i (unit: people per square meter). These N i values are arranged in chronological order, forming a sequence of the number of people per unit area corresponding to each time period.

[0095] Formula: The benefit of the formula is that this formula can dynamically calculate the target brightness level percentage, comprehensively considering two core factors: the number of people per unit area and the current ambient light level. By squaring the normalized number of people ((N' i ) 2 ), the weight of the lighting demand for high-density crowds is enhanced. At the same time, by taking the square root of the normalized ambient light and adding 1 to the denominator The stronger the existing light, the smaller the contribution to the demand for supplementary lighting, but its impact is smoothed, avoiding large fluctuations in lighting demand caused by minor light changes. Finally, it is normalized by dividing by the maximum brightness level series M of the lighting system to ensure the output of L p is a relative control value that matches the system's capabilities, thus achieving refined energy-saving control while ensuring visual requirements;

[0096] N′ i is the ratio of the number of people present per unit area in the i-th time period to the reference personnel density. The steps to obtain it are as follows: First, obtain the number of people present per unit area N i (person / m 2 ), and this data is from the output result of the previous step, that is, the i-th element in the "sequence of the number of people present per unit area corresponding to each time period". Then, obtain the reference personnel density N ref (person / m 2 ). Finally, through the formula N′ i =Ni / N ref calculate to obtain. N′ i is a dimensionless relative value that reflects the degree of the current personnel density relative to the reference standard. For example, if N i =0.6 person / m 2 , and N ref =0.75 person / m 2 , then N′ i =0.6 / 0.75 = 0.8.

[0097] N ref (reference personnel density) is obtained as follows: This parameter represents the reference benchmark of the personnel density considered reasonable or requiring attention in the subway platform design or operation management, with the unit of person per square meter. After evaluation by the operator, N ref is set to 0.75 person / m 2 .

[0098] E′ i is the ratio of the environmental light reference value in the i-th time period to the reference light intensity. The steps to obtain it are as follows: First, obtain the environmental light reference value E i (lux). For example, if the current environmental light reference value E i =120 lux, and the reference light intensity E ref =200 lux, then E′ i =120 / 200 = 0.6.

[0099] E refThe acquisition steps for (reference light intensity) are as follows: This parameter represents the standard light intensity required under ideal conditions or in design specifications for the subway platform. For example, if the standard stipulates that the average maintained illuminance in the subway platform area should not be lower than 200 lux, then the operation and management department can set E ref to 200 lux.

[0100] g is the total number of sampling time periods, and its acquisition steps are as follows: This parameter refers to the total number of time periods obtained by dividing the current L p calculation period (for example, evaluated based on data from the most recent 30 minutes) according to the "lighting control time segmentation" (for example, each segment is 5 minutes). If the evaluation period is 30 minutes and each segment is 5 minutes, then g = 30 minutes / 5 minutes / segment = 6 time periods.

[0101] M is the maximum number of brightness level grades supported by the lighting system, and its acquisition steps are as follows: This parameter is a dimensionless integer that represents the maximum number of controllable discrete levels into which the brightness of the lamps installed on the subway platform can be divided by the intelligent lighting control system. For example, if the DALI dimming system used on a certain platform is logically divided into 20 adjustable brightness levels (from the lowest to the highest), then the value of M is 20. This value is set and entered by the system integrator according to the hardware capabilities during deployment. For example, set M = 20.

[0102] Calculation process:

[0103] Let: The number of people present per unit area N i for the most recent g = 6 time periods obtained are respectively: [0.3, 0.45, 0.6, 0.5, 0.4, 0.3] people / m 2 .

[0104] The current ambient light reference value E actual obtained through the previous steps = 90 lux.

[0105] Reference personnel density N ref = 0.75 people / m 2 .

[0106] Reference light intensity E ref = 200 lux.

[0107] The maximum number of brightness level grades M supported by the lighting system = 20.

[0108] First, calculate N′ i and E′ i for each time period:

[0109] Since E i = E actual = 90 lux is the same for all i, so E′ i= E actual / E ref = 90 / 200 = 0.45 is the same for all i.

[0110] N'1 = 0.3 / 0.75 = 0.4, N'2 = 0.45 / 0.75 = 0.6, N'3 = 0.6 / 0.75 = 0.8,

[0111] N'4 = 0.5 / 0.75 ≈ 0.667, N'5 = 0.4 / 0.75 ≈ 0.533, N'6 = 0.3 / 0.75 = 0.4.

[0112] Calculate the contribution terms for each time period

[0113] Denominator

[0114] Term 1: (0.4) 2 / 1.6708 = 0.16 / 1.6708 ≈ 0.09576;

[0115] Term 2: (0.6) 2 / 1.6708 = 0.36 / 1.6708 ≈ 0.21546;

[0116] Term 3: (0.8) 2 / 1.6708 = 0.64 / 1.6708 ≈ 0.38305;

[0117] Term 4: (0.667) 2 / 1.6708 ≈ 0.44489 / 1.6708 ≈ 0.26627;

[0118] Term 5: (0.533) 2 / 1.6708 ≈ 0.28409 / 1.6708 ≈ 0.16997;

[0119] Term 6: (0.4) 2 / 1.6708 = 0.16 / 1.6708 ≈ 0.09576;

[0120] Sum: 0.09576 + 0.21546 + 0.38305 + 0.26627 + 0.16997 + 0.09576 ≈ 1.22627;

[0121] Average:

[0122] Calculate L p :

[0123] L p = 0.20438 ÷ M = 0.20438 ÷ 20 ≈ 0.010219;

[0124] The result shows that the calculated percentage L of the target brightness level p is approximately 0.010219, which is about 1.02%. This value is a ratio between 0 and 1 (or a percentage after multiplying by 100), and it indicates the proportion of the brightness that the lighting system should output within its maximum controllable brightness level range based on the current occupancy density and ambient light conditions. If this value is close to 1 (or 100%), it means high-brightness lighting is required; if it is close to 0, it means lower brightness is needed. This L p value will be used in subsequent steps to determine the specific lighting fill light output.

[0125] Based on the percentage L of the target brightness level calculated in the previous paragraph p , for example, its value is 0.010219, and the current ambient light reference value E actual obtained from the previous steps (before the description of this paragraph), for example, 90 lux, the system starts to determine the actual required lighting fill light output. First, convert the percentage L of the target brightness level p to an absolute target total illuminance value L target_total by referring to a preset "ideal full-load light intensity" L ideal_max . This L ideal_max is usually set to be consistent with the parameter E ref (reference light intensity, for example, 200 lux), which represents the light level that the system should be able to provide when there is no natural light and the highest lighting support is required for human activities. Then L target_total = L p ×L ideal_max . Substituting the values, we get L target_total = 0.010219×200 lux ≈ 2.04 lux. This L target_total represents the total ambient illuminance that the ground in the platform area should reach under the current conditions. Next, the system compares this target total illuminance L target_total with the current actual existing ambient light reference value E actual (90 lux), calculates the difference in the brightness components between the two, that is, the required supplementary light amount L supplement = L target_total - E actual . Substituting the values, L supplement = 2.04 lux - 90 lux = -87.96 lux. Since the calculated required supplementary light amount is negative, it indicates that the current ambient light far exceeds the target total illuminance. Therefore, the actual required artificial lighting fill light output should be 0 lux, and the system performs a linear compensation process, that is, to ensure that the fill light output is not negative and does not exceed the maximum physical output capacity L sys_max_output (for example, 300 lux). Specifically: if Lsupplement < 0, then the final supplementary light output L final_supplement = 0. If 0 ≤ L supplement ≤ L sys_max_output , then L final_supplement = L supplement . If L supplement > L sys_max_output , then L final_supplement = L sys_max_output . In this example, L final_supplement = 0 lux. This L final_supplement value (0 lux) after linear compensation and upper and lower limit constraints constitutes the target illumination brightness parameter transmitted to the downstream.

[0126] The steps to obtain the preliminary control code sequence are as follows:

[0127] Based on the target illumination brightness parameter, match the illumination unit control protocol mapping table preset in the illumination system. According to the level interval to which the current illumination brightness parameter belongs, retrieve the illumination unit control protocol number corresponding to the level, and perform encoding structure analysis on this number to obtain the illumination unit control protocol code;

[0128] Based on the illumination unit control protocol code, extract the illumination unit address structure, instruction field format, and function bit identification method defined in the control protocol code. According to the physical layout order and response priority order of the illumination units in the platform area, sort the same protocol codes according to the address structure to generate a structured illumination unit instruction set;

[0129] Based on the structured illumination unit instruction set, merge and encode the protocol codes of all illumination units and the corresponding control fields according to the operation timing to generate a preliminary control code sequence.

[0130] Specifically, based on the specific numerical target illumination brightness parameter finally formed by the above series of steps, this parameter is, for example, 0 lux (lux), indicating that the current calculation requirement is to turn off the illumination or maintain it at the minimum illuminance. The system first accesses the "illumination unit control protocol mapping table" stored internally. This mapping table is generated by lighting engineers during the lighting system design and debugging phase according to the technical manuals and supported dimming instruction sets of various lighting units (such as different models of LED lamps, fluorescent lamps, etc.) installed on the platform and their controllers. The core structure of this table is multi-column data, including "brightness parameter level interval" (such as 0 - 5 lux, 6 - 20 lux, etc.), "corresponding brightness level" (such as level 0, level 1, etc.), "illumination unit type identifier" (such as "DALI-TypeA", "KNX-Zone1-Luminaire"), and the key "illumination unit control protocol number" (an internal code or index, such as "DALI_SET_LEVEL_0",

[0131] "KNX_DIM_VALUE_10%"), the system matches the input target lighting brightness parameter of 0 lux with the "brightness parameter level interval" in the mapping table to determine its belonging level. For example, 0 lux falls within the "0 - 5 lux" interval, corresponding to "Level 0". Subsequently, based on this "Level 0" and the type of lighting unit to be controlled (for example, currently the main lighting fixtures on the platform are DALI-Type A fixtures), the system retrieves the corresponding "lighting unit control protocol number" from the mapping table. For example, the retrieved number is "DALI_CMD_OFF_00". Then, the system analyzes the encoding structure of this "lighting unit control protocol number", which means the system will query another internal protocol definition library that details the specific communication protocol frame structure, instruction bytes, and the meaning and format of data bytes corresponding to each protocol number. For example, "DALI_CMD_OFF_00" may be parsed to correspond to the "DIRECTARCPOWER" instruction in the DALI protocol, and its data byte is 0 (indicating 0% brightness or off), thus obtaining the specific lighting unit control protocol code for target shutdown or minimum brightness.

[0132] Based on the lighting unit control protocol code parsed from the previous paragraph, such as a standardized instruction structure representing the "DIRECTARCPOWER" instruction of the DALI protocol and setting the brightness to 0, the system then extracts the detailed control parameter formats from the definition of this protocol code. Specifically, it includes the address structure of the lighting unit (for example, in the DALI protocol, the short address ranges from 0 to 63, the group address ranges from 0 to 15, the broadcast address is 255, the address occupies 1 byte, the high bit represents the addressing type, and the low bit is the address value), the instruction field format (for example, DALI instructions usually have a 1-byte instruction code, and the operation code of the "DIRECTARCPOWER" instruction is 0x00), and the relevant dimming function bit identification method (for example, some instructions may include a fade time setting bit, but for the direct turn-off instruction, there may be no such function bit or it is set to execute immediately). This information is derived from international standards (such as the IEC62386 series for DALI) or the protocol specification documents provided by device manufacturers and has been integrated into the protocol definition library during system construction. Subsequently, the system will refer to the "Deployment List of Lighting Units in the Platform Area", which is generated during lamp installation and system commissioning, and details the unique physical identifier of each lighting unit, the type of control protocol it uses, its logical address on the control bus (such as the DALI short address), the physical installation location coordinates or area description (for example, "Near Column No. P05 in Area A of the Platform, Lamp No. L101, DALI Address 5"), and the preset response priority order (for example, safety exit indicator lighting and platform edge lighting have the highest response priority of 1, the main waiting area has a priority of 2, and the auxiliary area has a priority of 3, and this order is determined by safety specifications and operational requirements). The system will filter out all lighting units that match the type of the currently obtained lighting unit control protocol code (such as all DALI lamps) and sort them in ascending order according to their logical addresses (such as DALI short addresses). If there is a response priority order, they will be grouped by priority first, and then sorted by address within the group, finally forming an ordered, structured lighting unit instruction set that includes the address of each lighting unit to be controlled and its corresponding specific control instruction (the protocol code with parsed and filled parameters).

[0133] Based on the set of structured lighting unit instructions generated in the previous paragraph, which are sorted by address and priority, and each item in the set specifies the address of the target lighting unit and the complete protocol instruction to be executed on it (for example, execute the "DIRECTARCPOWER0" instruction on the luminaire with DALI address 5), the system begins to combine and encode these independent instruction information into a preliminary control code sequence suitable for transmission on the physical communication bus. First, the system traverses each instruction in the set of structured lighting unit instructions, extracts its protocol code (such as the complete byte sequence of the DALI instruction frame) and the corresponding control fields (such as the target brightness value, switch state, etc., in this example, the brightness 0 or off state), and then the system encodes according to the predefined operation timing rules, which specify the order and time interval of instruction sending. For example, for the DALI bus, there needs to be a minimum silent time between instructions. If point-to-point control is used, an instruction frame will be generated for each address and sent sequentially. If multicast or broadcast is used, the corresponding multicast or broadcast instruction frame will be generated. The operation timing may also include preparations for the collision detection and retransmission mechanism (although the actual execution is in the sending stage). During the merging and encoding process, the system assembles the address information, instruction opcode, data parameters, etc. of each instruction into one or more complete digital signal packets according to the frame format specified by the corresponding communication protocol (such as DALI IEC62386), including start bits, address bytes, data bytes, stop bits, and possible checks such as Manchester coding rules. If multiple instructions can be merged (for example, execute the same operation on a group of luminaires), the system will preferentially use group control instructions to improve efficiency. All the encoded instruction frames for individual lighting units or groups of lighting units are concatenated in the order in which they appear in the set of structured lighting unit instructions (this order has considered the physical layout and response priority) to form a continuous binary data stream or message queue to be sent. This data stream or queue is the preliminary control code sequence.

[0134] The steps for obtaining standardized lighting control instructions are as follows:

[0135] Based on the preliminary control code sequence, parse the lighting unit control protocol field and operation instruction field in each control frame, and sequentially extract the dimming level, execution delay, and switch state coding content according to the physical location number of the lighting unit to generate a set of dimming control fields for the lighting unit;

[0136] Based on the set of dimming control fields for the lighting unit, splice each control field with the corresponding lighting unit address instruction, unify the dimming level coding format, and supplement the protocol header identifier, synchronization bit, and check bit to form a dimming instruction sequence with a complete dimming logic structure;

[0137] Based on the dimming instruction sequence with a complete dimming logic structure, perform instruction encapsulation format conversion, integrate all lighting unit dimming instructions into the same control data frame structure, and generate a standardized lighting control instruction.

[0138] Specifically, based on the preliminary control code sequence generated in the previous paragraph, which is composed of a series of instructions following the original frame format of a specific communication protocol (such as the DALI protocol), the system first parses each control frame in this sequence one by one, identifies and separates the "lighting unit control protocol field" and the "operation instruction field" inside the control frame. The former usually contains the address information of the target lighting unit (such as DALI short address, group address, or broadcast address), and the latter contains the specific operation code and related parameters (such as the "DIRECTARCPOWER" instruction in the DALI protocol and its corresponding power percentage value of 0, representing off). Then, the system relies on a pre-established and stored "mapping list of physical location and logical address of lighting units", which is generated by scanning the bus, manually entering, or importing design files during the system installation and debugging phase, and contains the unique physical location number corresponding to each logical address (such as the DALI short address "5") (for example, "P2 - A1 - L003", representing the 3rd lamp in Area A1 of Platform 2). In the order of this physical location number, the system extracts the specific control parameters of each lighting unit from the parsed operation instruction field, including the "dimming level" (for example, for the DALI protocol, it is the direct power level from 0 to 254, 0 representing the off state), the "execution delay" (for example, the preset brightness change transition time through the DALI "SETFADETIME" instruction, or the default immediate or fast change of the protocol if there is no specific setting), and the clear "switch state coding content" (for example, the dimming level 0 is uniformly interpreted as the "off" state, and non - 0 levels are the "on" state with a specific brightness). These extracted parameter information is reorganized to generate a record containing the dimming level, execution delay, and switch state for each lighting unit at a physical location, thus forming a set of dimming control fields for lighting units indexed by physical location.

[0139] A set of dimming control fields for lighting units indexed by physical location, generated based on the previous paragraph. Each record in this set clearly defines the target state of a lamp at a specific physical location (e.g., dimming level 0, execution delay "instant", switch state "off"). Next, the system reconstructs these control fields into a complete instruction frame that meets the requirements of the underlying communication protocol. First, for each record in the set, the system extracts its control fields (dimming level, execution delay, switch state) and looks up again in the "Mapping List of Lighting Unit Physical Location and Logical Address" for the lighting unit logical address instruction corresponding to this physical location number (e.g., the address byte representation of DALI short address "5"). Then, it concatenates this address instruction with the control fields. During the concatenation process, the system performs the operation of "unifying the dimming level coding format" to ensure that the dimming level value conforms to the specific coding method of the target protocol. For example, if the internally represented dimming level is a percentage and the target is the DALI protocol, it needs to be converted into a byte value within the range of 0 - 254 (in this example, level 0 directly corresponds to DALI value 0). Subsequently, the system supplements the concatenated address and data fields with other elements necessary to form a complete communication frame, specifically including: "protocol header identifier". For the DALI protocol, this is not a separate field but is identified by a specific frame start condition (a high-level start bit and subsequent bit stream timing). Then comes the "synchronization bit". The DALI signal uses Manchester coding, and each bit contains one level transition, ensuring bit synchronization. Finally, there is the "check bit". The standard DALI forward command frame (from the master to the lamp) does not contain an explicit cyclic redundancy check (CRC) or checksum field, and its data integrity mainly depends on accurate bit timing, the self-checking feature of Manchester coding, and the confirmation of the response frame. If other protocols such as Modbus are used, the corresponding CRC check code will be calculated and appended here (e.g., calculating a 16-bit CRC value using a preset polynomial for the address and data bytes). Through the above field concatenation, format unification, and supplementation of necessary elements, an independent dimming instruction frame with a complete dimming logic structure and conforming to specific communication protocol standards is generated for each lighting unit. These frames together form a dimming instruction sequence.

[0140] Based on the dimming instruction sequence formed in the previous paragraph, which consists of a series of independent dimming instruction frames that conform to a specific communication protocol (such as DALI) and have a complete dimming logic structure, the system then performs the final encapsulation format conversion and integration of the instructions. The purpose of this step is to adapt these instruction sequences for the underlying hardware to a higher-level central control system or network transmission architecture. First, the system determines the current control architecture. If the subway lighting control system adopts a hierarchical structure, for example, connected to a regional DALI gateway through an IP network, then the "instruction encapsulation format conversion" may involve encapsulating each DALI instruction frame (or a batch of instruction frames) as the data payload into an IP packet (such as a UDP or TCP packet), and adding network layer and transport layer header information such as the IP address and port number of the target gateway. If the system uses a unified internal control bus protocol that can manage multiple different lighting subsystems, then this conversion process will convert the DALI instruction sequence into the message format of this internal standard protocol. Next, "integrating all lighting unit dimming instructions into the same control data frame structure" means that if the upper-level control system or gateway supports batch command processing, the system may aggregate multiple independent DALI dimming instructions (for example, instructions sent to lamps with different addresses on the same DALI bus, or instructions sent continuously in a short period of time), and organize them into a single, structured data packet or message body according to the interface specifications of the target gateway or controller. This data packet may contain an instruction count, an instruction list (each instruction contains the target DALI address and DALI data), and possibly batch processing control information. For example, the turn-off instructions for all DALI lamps in area A of the platform are packaged into a request message sent to DALI gateway A. After this conversion and integration process, the single or batch instruction packets ready to be sent to the upper-level controller or network node are the standardized lighting control instructions.

[0141] The steps for obtaining the list of abnormal lamps are as follows:

[0142] Obtain the lamp status data of the subway lighting equipment, read the current brightness level, electrical status, working response flag, and number information of each lighting unit through the protocol port, and form a set of current status data of the lighting equipment;

[0143] Based on the set of current status data of the lighting equipment, perform item-by-item comparison according to the matching rule between the lighting unit number and the standardized lighting control instructions, determine whether the current status of each lighting unit meets the brightness level and control response logic required by the standardized lighting control instructions, filter out the lighting unit numbers with inconsistent status, mark the non-conforming type as brightness anomaly, and summarize and store them uniformly as the list of abnormal lamps.

[0144] Specifically, the system actively initiates the process of collecting the lamp status data of all installed and networked lighting devices in the current subway platform area. This process is carried out through a preset communication protocol port, such as the Ethernet port of a gateway device connected to the DALI bus (e.g., using ModbusTCP or BACnet / IP to encapsulate DALI query instructions), or directly through the PLC polling interface connected by a serial port. For each lighting unit registered in the "Deployment List of Lighting Units in the Platform Area" (this list contains information such as the unique number, logical address such as DALI short address, and physical location of each lamp, and is generated during system debugging), the system sends specific query instructions. These instructions are used to obtain its operating status according to the communication protocol followed by the lighting unit (such as the DALI standard IEC62386). The specific information read includes: "Current brightness level", for example, obtaining a value in the range of 0 - 254 through the "QUERYACTUALLEVEL" instruction of DALI, or 255 indicating that the lamp does not support or has a fault; "Electrical status", for example, checking for lamp faults through the "QUERYLAMPFAILURE" instruction of DALI, checking for driver faults through "QUERYCONTROLGEARFAILURE", and confirming whether the lamp is powered on through "QUERYLAMPPOWERON". These queries will return corresponding status codes or boolean values; "Working response flag", this flag is indirectly obtained by determining whether the device responds correctly to the query instructions according to the protocol. If no valid response is received within the preset timeout period (for example, 22 positive pulse times plus 100 ms for the DALI bus), it is considered that the response is abnormal; and the "number information" of this lighting unit, that is, its logical address on the bus, such as DALI short address 0 to 63. The system integrates a set of status data (including its logical number, reported brightness level value, electrical status code sequence, flag indicating whether the response is successful, and timestamp of data collection) collected for each queried lighting unit into a record. After all these records are summarized, they jointly form a structured set of current status data of lighting devices.

[0145] Based on the set of current status data of lighting devices formed in the previous paragraph, this set contains the measured operating parameters of each lighting unit on the platform and the standardized lighting control instructions generated in the previous steps. The instructions record the expected control status for each lighting unit (for example, the instruction requires that the brightness level of the lamp with DALI address 5 be 0, that is, turned off). Next, the system performs item-by-item comparison and status judgment. First, the system traverses each record in the set of current status data of lighting devices, and through the lighting unit number (such as DALI short address) in it, searches for the content of the latest issued instruction corresponding to this number in the standardized lighting control instructions, including the required brightness level L cmdand the expected control response logic (e.g., the luminaire should be in the on / off state or at a specific dimming level). Then, the system compares the queried current state with the instruction requirements: In terms of the brightness level, if the instruction requires the luminaire to be turned off (e.g., L cmd = 0), it checks whether the currently reported brightness level L actual is also 0, or is lower than an extremely low "effective off threshold" (e.g., DALI level 2, which is determined by the combination of the lowest physically controllable brightness and visual perceptibility of the luminaire to ensure that values below this are considered effectively off); if the instruction requires a specific brightness level L cmd > 0, it determines whether |L actual - L cmd | exceeds a preset "brightness deviation tolerance range", which is set, for example, to the larger of ±10% of the instructed brightness and a fixed minimum deviation value (e.g., DALI level 5) (i.e., max(L cmd × 0.1, 5)). This tolerance range is set considering the DALI control accuracy, the non-linearity of the luminaire dimming curve, and the small deviations that may be caused by normal aging. In terms of the control response logic, it checks whether the working response flag is "response normal" and whether there is no fault report for the electrical state. If any of the above comparison results show that the current state does not match the instruction requirements (e.g., the brightness exceeds the tolerance range, or the instruction is to turn off but the luminaire is still on, or the luminaire has no response), the number of the lighting unit is screened out, and its non-conformance type is initially marked as "brightness anomaly" (if the main problem lies in the mismatch of the brightness value) or other corresponding fault types. After summarizing all the screened lighting unit numbers with inconsistent states, their non-conformance types, instruction required states, actual states, etc., they are uniformly stored as a structured abnormal luminaire list.

[0146] The steps to obtain the specified luminaire maintenance activation signal are as follows:

[0147] Based on the abnormal luminaire list, retrieve the fault type, cumulative fault duration, number of failed responses, and state switching frequency of each lighting unit, extract the number of each luminaire and the corresponding state parameters, and generate a list of abnormal state data corresponding to each luminaire number;

[0148] According to the list of abnormal state data corresponding to each luminaire number, calculate the maintenance trigger factor for each luminaire. The calculation formula is:

[0149]

[0150] where, F r,j is the maintenance trigger factor for the j-th faulty luminaire, CFT j is the cumulative fault duration of this luminaire in seconds, R jThe number of failure responses of the luminaire, in number of times, S j is the frequency of state switching of the luminaire, in number of times, ln(S j +2) ensures positive correlation with the growth of the frequency and no risk of division by zero. Z is a constant normalization coefficient, in seconds;

[0151] Based on the maintenance trigger factor, judge whether each luminaire exceeds the maintenance trigger threshold one by one, screen the luminaire numbers that meet the conditions, and generate a maintenance activation signal for the specified luminaire.

[0152] Specifically, based on the list of abnormal luminaires generated in the previous paragraph, which lists the lighting units whose current status does not match the instruction requirements and their preliminarily marked abnormal types, the system first traverses each lighting unit entry in this list. For the lighting unit number recorded in each entry (such as the DALI short address "5"), retrieve it from the system's historical status and event log database to collect more detailed fault-related parameters of the luminaire. The specific data items retrieved and extracted include: "fault type", which is directly obtained from the list of abnormal luminaires, such as "abnormal brightness", "response timeout", "driver failure", etc.; "cumulative fault duration", by querying the start timestamp and recovery timestamp (or the current time if the fault is still ongoing) when the luminaire entered the fault state each time, calculate the duration of each fault, and accumulate these durations to get a total number of seconds; "number of failure responses", by counting the number of times the luminaire fails to correctly respond to control instructions or status queries within a preset evaluation period (such as the past 30 days); and "frequency of state switching", by counting the specific number of changes in the operating status reported by the luminaire (such as the brightness level changing from X to Y, or from normal to fault) within the same evaluation period (such as the past 30 days). For example, the luminaire "DALI-05" is marked as "abnormal brightness" in the list of abnormal luminaires. The system retrieves from the log that its current "abnormal brightness" has lasted for 7200 seconds (2 hours), the historical cumulative fault (all types) duration is 86400 seconds (24 hours), the number of failure responses in the past 30 days is 3 times, and the state switching (including brightness changes and fault reports) has occurred 15 times in total. The system associates these retrieved and calculated parameters (fault type, cumulative fault duration, number of failure responses, frequency of state switching) with the luminaire number to form a detailed abnormal status record. After all these records are collected, an abnormal status data list corresponding to each luminaire number is generated.

[0153] Formula: The advantage of this formula is that it comprehensively combines multiple dynamic parameters related to the severity of luminaire faults and the urgency of maintenance. By taking the cumulative fault duration (CFT j ), the number of failure responses (R j ) and the frequency of state switching (Sj ) The combination of the three in the form of product and logarithmic function can more comprehensively evaluate the health status and potential risks of the luminaire, R j +1 and ln(S j +2) are designed to ensure that even when the number of responses or switching frequencies is 0, the factor is still meaningful and will not cause the entire factor to fail or result in calculation errors due to zero values. At the same time, the logarithmic function smooths the extreme impact of the state switching frequency, making the growth of the maintenance trigger factor more reasonable. The application of the constant normalization coefficient Z enables the factor values under different parameter combinations to be standardized to a comparable scale, facilitating the setting of a unified maintenance trigger threshold;

[0154] CFT j is the cumulative failure duration of the j-th faulty luminaire, in seconds. The acquisition steps are as follows: The system continuously monitors the status of each lighting unit. When it detects that luminaire j enters any predefined fault state (e.g., abnormal brightness, no response, driver failure, etc.), it records the current timestamp as the fault start time T start , and when the luminaire returns to the normal state or is manually confirmed to be repaired, it records the recovery timestamp T end , and the single-fault duration is T end -T start . The system stores the durations of all historical fault events of the luminaire since it was put into use and accumulates these single-fault durations to obtain CFT j . For example, luminaire numbered "P2-A1-L003" has had three faults in history, with durations of 3600 seconds, 18000 seconds, and 7200 seconds that is still ongoing (e.g., 7200 seconds have passed since the start of this fault at the current checkpoint), then its CFT j = 3600 + 18000 + 7200 = 28800 seconds.

[0155] R j is the number of failed responses of the j-th faulty luminaire, in times. The acquisition steps are as follows: After the system sends a control instruction (such as dimming, switching) or a status query instruction to lighting unit j, it expects to receive a valid response signal within a preset timeout period (e.g., the general response time upper limit specified by the DALI protocol is about 100 milliseconds). If no response is received within the timeout period, or the received response format is incorrect or the content does not meet the expectations, it is counted as one failed response. The system maintains a failed response counter R j for each luminaire within a specific evaluation period (e.g., since the last maintenance reset or within the past 30-day rolling period). Each time a failed response occurs, this counter is incremented by 1. For example, luminaire "P2-A1-L003" has had 5 control instructions without receiving valid responses in the past 30 days, then its R j = 5 times.

[0156] S j is the status switching frequency of the j-th faulty lamp, with the unit of times. The acquisition steps are as follows: The system records the number of times the key operating status of lighting unit j changes within a specific evaluation period (for example, within the past 7-day rolling period). These statuses include but are not limited to: non-instructional jumps in the reported brightness level, changes in the working mode (such as normal mode, emergency mode), and transitions of the electrical status (such as bulb failure, driver failure flag bit) from 0 to 1 or from 1 to 0. Each such valid status change is counted as one status switch. For example, for the lamp "P2-A1-L003" within the past 7 days, if its reported brightness level has 3 uncommanded fluctuations and reports 1 instantaneous driver failure and then recovers, then its S j = 3 + 1 = 4 times.

[0157] Z is a constant normalization coefficient, with the unit of seconds. The acquisition steps are as follows: The setting of Z aims to adjust the calculation result of the maintenance trigger factor F r,j to a numerical range that is convenient for understanding and application (such as 0 - 1000), and to coordinate it with the setting of the maintenance trigger threshold. Its value is obtained through statistical analysis and calibration of historical data. The specific steps are as follows: 1. Collect a large number of lamp maintenance cases that have occurred and been processed, and record the original value N of the numerator term calculated for each lamp before maintenance is triggered in each case j = CFT j ·(R j + 1)·ln(S j + 2), 2. Set a desired target mean or median of the maintenance trigger factor F target (for example, set to 100). This target value represents a typical factor level that should trigger maintenance. 3. Calculate Z = Average(N j ) / F target or Z = Median(N j ) / F target as the initial value. For example, if the average value of N j in historical cases is 6×10 7 s and the desired F target is 150, then Z = (6×10 7 s) / 150 = 400,000 s. 4. In actual application, fine-tune the Z value according to the distribution of F r,j and maintenance feedback to optimize the accuracy and sensitivity of maintenance warnings.

[0158] Calculation process:

[0159] Let the status parameters of the j-th lamp "P2-A1-L003" be as follows:

[0160] Cumulative Failure Duration CFT j = 28,800 seconds.

[0161] Number of Failure Responses R j = 5 times.

[0162] State Transition Frequency S j = 4 times.

[0163] Constant Normalization Coefficient Z = 400,000 seconds.

[0164] Calculate (R j + 1):

[0165] R j + 1 = 5 + 1 = 6;

[0166] Calculate ln(S j + 2):

[0167] ln(S j + 2) = ln(4 + 2) = ln(6) ≈ 1.79176;

[0168] Calculate the numerator CFT j ·(R j + 1)·ln(S j + 2):

[0169] Numerator = 28800s·6·1.79176 ≈ 172800s·1.79176 ≈ 309583.488s;

[0170] Calculate the Maintenance Trigger Factor F r,j :

[0171]

[0172] The result shows that: The maintenance trigger factor F of the lamp "P2 - A1 - L003" r,j The calculated value is 0.77396, which is a quantitative indicator comprehensively evaluating the current failure history, response stability, and state volatility of the lamp. This value itself is a relative value (in this example, through the setting of Z, its value is usually not very large), and its magnitude directly reflects the urgency of maintenance or the potential risk level of the lamp. If the calculated F r,j value exceeds the preset maintenance trigger threshold, it indicates that the health status of the lamp has reached the level that requires manual inspection or preventive maintenance. The higher the value, the higher the maintenance priority usually is.

[0173] Based on the maintenance trigger factor F calculated for each abnormal lamp in the previous paragraph r,j, for example, the F value of the lamp "P2-A1-L003" is 0.77396. Next, the system compares this factor with a preset "maintenance trigger threshold" F. r,j The "maintenance trigger threshold" is comprehensively set by the maintenance management department according to historical maintenance data, the mean time between failures (MTBF) of lamps, maintenance resource limitations, and the desired equipment reliability goals during system configuration. Its setting process generally includes: analyzing the F value distribution of lamps that had serious failures or required emergency repairs in history before the failure, and selecting a boundary that can effectively distinguish potentially high-risk lamps. For example, through statistics on past data, it is found that when the F value (for example, Z, where the F value is generally in the range of 0 - 10) is greater than 3.5, the probability of complete failure of the lamp within the next month increases significantly, and this threshold can cover 85% of such events at an acceptable false alarm rate. Then, F can be set to 3.5. The system executes this judgment logic for each lamp in the "abnormal status data list corresponding to each lamp number": compares the calculated F value of the lamp with F (for example, 3.5). If F (for example, if the F value of a certain lamp is 4.2, then the condition 4.2 > 3.5 holds), then the lamp is determined to meet the maintenance trigger condition. The system records the unique number of this lamp (for example, "P2-A1-L003"). After completing this judgment for all lamps, the numbers of all lamps that meet the conditions are summarized in a list. Based on this list, the system generates a specified lamp maintenance activation signal, which may be a work order creation request sent to the maintenance management system (CMMS), or highlighting and alarming these lamps on the monitoring interface, along with their F values and relevant fault parameters for maintenance personnel to reference and schedule. Thresh The "maintenance trigger threshold" is comprehensively set by the maintenance management department according to historical maintenance data, the mean time between failures (MTBF) of lamps, maintenance resource limitations, and the desired equipment reliability goals during system configuration. Its setting process generally includes: analyzing the F value distribution of lamps that had serious failures or required emergency repairs in history before the failure, and selecting a boundary that can effectively distinguish potentially high-risk lamps. r,j For example, through statistics on past data, it is found that when the F value (for example, Z, where the F value is generally in the range of 0 - 10) is greater than 3.5, the probability of complete failure of the lamp within the next month increases significantly, and this threshold can cover 85% of such events at an acceptable false alarm rate. Then, F can be set to 3.5. r,j value (such as Z such that the F value is generally in the range of 0 - 10) is greater than 3.5, the probability of complete failure of the lamp within the next month increases significantly, and this threshold can cover 85% of such events at an acceptable false alarm rate. Then, F can be set to 3.5. r,j is generally in the range of 0 - 10) is greater than 3.5, the probability of complete failure of the lamp within the next month increases significantly, and this threshold can cover 85% of such events at an acceptable false alarm rate. Then, F can be set to 3.5. Thresh is set to 3.5. The system executes this judgment logic for each lamp in the "abnormal status data list corresponding to each lamp number": compares the calculated F value of the lamp with F. r,j with F. Thresh (for example, 3.5). If F. r,j >F. Thresh (for example, if the F value of a certain lamp is 4.2, then the condition 4.2 > 3.5 holds), then the lamp is determined to meet the maintenance trigger condition. The system records the unique number of this lamp (for example, "P2-A1-L003"). r,j = 4.2, then the condition 4.2 > 3.5 holds), then the lamp is determined to meet the maintenance trigger condition. The system records the unique number of this lamp (for example, "P2-A1-L003"). After completing this judgment for all lamps, the numbers of all lamps that meet the conditions are summarized in a list. Based on this list, the system generates a specified lamp maintenance activation signal, which may be a work order creation request sent to the maintenance management system (CMMS), or highlighting and alarming these lamps on the monitoring interface, along with their F values and relevant fault parameters for maintenance personnel to reference and schedule. r,j value and relevant fault parameters for maintenance personnel to reference and schedule.

[0174] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A subway intelligent lighting energy-saving control method, characterized in that The following steps are involved: Based on the preloaded train route and real-time train entry detection signal, the train approaching status in the platform area is judged. Combined with the passenger flow, the lighting activation start time and duration of the platform area are set to generate the platform activation time window; Based on the platform activation time window, read the ambient light sensor value of the platform area, establish a current ambient light reference value, compare the current ambient light reference value with the number of people, determine the target brightness level percentage, and obtain the target lighting brightness parameter; Based on the target lighting brightness parameter, a lighting unit control protocol code is selected to establish a preliminary control code sequence, and based on the preliminary control code sequence, a complete dimming instruction for the lighting unit is combined to generate a standardized lighting control instruction; Obtain the lamp status data of the subway lighting equipment, identify the lighting unit numbers that do not match the standardized lighting control instruction status, establish a list of abnormal lamps, and based on the abnormal lamp list, match the faulty lamp numbers in the list with preset maintenance trigger conditions to generate a designated lamp maintenance activation signal.

2. The subway intelligent lighting energy-saving control method according to claim 1, characterized in that The steps for obtaining the station activation time window are: According to the train route, the track number and train direction of the target platform are retrieved, the train entry timestamp in the current scheduling cycle is extracted from the real-time train entry detection signal, and the corresponding departure time stamp of the train is found from the scheduling plan, and the difference between the entry and departure is calculated to obtain the planned stay time of the train in the platform section; According to the planned stop time of the train in the platform section, collect the personnel flow per unit area and the personnel flow velocity data of the corresponding channel in each sampling unit time period in the platform area during this period, normalize the flow value with the standard reference density, normalize the flow velocity value with the reference flow velocity value, and calculate the lighting duration; The train entry detection time is set as the lighting activation start time point, and the lighting activation end time point is set according to the lighting duration to generate a platform activation time window.

3. The subway intelligent lighting energy-saving control method according to claim 1, characterized in that, The steps for obtaining the current ambient light reference value are: Based on the platform activation time window, locate the start time and end time boundaries of the corresponding activation period in the current platform area, extract the original numerical sequence of illuminance in the activation time period, and generate a set of original values ​​of ambient illuminance in the platform area; Based on the original value set of ambient light illuminance in the platform area, the original value sequence of illuminance is classified and aggregated according to the sensor number, and the sampling period consistency check is performed on each type of original value sequence of illuminance, and the data segments with discontinuity, lag or drift are eliminated to obtain the effective value set of ambient light illuminance in the platform area; Based on the effective value set of ambient light illumination in the platform area, priority screening is performed according to the spatial arrangement position and stability of each group of sensors, and the illumination value representing the overall brightness level of the current platform is extracted to generate the current ambient light reference value.

4. The subway intelligent lighting energy-saving control method according to claim 1, characterized in that The steps for obtaining the target lighting brightness parameters are: Based on the current environmental light reference value, retrieve all the turnstile statistical data in the platform area at the current time point. Accumulate all the entry and exit records according to the lighting control time segments, count the total number of people present in each time segment, and divide the total number of people in each segment by the effective usable area of the platform area to obtain the sequence of the number of people present per unit area corresponding to each time segment; According to the sequence of the number of people present per unit area and the current environmental light reference value, perform normalization processing on each pair of data items in the order of time segment indexing, and calculate the target brightness level percentage; Based on the target brightness level percentage, calculate the difference in brightness components with the current environmental light reference value, and obtain the required lighting supplementary light output through linear compensation to form the target lighting brightness parameters.

5. The subway intelligent lighting energy-saving control method according to claim 1, characterized in that The steps for obtaining the preliminary control code sequence are as follows: Based on the target lighting brightness parameters, match the lighting unit control protocol mapping table preset in the lighting system. According to the level interval to which the current lighting brightness parameters belong, retrieve the lighting unit control protocol number corresponding to the level, and perform encoding structure analysis on this number to obtain the lighting unit control protocol code; Based on the lighting unit control protocol code, extract the lighting unit address structure, instruction field format, and function bit identification method defined in the control protocol code. According to the physical layout order and response priority order of the lighting units in the platform area, sort the same protocol codes according to the address structure to generate a structured lighting unit instruction set; Based on the structured lighting unit instruction set, merge and encode the protocol codes and corresponding control fields of all lighting units according to the operation timing to generate a preliminary control code sequence.

6. The subway intelligent lighting energy-saving control method according to claim 1, characterized in that The steps for obtaining the standardized lighting control instruction are as follows: Based on the preliminary control code sequence, analyze the lighting unit control protocol field and operation instruction field in each control frame, and sequentially extract the dimming level, execution delay, and switch status coding content according to the physical position number of the lighting unit to generate a dimming control field set for the lighting unit; Based on the dimming control field set for the lighting unit, splice each control field with the corresponding lighting unit address instruction, unify the dimming level coding format, and supplement the protocol header identifier, synchronization bit, and check bit to form a dimming instruction sequence with a complete dimming logic structure; Based on the dimming instruction sequence with a complete dimming logic structure, perform instruction encapsulation format conversion, and integrate all lighting unit dimming instructions into the same control data frame structure to generate a standardized lighting control instruction.

7. The subway intelligent lighting energy-saving control method according to claim 1, characterized in that The steps for obtaining the list of abnormal lamps are as follows: Obtain the lamp status data of the subway lighting equipment, and read the current brightness level, electrical state, working response flag, and number information of each lighting unit through the protocol port to form a set of current status data of the lighting equipment; Based on the current state data set of the lighting device, compare item by item according to the matching rule between the lighting unit number and the standardized lighting control instruction, judge whether the current state of each lighting unit meets the brightness level and control response logic required by the standardized lighting control instruction, screen the lighting unit numbers with inconsistent states, mark the non-conformity type as brightness anomaly, and summarize and store them uniformly as the list of abnormal lamps.

8. The subway intelligent lighting energy-saving control method according to claim 1, characterized in that The steps for obtaining the specified lamp maintenance activation signal are as follows: Based on the list of abnormal lamps, retrieve the fault type, cumulative fault duration, failure response times and state switching frequency of each lighting unit, extract the number of each lamp and the corresponding state parameters, and generate a list of abnormal state data corresponding to each lamp number. According to the list of abnormal state data corresponding to each lamp number, calculate the maintenance trigger factor for each lamp. Based on the maintenance trigger factor, judge one by one for all lamps whether it exceeds the maintenance trigger threshold, screen the lamp numbers that meet all conditions, and generate a specified lamp maintenance activation signal.

Images