Road tunnel vehicle-mounted lighting system based on interval intelligent measurement of lamp light-up time

Abstract

The application discloses a highway tunnel vehicle-mounted lighting system based on interval intelligent measurement of lamp light-up time, and relates to the technical field of highway tunnel lighting energy saving. The highway tunnel vehicle-mounted lighting system based on interval intelligent measurement of lamp light-up time comprises an interval division module, a lighting analysis module and a lighting control module. The application can accurately predict the lighting path of each vehicle by performing fine interval division on the highway tunnel, dynamically obtaining vehicle passing key features in combination with vehicle shielding features, identification information and time stamps, can realize vehicle response and on-demand lighting, can fuse front and rear external environment light change information, can extract sunlight reflection lighting features, can eliminate strong sunlight reflection, entrance and exit light difference and other interference on lighting control from the source, and can significantly improve the environmental adaptability and identification accuracy of the lighting system.

Description

Technical Field

[0001] This invention relates to the field of energy-saving technology for highway tunnel lighting, and in particular to a vehicle-mounted lighting system for highway tunnels based on interval intelligent calculation of lamp adjustment time. Background Technology

[0002] Existing highway tunnel lighting systems mostly use fixed-duration or sensor-triggered control methods to control the lights, which cannot accurately adjust according to actual vehicle traffic conditions and changes in ambient light. This easily leads to excessive lighting energy consumption or insufficient local lighting, affecting traffic safety. Especially in special scenarios such as strong sunlight reflection, misjudgment of following distance, and abnormal deceleration, traditional systems struggle to respond in time, resulting in blind spots and control delays. Furthermore, existing systems typically lack the ability to dynamically integrate predicted vehicle paths with natural light interference, and their lighting control strategies are outdated, making it difficult to meet the needs of high-speed, dense, or special vehicle traffic.

[0003] Therefore, it is necessary to consider an intelligent lighting control method based on the linkage between vehicle traffic characteristics and changes in external lighting to achieve precise vehicle-mounted lighting and dynamic strategy updates within the designated area. Summary of the Invention

[0004] The present invention aims to provide a highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time, so as to realize precise interval vehicle lighting and dynamic strategy update.

[0005] A highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time includes:

[0006] The interval division module is used to define the number of tunnel lighting intervals contained in the current highway tunnel; and to divide the number of tunnel lighting intervals into 5 tunnel segments D according to their location. n n = 1, 2, ..., 5;

[0007] The lighting analysis module includes an information acquisition unit and a lighting analysis unit. The information acquisition unit acquires real-time vehicle identification information, vehicle passage timestamps, and vehicle occlusion features in tunnel section D1. Based on the vehicle identification information, vehicle passage timestamps, and vehicle occlusion features, it analyzes the data to obtain key vehicle traffic features and predict vehicle lighting paths. Based on the predicted vehicle lighting paths, it generates a predicted traffic lighting demand Q. n Simultaneously, acquire external ambient light data and output sunlight reflection illumination characteristics based on the external ambient light data;

[0008] The lighting analysis unit is used for tunnel section D n Establish a lighting demand queue Z n Lighting demand queue Z n It contains all currently received predicted traffic lighting requirements; based on the lighting demand queue Z nThe output of sunlight reflection lighting characteristics and tunnel section lighting control strategy P n ;

[0009] The lighting control module is used to continuously update the lighting demand queue Z. n and tunnel section lighting control strategy P n The lighting fixtures in the current highway tunnels are controlled according to the tunnel lighting zones.

[0010] As a preferred technical solution of the present invention, the specific steps for analysis based on vehicle identification information, vehicle passage timestamps, and vehicle occlusion features include:

[0011] Based on vehicle occlusion features, occlusion waveform analysis is performed to extract occlusion duration, occlusion intensity variation curves, and occlusion integrity indicators; a pre-trained vehicle feature recognition model is used for analysis to output vehicle passage labels.

[0012] Vehicle category labels are obtained by identifying special tags based on vehicle identification information.

[0013] The initial speed of the vehicle is calculated based on the vehicle's time stamp; the predicted vehicle lighting path is then calculated based on the initial speed of the vehicle and the current highway tunnel.

[0014] By combining vehicle access tags and vehicle category tags, key characteristics of vehicle access can be obtained.

[0015] As a preferred embodiment of the present invention, the specific steps for predicting the vehicle lighting path based on the initial speed of the vehicle and the current highway tunnel are as follows:

[0016] Based on historical vehicle traffic data, for tunnel section D n Establish tunnel traffic impact factor Y n Let the initial speed of the vehicle be V1; in tunnel section D n In the middle, the predicted average vehicle speed V is corrected. n , where V n =V n-1 *Y n ;

[0017] Based on the predicted average vehicle speed V n and tunnel section D n Tunnel section length L n Predicted in tunnel section D n Vehicle entry time T n1 and vehicle departure time T n2 ; Traverse all tunnel segments D n The combined data yields the predicted vehicle lighting path.

[0018] As a preferred technical solution of the present invention, a predicted traffic lighting demand Q is generated based on the predicted vehicle lighting path. n The specific steps include:

[0019] For tunnel section D n Construct a standard lighting response template R n ;

[0020] Matching vehicle traffic lighting demand factors based on key vehicle traffic features.

[0021] Based on the vehicle traffic lighting demand factor, the standard lighting response template R is applied. n Lighting corrections are performed to obtain the predicted traffic lighting demand Q. n .

[0022] As a preferred embodiment of the present invention, the specific steps for outputting sunlight reflection illumination characteristics based on external ambient light data include:

[0023] External ambient lighting data includes both front and back external ambient lighting data.

[0024] Both the pre-external ambient lighting data and the post-external ambient lighting data include ambient lighting data, reflected light data, and illumination intensity data; a pre-external ambient lighting change sequence is constructed based on the pre-external ambient lighting data; a post-external ambient lighting change sequence is constructed based on the post-external ambient lighting data.

[0025] The rate of change of illumination gradient is calculated based on the previous external ambient illumination change sequence; the synchronization of ambient illumination data, reflected light data and illumination intensity data in the previous external ambient illumination data is judged to obtain illumination synchronization characteristics; the spectral time difference is extracted based on the previous external ambient illumination change sequence; the previous solar reflected illumination characteristics are obtained based on the illumination synchronization characteristics and the spectral time difference.

[0026] Similarly, the characteristics of post-sunlight reflection illumination were obtained by analyzing the post-external ambient light change sequence;

[0027] The characteristics of the front and rear sunlight reflection illumination are combined to obtain the characteristics of sunlight reflection illumination.

[0028] As a preferred technical solution of the present invention, based on the lighting demand queue Z n The output of sunlight reflection lighting characteristics and tunnel section lighting control strategy P n The specific steps include:

[0029] Based on lighting demand queue Z n All predicted traffic lighting requirements received internally are pooled to obtain the initial control strategy C for the tunnel segment. n ;

[0030] Based on the characteristics of front and rear sunlight reflection illumination and the swarm optimization algorithm, the initial control strategy C for the tunnel section is proposed. n After optimization and adjustment, the tunnel section lighting control strategy P was obtained. n .

[0031] In a preferred embodiment of the present invention, the swarm optimization algorithm is a tuna swarm optimization algorithm.

[0032] The present invention has the following advantages:

[0033] 1. This invention, by finely dividing highway tunnels into sections and combining vehicle occlusion features, identification information, and timestamps, dynamically acquires key characteristics of vehicle passage, enabling accurate prediction of the lighting path for each vehicle and achieving vehicle-responsive, on-demand lighting. By integrating information on changes in ambient light before and after the vehicle, it extracts sunlight reflection lighting features, eliminating interference from strong sunlight reflection and entrance / exit light differences on lighting control at the source, significantly improving the environmental adaptability and recognition accuracy of the lighting system.

[0034] 2. This invention constructs a lighting demand queue and performs pooling fusion, enabling unified scheduling of lighting demands in multi-vehicle traffic scenarios, avoiding lighting gaps and resource waste. With the help of a swarm optimization algorithm, it further optimizes lighting control strategies intelligently while ensuring traffic safety and visual continuity, achieving multi-dimensional optimal configuration of lighting areas, brightness levels, and durations. Overall, it has comprehensive advantages of high responsiveness, high energy efficiency, and high adaptability, effectively improving tunnel traffic safety and energy efficiency management. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the structure of a highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time, used in an embodiment of the present invention. Detailed Implementation

[0036] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this invention.

[0037] A highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time, see [reference]. Figure 1 As shown, it includes:

[0038] The interval division module is used to define the number of tunnel lighting intervals contained in the current highway tunnel; and to divide the number of tunnel lighting intervals into 5 tunnel segments D according to their location. n n = 1, 2, ..., 5; each of the several tunnel lighting sections contains LED tunnel light groups for illumination;

[0039] The tunnel section D1 is the pre-entry zone, used for vehicle preparation and pre-judgment before entry. It may appear in areas with strong sunlight, creating a contrast between light and dark at the tunnel entrance. Drivers' vision has not yet adapted to the darkness of the tunnel, so they need to be aware of this in advance. The entrance should not be suddenly bright, and natural light reflection should be avoided to prevent interference with sensor triggering. Tunnel section D2 is the entrance adaptation zone, where the driver's vision adapts from light to dark. High-brightness lighting should be used to simulate the bright environment outside the tunnel and alleviate the visual gap. Tunnel section D3 is the middle zone of the tunnel, which is the main passage area during driving. Sufficient visibility and visual clarity should be maintained. Tunnel section D4 is the exit adaptation zone, where the vision transitions from dark to bright. The driver's vision has adapted to the dark environment inside the tunnel, and suddenly entering strong sunlight will cause glare. A "gradual darkening zone" should be set up to transition outward to avoid abrupt changes in vision. Tunnel section D5 is the exit zone, which is a short adjustment and light release zone after leaving the tunnel. The lighting should be maintained at a medium to low brightness for a few seconds to avoid sudden loss of lighting in the dark.

[0040] The lighting analysis module includes an information acquisition unit and a lighting analysis unit. The information acquisition unit acquires real-time vehicle identification information, vehicle passage timestamps, and vehicle occlusion features in tunnel section D1. Based on the vehicle identification information, vehicle passage timestamps, and vehicle occlusion features, it analyzes the data to obtain key vehicle traffic features and predict vehicle lighting paths. Based on the predicted vehicle lighting paths, it generates a predicted traffic lighting demand Q. n Simultaneously, acquire external ambient light data and output sunlight reflection illumination characteristics based on the external ambient light data;

[0041] An infrared vehicle detector and camera with occlusion detection are installed at the tunnel entrance to acquire vehicle identification information, vehicle passage timestamps, and vehicle occlusion features. The vehicle identification information is used to determine whether the vehicle is a special vehicle and its model. The vehicle passage timestamp is marked at the moment when vehicle occlusion occurs to calculate the initial speed of the vehicle, which serves as the zero point for all lighting prediction paths and lighting time windows. It is used to determine the time difference between vehicles ahead and behind and to identify the following vehicle or convoy structure. The vehicle occlusion features are used to identify the duration of occlusion, waveform integrity, whether there are abrupt changes or multiple peaks, and can also distinguish whether it is a false trigger by other interfering objects such as billboards.

[0042] The specific steps for analysis based on vehicle identification information, vehicle passage timestamps, and vehicle occlusion features include:

[0043] Based on vehicle occlusion features, occlusion waveform analysis is performed to extract occlusion duration, occlusion intensity variation curves, and occlusion integrity indicators; a pre-trained vehicle feature recognition model is used for analysis to output vehicle passage labels.

[0044] In occlusion waveform analysis, the occlusion duration represents the total time that the vehicle occludes the infrared signal within the sensor's field of view, reflecting the vehicle's length or speed. The occlusion intensity variation curve describes the trend of infrared signal intensity changing over time during the occlusion process, which can be used to determine the vehicle's outline, driving posture, and whether there is any abnormal interference. The occlusion integrity index comprehensively measures the continuity, smoothness, and morphological integrity of the waveform, used to determine whether the occlusion event is a real and stable vehicle passage, or whether there are any abnormalities such as jumps, discontinuities, or artifacts.

[0045] The training process of the vehicle feature recognition model includes: first, collecting infrared sensor data covering various vehicle types, traffic states, and occlusion waveforms, and establishing a fully labeled training set by combining image data; then, selecting a suitable basic model for time series or image processing (such as 1D-CNN or ResNet), setting classification and validity recognition as training objectives, and optimizing using cross-entropy or focus loss functions; during training, improving generalization ability by adjusting the learning rate, introducing data augmentation, and early stopping strategies; once the validation set accuracy and occlusion validity recognition accuracy reach the set standards, the model training is completed and used for subsequent vehicle passage label determination and false trigger recognition; in the vehicle feature recognition model, the vehicle passage label recognition process is based on occlusion waveform features. For example, if the occlusion duration exceeds a set threshold and the waveform contour is continuous and stable, it is determined to be a valid vehicle; if the occlusion time is too short, the waveform shows jump interference, or cannot match a valid vehicle ID, it is determined to be a false trigger, and subsequent analysis is terminated; if there are two obvious peaks in the waveform, and they match the vehicle length and time interval characteristics, they are recorded as suspected following structures for subsequent lighting compensation judgment.

[0046] Special tag identification is performed based on vehicle identification information to obtain vehicle category tags; the vehicle identification information is matched with a preset vehicle feature database to extract the vehicle's registration type, appearance features and access permissions; based on the matching results, it is determined whether the vehicle belongs to a special vehicle category (such as ambulance, fire truck, hazardous chemical transport vehicle, etc.) or a large or medium-sized vehicle (such as truck, bus), and a corresponding vehicle category tag is generated accordingly, which serves as an important basis for subsequent lighting dispatch priority and response strategy adjustment.

[0047] The pre-built vehicle feature database is constructed through multi-source data collection and classification. It mainly includes vehicle registration information from traffic management systems, vehicle type archives from highway toll systems, historical passage records, and special vehicle filing data. All of the above data sources are publicly available. The database establishes a unique identifier for each vehicle and associates it with its vehicle category (such as passenger car, truck, special vehicle), external parameters (such as length and height), passage priority, and special passage rules. To improve real-time matching efficiency, the database is updated and cleaned regularly, and supports configuration by region and dynamic loading to ensure the accuracy and timeliness of tag recognition.

[0048] The initial speed of the vehicle is calculated based on the vehicle's time stamp; the predicted vehicle lighting path is then calculated based on the initial speed of the vehicle and the current highway tunnel.

[0049] When a vehicle enters tunnel section D1, the initial obstruction timestamp is recorded. The speed is then measured by combining the obstruction duration with the vehicle length to obtain the initial vehicle speed. The vehicle passage label and vehicle category label are combined to obtain the key characteristics of vehicle passage.

[0050] Based on the initial speed of the vehicle and the current highway tunnel, the specific steps for predicting the vehicle lighting path are calculated, including:

[0051] Based on historical vehicle traffic data, for tunnel section D n Establish tunnel traffic impact factor Y n Let the initial speed of the vehicle be V1; in tunnel section D n In the middle, the predicted average vehicle speed V is corrected. n , where V n =V n-1 *Y n ;

[0052] Based on the predicted average vehicle speed V n and tunnel section D n Tunnel section length L n Predicted in tunnel section D n Vehicle entry time T n1 and vehicle departure time T n2 ; Traverse all tunnel segments D n The predicted vehicle lighting path is obtained by combining the results.

[0053] Tunnel Section D n Tunnel traffic impact factor Y nIt is a dynamically corrected parameter constructed based on historical vehicle traffic data statistics, used to reflect the typical impact of a specific tunnel section on vehicle speed. Its construction process includes: collecting a large amount of data on the actual and initial speeds of vehicles in each tunnel section; calculating the speed reduction or acceleration ratio through regression analysis or moving average; and simultaneously weighting and correcting the parameters by combining the physical characteristics of the tunnel section (such as slope, curves, and lighting conditions), traffic environment (such as traffic density and accident frequency), and control factors (such as speed limit signs and surveillance coverage). Finally, Y... n As a factor reflecting the trend of average traffic speed variation in this section, it is used to correct the speed of the previous section to estimate the speed of this section, thereby improving the accuracy and adaptability of lighting path prediction. For example, analysis of historical vehicle traffic data shows that vehicles exhibit typical speed change trends in different tunnel sections: in the pre-tunnel entry zone (D1), due to strong lighting and the black hole effect, drivers often experience slight deceleration or speed fluctuations; after entering the entrance adaptation zone (D2), due to the transition from bright to dark vision, the vehicle speed drops significantly, reaching the lowest speed point in the entire tunnel; after entering the middle tunnel zone (D3), as visual adaptation is completed, the vehicle speed gradually increases and tends to stabilize; in the exit adaptation zone (D4), affected by the white hole effect, some vehicles will again experience slight deceleration or speed fluctuations; finally, in the post-tunnel exit zone (D5), the driver's field of vision is wide and the environment becomes clear, and the vehicle speed generally returns to normal or accelerates, showing an accelerating traffic trend.

[0054] The predicted vehicle lighting path is calculated by traversing each tunnel segment within the tunnel, based on the predicted average speed of the vehicle in each segment and the length of the tunnel segment, to determine the vehicle's entry and exit times in that segment, and then combining the lighting time windows of each segment into a continuous lighting interval. Whether each tunnel segment is included in the predicted lighting path depends on the expected coverage period of the vehicle in that segment, thereby determining when the lighting should be turned on during that period. The path range is also adjusted in real time based on the actual traffic conditions of the vehicles, thus ensuring that the lighting is synchronized with the vehicles throughout the entire passage, meeting both safety vision requirements and achieving dynamic energy-saving control.

[0055] Predicted traffic lighting demand Q is generated based on predicted vehicle lighting paths. n The specific steps include:

[0056] For tunnel section D n Construct a standard lighting response template R n ;

[0057] Matching vehicle traffic lighting demand factors based on key vehicle traffic features.

[0058] Based on the vehicle traffic lighting demand factor, the standard lighting response template R is applied. n Lighting corrections are performed to obtain the predicted traffic lighting demand Q. n ;

[0059] The standard lighting response template is a set of basic lighting control parameters preset for each tunnel section, representing the default lighting strategy under typical vehicle traffic conditions. It is constructed based on the physical function of the tunnel section, visual adaptation requirements, and traffic safety strategies. For example: the pre-entry zone (D1) mainly serves a guiding function, using low-intensity lighting for visual pre-adaptation; the entrance adaptation zone (D2) needs to provide high-brightness, gradually brightening lighting to alleviate the visual drop caused by the black hole effect; the middle zone (D3) uses a baseline brightness, dynamically lighting the corresponding sections based on the predicted vehicle travel time to ensure visibility and energy saving; the exit adaptation zone (D4) is set with a gradual dimming transition to prevent sudden glare caused by the white hole effect, while strengthening the exit guidance lighting; the rear zone (D5) of the exit section provides a short-term low-brightness lighting buffer to help drivers smoothly transition from the dark environment to the external environment, ensuring visual safety and continuity. The specific lighting parameters are set manually.

[0060] In generating the predicted traffic lighting demand Q n During the process, the corresponding vehicle traffic lighting demand factors are matched based on the key characteristics of vehicle traffic, and the standard lighting templates for each tunnel section are dynamically adjusted. For example, for small vehicles, the standard template is usually used, and the lighting advance time can be appropriately shortened in the entrance and exit areas; for medium or large vehicles, the lighting range needs to be expanded and the lights need to be turned on in advance to eliminate potential blind spots; for special vehicles such as hazardous chemical transport vehicles or fire trucks, synchronous lighting is triggered for the entire tunnel section to ensure continuous visibility; if a suspected following vehicle structure is identified, the lighting paths of the two vehicles in front and behind will be merged, and the lighting time window will be uniformly extended to prevent lighting gaps; if the vehicle speed is abnormally low, the duration of lighting in each section will be extended to avoid the lights going out before the vehicle has left; when a vehicle has an emergency tag, the pre-lighting of the entire tunnel will be turned on first, and the lighting of the front and rear sections will be synchronized to avoid lighting blind spots. In situations with light interference, such as strong forward reflection, the brightness threshold of the entrance area should be increased to prevent the sensor from misjudging and not turning on the lights. In scenarios where the white hole effect may occur at the exit, the dimming time of the exit area lighting should be increased and the tunnel exit buffer lighting time should be extended to ensure traffic safety and visual continuity.

[0061] The lighting analysis unit is used for tunnel section D n Establish a lighting demand queue Z n Lighting demand queue Z n It contains all currently received predicted traffic lighting requirements; based on the lighting demand queue Z n The output of sunlight reflection lighting characteristics and tunnel section lighting control strategy P n ;

[0062] The specific steps for outputting sunlight reflection illumination characteristics based on external ambient light data include:

[0063] External ambient lighting data includes both front and back external ambient lighting data.

[0064] Acquire the external ambient lighting data between tunnel segment D1 and tunnel segment D2, and acquire the external ambient lighting data between tunnel segment D4 and tunnel segment D5.

[0065] External ambient light data acquisition points were set up between tunnel sections D1 and D2, and between tunnel sections D4 and D5, to capture changes in ambient light before and after vehicles enter and exit the tunnel, reflecting the true intensity of sunlight interference and reflection. Areas D1 and D2 are transition zones where vehicles move from strong sunlight into a relatively dim tunnel, prone to black hole effects or strong sunlight reflection, affecting driving safety and sensor judgment. Therefore, accurate acquisition of ambient and reflected light characteristics before entering the tunnel is necessary. Areas D4 and D5 are transition zones where vehicles exit the tunnel into a bright light environment, often accompanied by white hole effects, which can cause temporary loss of driver vision. Therefore, acquisition of external light characteristics before and after exiting the tunnel is necessary to assist in developing reasonable exit lighting dimming strategies and delayed lighting buffer mechanisms, thereby achieving dynamic perception and intelligent adaptation of the overall lighting system to the natural light environment.

[0066] Both the front and rear ambient lighting data include ambient lighting data, reflected light data, and lighting intensity data;

[0067] Starting from the vehicle's passage timestamp, the front illumination sampling period is calculated based on the vehicle's initial speed. Within the front illumination sampling period, front external ambient illumination data is acquired, and a front external ambient illumination change sequence is constructed based on the front external ambient illumination data.

[0068] After identifying the timestamp of a vehicle passing through tunnel segment D1, the estimated travel time of the vehicle in the area before entering the tunnel is calculated starting from this time point and combined with the initial speed of the vehicle, thereby determining the forward illumination sampling period. Subsequently, the forward illumination acquisition operation is started synchronously within this time window to acquire ambient illumination data, reflected light data, and lighting intensity data of the area before the vehicle enters in real time. The acquired data are arranged in chronological order to form a sequence of changes in the external ambient illumination, reflecting the dynamic trend of light intensity changes over time, providing continuous and accurate raw illumination information support for subsequent illumination synchronization analysis, illumination gradient calculation, and extraction of sunlight reflection features.

[0069] Based on the predicted average vehicle speeds V3 and V4, an early illumination sensing window is obtained; the subsequent illumination sampling period is set based on the early illumination sensing window; subsequent external ambient illumination data is acquired within the subsequent illumination sampling period, and a subsequent external ambient illumination change sequence is constructed based on the subsequent external ambient illumination data.

[0070] After predicting that a vehicle is about to enter tunnel section D3, the estimated time for the vehicle to reach the exit adaptation zone is calculated based on its average speed in areas D3 and D4. This establishes an advance perception window to predict changes in the lighting environment before and after the tunnel. Within this advance perception window, a post-lighting sampling period is set, and ambient light data, reflected light data, and illumination intensity data of the exit area are continuously collected during this period. The collected data are organized in chronological order into a post-external ambient light change sequence to dynamically reflect the trend of natural light changes before and after the vehicle exits the tunnel, providing data support for optimizing the dimming control and buffering strategy of the exit section lighting.

[0071] The rate of change of illumination gradient is calculated based on the previous external ambient illumination change sequence; the synchronization of ambient illumination data, reflected light data and illumination intensity data in the previous external ambient illumination data is judged to obtain illumination synchronization characteristics; the spectral time difference is extracted based on the previous external ambient illumination change sequence; the previous solar reflected illumination characteristics are obtained based on the illumination synchronization characteristics and the spectral time difference.

[0072] The analysis of the ambient light change sequence was performed to calculate the change in light intensity per unit time, obtaining the light gradient change rate to determine the severity of light changes in the tunnel entry area. The temporal synchronicity of the collected ambient light data, reflected light data, and illumination intensity data was assessed to evaluate whether their trends were consistent, thereby extracting light synchronicity characteristics and identifying the degree of coupling between natural light and reflected interference. Simultaneously, based on the time-dimension response differences of different spectral components, a spectral time difference index was extracted to identify the presence of highly reflective materials or interfering light sources. Combining the light synchronicity characteristics and spectral time difference, a comprehensive determination was made regarding the presence of significant sunlight reflection interference in the current tunnel entry environment, thereby extracting accurate pre-sunlight reflection illumination characteristics for optimizing the triggering strategy and brightness adjustment logic of the tunnel entry section lighting.

[0073] Similarly, based on the post-external ambient light change sequence, the post-sunlight reflection illumination characteristics are obtained. Based on this sequence, the rate of change in light intensity per unit time is first calculated to obtain the light gradient change rate, which is used to assess the drastic change in natural light in the tunnel exit area. Subsequently, in the post-external ambient light data, time axis alignment and correlation analysis are performed on the ambient light data, reflected light data, and illumination intensity data to determine the consistency of the three types of data in their change trends, thereby extracting illumination synchronicity characteristics. The response delay of different spectral channels during the change process is analyzed, and the spectral time difference index is extracted to identify whether there is strong reflection or sudden light source interference in the exit area. Combining the illumination synchronicity characteristics and the spectral time difference, post-sunlight reflection illumination characteristics are generated to assist in the dynamic adjustment of the lighting strategy in the exit adaptation area, optimizing the dimming control, delay duration, and buffer section brightness settings of the tunnel exit lighting.

[0074] The characteristics of the front and rear sunlight reflection illumination are combined to obtain the characteristics of sunlight reflection illumination.

[0075] Based on lighting demand queue Z n The output of sunlight reflection lighting characteristics and tunnel section lighting control strategy P n The specific steps include:

[0076] Based on lighting demand queue Z n All predicted traffic lighting requirements received internally are pooled to obtain the initial control strategy C for the tunnel segment. n After receiving all predicted traffic lighting demands from the tunnel section lighting demand queue, the system first parses and aligns the lighting control parameters contained in each demand with the time axis. Then, it uses pooling to merge overlapping or adjacent lighting demands, extracts the common features of the lighting demands, and forms a unified time interval and brightness configuration strategy. Based on this, according to the demand density, vehicle type distribution, and degree of interval overlap, it rationally arranges the lighting group activation sequence, lighting duration, and spatial distribution, thereby generating an initial control strategy for the tunnel section that covers all current effective predicted lighting demands, providing a structured lighting baseline for subsequent optimized scheduling.

[0077] Based on the characteristics of front and rear sunlight reflection illumination and the swarm optimization algorithm, the initial control strategy C for the tunnel section is proposed. n After optimization and adjustment, the tunnel section lighting control strategy P was obtained. n Among them, the swarm optimization algorithm is the tuna swarm optimization algorithm;

[0078] First, the initial lighting control strategy C for the tunnel section... n The lighting control parameters are encoded as optimizable solution vectors, serving as the basic structure in the search space. A tuna swarm consisting of candidate solutions is then generated, with each individual representing a possible lighting strategy. Key parameters of the tuna swarm optimization algorithm, such as population size, maximum number of iterations, migration probability, and adaptive weighting factor, are set to provide a balance between global convergence and local jump capabilities for subsequent searches. These key parameters are set manually.

[0079] For each candidate lighting strategy, a fitness function is constructed that comprehensively considers energy efficiency, safety, and illumination adaptability. Energy efficiency is estimated by multiplying the lighting time and brightness, while safety is measured by the overlap with the target time window in the lighting demand queue. Illumination adaptability incorporates the characteristics of front and rear sunlight reflection illumination, which are applied to the brightness correction factors of the tunnel entry and exit sections, respectively, to evaluate the robustness of the strategy under complex illumination interference scenarios. The fitness function aims to minimize energy consumption while imposing constraints to ensure that the safe lighting range is met.

[0080] In each iteration, individual tuna update their strategies based on their fitness value, current optimal individual position, and migration behavior model: some individuals move closer to the current optimal strategy to swim in a herd to accelerate global convergence; others simulate random walks to explore boundary solutions and escape local optima. The update process uses an adaptive perturbation factor to guide the adjustment of parameters such as the lighting time window and brightness level of the light group within a reasonable range. In the updated candidate strategies, the brightness threshold and lighting advance time of the tunnel entry section (D1, D2) are further adjusted by integrating the pre-sunlight reflection characteristics to counteract the black hole effect; at the same time, the dimming intensity and delay time of the exit section (D4, D5) are optimized by integrating the post-sunlight reflection characteristics to ensure the mitigation of the white hole effect. This stage plays a role in local illumination adjustment, performing target-specific compensation for key transition intervals that were not refined in the global optimization, and improving the lighting system's response to extreme environments. The system terminates when the preset number of iterations is reached or the fitness function converges. The system selects the lighting strategy represented by the current best individual as the final lighting control strategy for the tunnel section. This strategy takes into account lighting demand coverage, ambient light interference adaptation, and overall energy consumption optimization. It supports vehicle-mounted segment triggering and dynamic adjustment, providing an efficient, safe, and intelligent lighting control solution for practical applications.

[0081] The lighting control module is used to continuously update the lighting demand queue Z. n and tunnel section lighting control strategy P n The current highway tunnel lighting is controlled according to the tunnel lighting intervals. In this embodiment, to achieve dynamic and precise control of highway tunnel lighting, it is necessary to continuously receive and update the lighting demand queue and tunnel section lighting control strategy, and to perform real-time lighting scheduling according to the tunnel lighting intervals. To avoid logical errors during lighting replacement (such as premature extinguishing, repeated lighting, or lighting gaps), a time window queue mechanism is adopted to maintain an ordered task buffer queue for each lighting interval. All lighting control commands are processed in three stages: judgment, issuance, and execution, based on the predicted traffic time window. When a new demand arrives, the lighting strategy is prioritized and tasks are merged to ensure seamless connection between the old and new strategies. At the same time, a redundant lighting maintenance mechanism is set up to reserve a minimum brightness buffer time at the lighting switching point to ensure a smooth transition of lighting changes and avoid visual abrupt changes or control conflicts, thereby achieving a safe, continuous, and logically flawless intelligent lighting replacement process.

[0082] For example, in the actual use of a vehicle-mounted lighting system in a highway tunnel, the following situation may occur: A large truck A enters a highway tunnel at a normal speed. Its traffic characteristics have been identified by sensors in tunnel section D1, and a predicted lighting path has been established. Based on this, the lighting requirements Q for tunnel sections D2 to D5 have been generated. n Lighting demand queue Z n The system then updates and outputs the initial lighting control strategy P. nThe relevant lights are controlled to illuminate according to the path. About 5 seconds after truck A passes through, a small car B follows closely into the tunnel. Due to its higher speed, it is predicted that it will approach or overtake the rear of truck A in section D3.

[0083] After car B is identified at point D1, the system also collects its occlusion waveform, identification information, and passage timestamp to calculate its initial speed and predicted lighting path. Due to the short distance between vehicles, the system marks it as a following vehicle. The system finds that the predicted lighting path of car B overlaps significantly with that of the preceding vehicle A, especially in the D3-D4 section. At this point, the lighting analysis module calculates the lighting requirements Q for the preceding and following vehicles. n Perform pooling and fusion processing to form a unified lighting demand window; if the lighting control strategy P n If a segment of lights is already on, its illumination duration will only be extended, and repeated on / off commands will be avoided to prevent wasted system resources or control logic conflicts. The lighting control module will then operate in the merged queue Z. n Based on this, the system combines current characteristics of sunlight reflection lighting (such as strong reflection at the entrance and white hole effect at the exit) with the tuna swarm optimization algorithm to jointly optimize lighting brightness, on-time, and transition zone settings. In particular, for the scenario where car B is predicted to quickly exit the D4-D5 area, the system avoids visual discrepancies by extending the tail time of the tunnel exit lighting and adjusting the dimming strategy.

[0084] In multi-vehicle traffic scenarios, the system first uses sensors in tunnel section D1 to identify vehicle information, occlusion features, and timestamps for each vehicle, and then generates their respective predicted lighting paths and traffic lighting requirements Q. n The lighting analysis module will analyze the Q values ​​of multiple vehicles. n All are uniformly added to the lighting demand queue Z of the corresponding tunnel section. n It then performs time axis alignment and pooling to merge overlapping or adjacent lighting requirements into continuous lighting intervals, forming the initial lighting control strategy C. n Subsequently, the system, based on the current characteristics of sunlight reflection illumination, employs a population optimization algorithm to optimize C. n Perform dynamic optimization to generate the final control strategy P n Ensure that all vehicle path lighting needs are covered without conflict; the lighting control module updates Z in real time. n With P n The system precisely controls the timing of the lights turning on and off according to the interval, while setting up a buffer window and lighting redundancy mechanism to prevent lighting gaps, premature lights turning off, or repeated lights turning on caused by changes in vehicle speed, following vehicles, or differences in vehicle type. This ensures smooth and intelligent lighting transitions and energy-saving control in scenarios with multiple vehicles converging or dense traffic.

[0085] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims. Parts not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time, characterized in that, include: The interval division module is used to define how many tunnel lighting intervals the current highway tunnel contains; The tunnel lighting sections are divided into 5 tunnel segments D according to their location. n n = 1, 2, ..., 5; The lighting analysis module includes an information acquisition unit and a lighting analysis unit. The information acquisition unit acquires real-time vehicle identification information, vehicle passage timestamps, and vehicle occlusion features in tunnel section D1. Based on the vehicle identification information, vehicle passage timestamps, and vehicle occlusion features, it analyzes the data to obtain key vehicle passage features and predict vehicle lighting paths. Based on the predicted vehicle lighting paths, it generates a predicted traffic lighting demand Q. n Simultaneously, acquire external ambient light data and output sunlight reflection illumination characteristics based on the external ambient light data; Vehicle occlusion features are used to identify the duration of occlusion, waveform integrity, whether there are abrupt changes or multiple peaks, and to distinguish whether it is a false triggering by interfering objects. Predicting vehicle lighting paths involves traversing each tunnel segment, calculating the vehicle's entry and exit times within each segment based on the predicted average speed of the vehicle in each segment and the length of the tunnel segment, and then combining the lighting time windows of each segment into a continuous lighting interval. The lighting analysis unit is used for tunnel section D n Establish a lighting demand queue Z n Lighting demand queue Z n It contains all currently received predicted traffic lighting requirements; based on the lighting demand queue Z n The output of sunlight reflection lighting characteristics and tunnel section lighting control strategy P n ; The lighting control module is used to continuously update the lighting demand queue Z. n and tunnel section lighting control strategy P n The lighting fixtures in the current highway tunnels are controlled according to the tunnel lighting zones; The specific steps for analysis based on vehicle identification information, vehicle passage timestamps, and vehicle occlusion features include: Based on vehicle occlusion features, occlusion waveform analysis is performed to extract occlusion duration, occlusion intensity variation curves, and occlusion integrity indicators; a pre-trained vehicle feature recognition model is used for analysis to output vehicle passage labels. In the occlusion waveform analysis, the occlusion duration represents the total time that the vehicle occludes the infrared signal within the sensor's field of view, reflecting the vehicle's length or speed. The occlusion intensity change curve describes the trend of the infrared signal intensity changing over time during the occlusion process, used to determine whether there is abnormal interference. The occlusion integrity index comprehensively measures the continuity, smoothness, and morphological integrity of the waveform, used to determine whether the occlusion event is a real and stable vehicle passage, or whether there is an abnormal situation. Vehicle category labels are obtained by identifying special tags based on vehicle identification information. The initial speed of the vehicle is calculated based on the vehicle's time stamp; the predicted vehicle lighting path is then calculated based on the initial speed of the vehicle and the current highway tunnel. By combining vehicle access tags and vehicle category tags, key characteristics of vehicle access can be obtained. Based on the initial speed of the vehicle and the current highway tunnel, the specific steps for predicting the vehicle lighting path are calculated, including: Based on historical vehicle traffic data, for tunnel section D n Establish tunnel traffic impact factor Y n Let the initial speed of the vehicle be V1; in tunnel section D n In the middle, the predicted average vehicle speed V is corrected. n , where V n =V n-1 *Y n ; Based on the predicted average vehicle speed V n and tunnel section D n Tunnel section length L n Predicted in tunnel section D n Vehicle entry time T n1 and vehicle departure time T n2 ; Traverse all tunnel segments D n The predicted vehicle lighting path is obtained by combining the results. The specific steps for outputting sunlight reflection illumination characteristics based on external ambient light data include: External ambient lighting data includes both front and back external ambient lighting data. Both the pre-external ambient lighting data and the post-external ambient lighting data include ambient lighting data, reflected light data, and illumination intensity data; a pre-external ambient lighting change sequence is constructed based on the pre-external ambient lighting data; a post-external ambient lighting change sequence is constructed based on the post-external ambient lighting data. The rate of change of illumination gradient is calculated based on the previous external ambient illumination change sequence; the synchronization of ambient illumination data, reflected light data and illumination intensity data in the previous external ambient illumination data is judged to obtain illumination synchronization characteristics; the spectral time difference is extracted based on the previous external ambient illumination change sequence; the previous solar reflected illumination characteristics are obtained based on the illumination synchronization characteristics and the spectral time difference. Similarly, the characteristics of post-sunlight reflection illumination were obtained by analyzing the post-external ambient light change sequence; The characteristics of the front and rear sunlight reflection illumination are combined to obtain the characteristics of sunlight reflection illumination.

2. The highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time according to claim 1, characterized in that, Predicted traffic lighting demand Q is generated based on predicted vehicle lighting paths. n The specific steps include: For tunnel section D n Construct a standard lighting response template R n ; Matching vehicle traffic lighting demand factors based on key vehicle traffic features. Based on the vehicle traffic lighting demand factor, the standard lighting response template R is applied. n Lighting corrections are performed to obtain the predicted traffic lighting demand Q. n .

3. The highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time according to claim 2, characterized in that, Based on lighting demand queue Z n The output of sunlight reflection lighting characteristics and tunnel section lighting control strategy P n The specific steps include: Based on lighting demand queue Z n All predicted traffic lighting requirements received internally are pooled to obtain the initial control strategy C for the tunnel segment. n ; Based on the characteristics of front and rear sunlight reflection illumination and the swarm optimization algorithm, the initial control strategy C for the tunnel section is proposed. n After optimization and adjustment, the tunnel section lighting control strategy P was obtained. n .

4. The highway tunnel vehicle lighting system based on interval intelligent calculation of lamp adjustment time according to claim 3, characterized in that, in, The swarm optimization algorithm is the tuna swarm optimization algorithm.

Images