A tunnel lighting adaptive regulation method and system based on multi-source data fusion

Abstract

The application provides a tunnel lighting adaptive regulation method and system based on multi-source data fusion, belongs to the technical field of lighting control, performs multi-source sensing on the environment inside and outside the tunnel hole, generates an original multi-source data set of the current tunnel environment; performs driving visual recognition evaluation based on the original multi-source data set to obtain driving visual recognition features; inverses according to the relationship between the driving visual recognition features and the vehicle running speed and the safe parking distance constraint condition to obtain a lighting parameter threshold set meeting the safe driving demand; generates a lighting regulation strategy of each lighting section in the tunnel according to the lighting parameter threshold set and the running state vector of the current tunnel, and regulates the lighting lamps of each lighting section in the tunnel according to the lighting regulation strategy. The application can perform dynamic inversion based on multi-source data and driving visual recognition safety to generate a segmented adaptive regulation strategy, so as to realize the safety and comfort of tunnel lighting.

Description

Technical Field

[0001] This invention relates to the field of lighting control technology, and more specifically, to a method and system for adaptive control of tunnel lighting based on multi-source data fusion. Background Technology

[0002] With the rapid development of highway networks, tunnels are increasingly used in complex terrain. Tunnel lighting, as a crucial infrastructure for ensuring driving safety, directly impacts drivers' visibility and driving safety through its light environment quality. Most existing tunnel lighting systems employ fixed brightness control methods based on experience or specifications, typically adjusting brightness only according to the ambient light outside the tunnel. This lacks comprehensive consideration of factors such as traffic flow, vehicle speed, weather conditions, and light source color temperature, leading to significant shortcomings in actual operation. Because a quantitative relationship between the light environment and visibility has not been established based on drivers' visual mechanisms, current technologies struggle to accurately assess drivers' visual distance and reaction time under different conditions, thus failing to ensure that lighting parameters consistently meet safe stopping distance requirements. Therefore, how to dynamically invert data based on multi-source data and driving visual safety to generate segmented adaptive control strategies, thereby achieving both safety and comfort in tunnel lighting, remains a challenge for the industry. Summary of the Invention

[0003] This invention provides a method and system for adaptive control of tunnel lighting based on multi-source data fusion. It can perform dynamic inversion based on multi-source data and driving visual safety to generate a segmented adaptive control strategy, thereby achieving both safety and comfort in tunnel lighting.

[0004] In a first aspect, the present invention provides a method for adaptive control of tunnel lighting based on multi-source data fusion, comprising the following steps: Multi-source sensing of the environment inside and outside the tunnel is performed to generate the original multi-source dataset of the current tunnel environment; Based on the original multi-source dataset, driving vision assessment is performed to obtain driving vision features in the current tunnel environment; Based on the relationship between the driving visual characteristics and the vehicle speed, and the safety parking distance constraint, an inversion is performed to obtain a set of lighting parameter thresholds that meet the requirements of safe driving. Based on the set of lighting parameter thresholds and the current tunnel operating state vector, a lighting control strategy for each lighting segment in the tunnel is generated, and the lighting fixtures for each lighting segment in the tunnel are controlled according to the corresponding lighting control strategy.

[0005] Preferably, multi-source sensing of the environment inside and outside the tunnel to generate the original multi-source dataset of the current tunnel environment specifically includes: Ambient light sensing units and meteorological monitoring units are deployed in typical impact areas outside the tunnel to acquire ambient light data and meteorological data outside the tunnel. The ambient light data outside the tunnel includes ambient brightness and ambient color temperature, and the meteorological data includes weather type and visibility. The tunnel lighting sensing unit is synchronously triggered, and multi-point brightness sampling devices are deployed in the entrance section, transition section, middle section and exit section to collect the ambient light data inside the tunnel. The ambient light data inside the tunnel includes the brightness information of each lighting section in the tunnel. The traffic operation detection unit is invoked synchronously to acquire traffic operation data reflecting the traffic operation status through video recognition and radar detection. The traffic operation data includes traffic flow and vehicle speed. The original multi-source dataset of the current tunnel environment is composed of the ambient light data outside the tunnel, the meteorological data, the ambient light data inside the tunnel, and the traffic operation data.

[0006] Preferably, the driving vision assessment based on the original multi-source dataset, to obtain the driving vision features in the current tunnel environment, specifically includes: Multidimensional feature extraction is performed on the original multi-source dataset to obtain the current tunnel operation status vector, which includes the ambient brightness and color temperature inside and outside the tunnel, weather level, traffic flow and vehicle speed features; The operating state vector is input into a pre-constructed driving vision performance model, which then outputs the driving vision features in the current tunnel environment.

[0007] Preferably, inputting the operating state vector into a pre-constructed driving vision performance model, and then outputting the driving vision features in the current tunnel environment, specifically includes: The components in the running state vector are subjected to structured parsing to generate an input feature set for visual recognition analysis; The input feature set is input into a pre-constructed driving vision performance model, which is composed of a physical model based on the human visual perception mechanism and a data-driven model trained based on experimental data. Based on the physical model, the contrast between the target object and the background is determined according to the principle of brightness contrast, and the minimum distinguishable contrast threshold under the current light environment is determined by combining the contrast sensitivity characteristics of the human eye, thereby obtaining the initial visual distance estimate. The initial visual distance estimate, along with traffic flow and vehicle speed features in the operating state vector, is input into the data-driven model. The initial visual distance estimate is corrected through a nonlinear mapping relationship. Based on the corrected visual distance estimate and combined with vehicle speed features, the driver's visual reaction time and safe braking margin are determined. By introducing a visual comfort evaluation function, the degree of visual fatigue under different combinations of brightness and color temperature is quantitatively evaluated, and a visual comfort index is obtained. The corrected visual distance estimate, the visual comfort index, the visual reaction time, and the safety braking margin are integrated to output the driving visual characteristics in the current tunnel environment.

[0008] Preferably, the set of lighting parameter thresholds that meet safe driving requirements is obtained by inversion based on the relationship between the driving visual features and the vehicle speed, as well as the safe stopping distance constraint. Specifically, this includes: The visual reaction time and vehicle speed in the driving visual recognition features are coupled and analyzed to obtain the safe stopping distance in the current tunnel environment; Based on the safe parking distance and the corrected visual distance estimate, a safety constraint criterion is constructed, and a visual comfort index is introduced as an adjustment factor, thereby establishing a reverse mapping mechanism from visual distance to lighting parameters. In the inversion process of the reverse mapping mechanism, a safety braking margin is introduced as a redundancy constraint, and then the initial lighting parameter threshold set of each lighting segment of the tunnel is determined according to the reverse mapping mechanism. The initial lighting parameter threshold set for each lighting segment is corrected by using traffic flow changes, thereby obtaining a lighting parameter threshold set that meets the requirements of safe driving.

[0009] Preferably, generating lighting control strategies for each lighting segment in the tunnel based on the set of lighting parameter thresholds and the current tunnel operating state vector specifically includes: The set of lighting parameter thresholds and the current tunnel operating state vector are correlated and mapped to generate a rigid constraint framework in the current tunnel environment. Based on the rigid constraint framework, a multi-objective optimization decision model with a three-layer hierarchical structure is constructed. The first layer of the multi-objective optimization decision model is the safety layer objective function, the second layer is the comfort layer objective function, and the third layer is the energy-saving layer objective function. The multi-objective optimization decision model is solved using the lexicographical order optimization method, and then outputs the lighting control strategy for each lighting segment in the tunnel.

[0010] Preferably, the specific methods for controlling the lighting fixtures in each lighting section of the tunnel according to the corresponding lighting control strategy include: For each lighting segment in the tunnel, a longitudinal lighting distribution curve is generated based on the corresponding lighting control strategy; The longitudinal lighting distribution curve is mapped to each specific physical lamp in the corresponding lighting segment, thereby generating digital control instructions for each lighting segment that can be directly identified and executed by the lamp driver power supply. The digital control instructions include entry segment control instructions, transition segment control instructions, intermediate segment control instructions and exit segment control instructions. The estimated time when the vehicle arrives at the tunnel entrance is recorded as the baseline time, and the estimated time when the vehicle arrives at the starting point of each lighting section is determined based on the starting mileage marker of each lighting section in the tunnel. At a preset lead time before the arrival of the reference time, the entrance section control command is issued to the entrance section and the approach lighting fixtures in front of the entrance section, and the vehicle tracking triggering mechanism is used to issue control commands to the transition section, intermediate section and exit section respectively, thereby realizing the control of the lighting fixtures in each lighting section.

[0011] Secondly, the present invention provides a tunnel lighting adaptive control system based on multi-source data fusion, used to execute a tunnel lighting adaptive control method based on multi-source data fusion, comprising: The multi-source sensing module is used to perform multi-source sensing of the environment inside and outside the tunnel and generate the original multi-source dataset of the current tunnel environment. The feature extraction module is used to perform driving vision assessment based on the original multi-source dataset to obtain driving vision features in the current tunnel environment; The parameter inversion module is used to perform inversion based on the relationship between the driving visual characteristics and the vehicle operating speed, as well as the safe parking distance constraint, to obtain a set of lighting parameter thresholds that meet the requirements of safe driving. The lighting control module is used to generate lighting control strategies for each lighting segment in the tunnel based on the set of lighting parameter thresholds and the current tunnel operating state vector, and to control the lighting fixtures of each lighting segment in the tunnel according to the corresponding lighting control strategies.

[0012] Thirdly, the present invention provides a computer device, the computer device including a memory and a processor, the memory storing code, the processor being configured to acquire the code and execute the above-described adaptive control method for tunnel lighting based on multi-source data fusion.

[0013] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described adaptive control method for tunnel lighting based on multi-source data fusion.

[0014] The technical solution provided by this invention has the following beneficial effects: This invention provides a method and system for adaptive control of tunnel lighting based on multi-source data fusion. First, by fusing and analyzing multi-source data and mapping it to driver visual characteristics, the originally dispersed and single-dimensional environmental and traffic information can be transformed into a comprehensive evaluation result centered on driver visual perception, thus accurately reflecting the essential visual needs of drivers under different operating conditions. Then, by coupling driver visual characteristics with vehicle speed and introducing a safe stopping distance as a rigid constraint, the safety margin under the current lighting environment can be accurately determined, and the optimal brightness, color temperature, and uniformity range required to meet this safety margin can be derived in reverse. Finally, by coupling the set of lighting parameter thresholds with the operating state vector, the target brightness, color temperature, and uniformity range of each lighting segment under the current operating conditions can be accurately determined, and further transformed into an executable control strategy. Through a segmented control mechanism, differentiated control is implemented for the different visual needs of the entrance, transition, middle, and exit segments, effectively mitigating the black hole and white hole effects, improving the continuity of driver visual adaptation, and thus achieving synergistic optimization of tunnel lighting in terms of safety, comfort, and energy efficiency.

[0015] In summary, the technical solution adopted in this invention can perform dynamic inversion based on multi-source data and driving visual safety to generate a segmented adaptive control strategy, thereby achieving both safety and comfort in tunnel lighting. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is an exemplary flowchart of an adaptive control method for tunnel lighting based on multi-source data fusion, as shown in some embodiments of the present invention. Figure 2 This is a schematic diagram of the application scenario architecture of the tunnel lighting adaptive control method based on multi-source data fusion according to some embodiments of the present invention; Figure 3 This is an exemplary flowchart illustrating the determination of driving visual features according to some embodiments of the present invention; Figure 4 This is a schematic diagram of the structure of a tunnel lighting adaptive control system based on multi-source data fusion, according to some embodiments of the present invention; Figure 5This is a schematic diagram of the structure of a computer device for implementing an adaptive control method for tunnel lighting based on multi-source data fusion, as shown in some embodiments of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] This invention provides a method and system for adaptive control of tunnel lighting based on multi-source data fusion. The core of this method involves multi-source perception of the tunnel's internal and external environment to generate an original multi-source dataset of the current tunnel environment; driving visual perception assessment based on this dataset to obtain driving visual perception features in the current tunnel environment; inversion based on the relationship between these features and vehicle speed, as well as safe stopping distance constraints, to obtain a set of lighting parameter thresholds that meet safe driving requirements; and generation of lighting control strategies for each lighting segment within the tunnel based on the lighting parameter threshold set and the current tunnel's operating state vector. The lighting fixtures in each segment are then controlled according to the corresponding lighting control strategies. This approach allows for dynamic inversion based on multi-source data and driving visual safety to generate segmented adaptive control strategies, thereby achieving both safety and comfort in tunnel lighting.

[0020] To better understand the above technical solutions, a detailed description of the technical solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. (Refer to...) Figure 1 This figure is an exemplary flowchart of an adaptive control method for tunnel lighting based on multi-source data fusion according to some embodiments of the present invention. The figure mainly includes the following steps: In step S101, multi-source sensing is performed on the environment inside and outside the tunnel to generate the original multi-source dataset of the current tunnel environment.

[0021] In some embodiments, multi-source sensing of the environment inside and outside the tunnel to generate the original multi-source dataset of the current tunnel environment can be performed in the following ways: Ambient light sensing units and meteorological monitoring units are deployed in typical impact areas outside the tunnel to acquire ambient light data and meteorological data outside the tunnel. The ambient light data outside the tunnel includes ambient brightness and ambient color temperature, and the meteorological data includes weather type and visibility. The tunnel lighting sensing unit is synchronously triggered, and multi-point brightness sampling devices are deployed in the entrance section, transition section, middle section and exit section to collect the ambient light data inside the tunnel. The ambient light data inside the tunnel includes the brightness information of each lighting section in the tunnel. The traffic operation detection unit is invoked synchronously to acquire traffic operation data reflecting the traffic operation status through video recognition and radar detection. The traffic operation data includes traffic flow and vehicle speed. The original multi-source dataset of the current tunnel environment is composed of the ambient light data outside the tunnel, the meteorological data, the ambient light data inside the tunnel, and the traffic operation data.

[0022] In this invention, Figure 2 This is a schematic diagram of an application scenario architecture for an adaptive control method for tunnel lighting based on multi-source data fusion, as shown in some embodiments of the present invention. Figure 2 As shown, in specific implementation, firstly, ambient light sensing units and meteorological monitoring units are deployed in the typical visual adaptation zone outside the tunnel before vehicles enter. The ambient light sensing unit samples the incident light using multispectral photosensitive devices, converts the light signal into an electrical signal using the photoelectric conversion principle, and calculates the ambient brightness and comprehensive color temperature parameters using a calibration model. Simultaneously, the meteorological monitoring unit inverts the air particle concentration based on the light scattering and transmission attenuation mechanism, thereby identifying the current weather type and calculating the visibility index, thus forming time-stamped ambient light and meteorological data outside the tunnel. After completing the data acquisition outside the tunnel, the corresponding time reference is used as a synchronous trigger signal to drive the tunnel lighting sensing unit to start segmented sampling. Multi-point brightness sampling devices are deployed in the entrance section, transition section, middle section, and exit section. The brightness values ​​of each sampling point are collected in real time, and the average brightness and brightness distribution characteristics of each segment are calculated using spatial weighted average and extreme value analysis methods, thus obtaining ambient light data inside the tunnel with spatial segmentation attributes. Based on this, using the acquired indoor and outdoor light environment data as a reference time series, the traffic operation detection unit is synchronously invoked. Vehicle targets are detected and tracked using video image recognition technology, and instantaneous vehicle speed and traffic flow information are obtained by combining radar speed measurement principles. Statistical analysis of the number and speed distribution of vehicles per unit time yields traffic operation data reflecting the current traffic status. Finally, the aforementioned outdoor ambient light data, meteorological data, indoor ambient light data, and traffic operation data are aligned according to a unified timestamp, and the data is integrated and structured according to the tunnel's spatial segmentation structure, ultimately generating the original multi-source dataset of the current tunnel environment.

[0023] In step S102, driving vision assessment is performed based on the original multi-source dataset to obtain driving vision features in the current tunnel environment.

[0024] Preferably, in some embodiments, reference is made to Figure 3 As shown, this figure is an exemplary flowchart for determining driving vision features according to some embodiments of the present invention. In this embodiment, driving vision evaluation based on the original multi-source dataset is performed to obtain driving vision features in the current tunnel environment, which can be achieved by the following steps: In step S1021, multidimensional feature extraction is performed on the original multi-source dataset to obtain the current tunnel operation status vector. The operation status vector includes the ambient brightness and color temperature inside and outside the tunnel, weather level, traffic flow and vehicle speed features. In step S1022, the operating state vector is input into a pre-constructed driving vision performance model, and then the driving vision features in the current tunnel environment are output.

[0025] In practical implementation, firstly, multidimensional feature extraction can be performed on the original multi-source dataset. This involves using a unified time series as a benchmark to perform time alignment and outlier correction on various data types within the original multi-source dataset, and eliminating noise and missing data through sliding window filtering and interpolation compensation methods. Secondly, the data can be segmented and mapped according to the tunnel's spatial structure. This involves associating external ambient light and meteorological data with the entrance influence zone, integrating internal ambient light data into entrance, transition, intermediate, and exit sections, and mapping traffic operation data to corresponding spatial segments according to their corresponding times, thus forming a multidimensional data structure with spatiotemporal coupling characteristics. Finally, feature extraction and dimensional unification processing can be performed on the above data. Normalization methods can eliminate scale differences between different physical quantities, and key feature components reflecting changes in the light environment, meteorological influences, and traffic operation characteristics can be extracted. These include internal and external ambient brightness and color temperature, meteorological levels, traffic flow, and vehicle speed characteristics. The feature vector composed of all extracted features is then used as the current tunnel operation state vector.

[0026] In some embodiments, inputting the operating state vector into a pre-built driving vision performance model, and then outputting the driving vision features in the current tunnel environment, can be done in the following way: The components in the running state vector are subjected to structured parsing to generate an input feature set for visual recognition analysis; The input feature set is input into a pre-constructed driving vision performance model, which is composed of a physical model based on the human visual perception mechanism and a data-driven model trained based on experimental data. Based on the physical model, the contrast between the target object and the background is determined according to the principle of brightness contrast, and the minimum distinguishable contrast threshold under the current light environment is determined by combining the contrast sensitivity characteristics of the human eye, thereby obtaining the initial visual distance estimate. The initial visual distance estimate, along with traffic flow and vehicle speed features in the operating state vector, is input into the data-driven model. The initial visual distance estimate is corrected through a nonlinear mapping relationship. Based on the corrected visual distance estimate and combined with vehicle speed features, the driver's visual reaction time and safe braking margin are determined. By introducing a visual comfort evaluation function, the degree of visual fatigue under different combinations of brightness and color temperature is quantitatively evaluated, and a visual comfort index is obtained. The corrected visual distance estimate, the visual comfort index, the visual reaction time, and the safety braking margin are integrated to output the driving visual characteristics in the current tunnel environment.

[0027] In practical implementation, firstly, the components in the operating state vector can be structurally analyzed. This involves reorganizing and filtering the ambient brightness and color temperature, weather level, traffic flow, and vehicle speed characteristics inside and outside the tunnel according to the human visual perception mechanism. Specifically, ambient brightness, ambient color temperature, and the brightness of each lighting segment are categorized into a light environment feature subset, traffic flow and vehicle speed are categorized into a traffic dynamic feature subset, and weather level and visibility information are categorized into an environmental disturbance feature subset, thereby generating an input feature set for visual perception analysis. Then, the input feature set can be input into a pre-constructed driving visual perception performance model. This model consists of a physical model based on the human visual perception mechanism coupled with a data-driven model trained based on experimental data. In other words, the driving visual perception performance model adopts a coupled architecture that combines mechanistic constraints and data learning. Furthermore, a physical model layer based on the human visual perception mechanism can be used to calculate the brightness contrast between the target object and the background based on the input ambient brightness and color temperature parameters. This involves constructing the contrast relationship between the target object and the background according to the principle of brightness contrast calculation. The contrast value is obtained by normalizing the difference between the target object's brightness and the background brightness, thus quantitatively describing the discernibility of the target object in the current lighting environment. After obtaining the contrast value, a contrast sensitivity characteristic function from the human visual perception mechanism can be introduced. This function comprehensively considers factors such as the current ambient brightness level, light color conditions, and spatial frequency to determine the minimum discernible contrast threshold that can be reliably identified under the human visual system. This minimum contrast threshold reflects the discernible contrast under specific lighting conditions. The minimum resolution of brightness differences by the human eye in the environment; the calculated contrast value can be compared with the minimum distinguishable contrast threshold. When the contrast value is higher than the minimum distinguishable contrast threshold, it means that the target object can be identified at that distance, otherwise it cannot be identified. Based on this, the mapping relationship between the target object distance and its viewing angle is constructed by combining the target object size, viewing angle and field of view geometry. Through step-by-step iterative or inverse calculation methods, the maximum identifiable distance of the target object is solved under the critical condition that the contrast value is equal to the minimum distinguishable contrast threshold, so as to obtain the initial visual distance estimate. This initial visual distance estimate reflects the limit distance at which the driver can identify the target object in front under the current light environment and visual conditions.

[0028] In addition, in specific implementation, firstly, the initial visual distance estimate, along with traffic flow and vehicle speed features from the operating state vector, can be input into the data-driven model. The initial visual distance estimate is then corrected through a nonlinear mapping relationship. This data-driven model is built based on a large amount of real-vehicle test and simulation experimental data. It uses a nonlinear mapping relationship to characterize the coupled influence between complex traffic conditions, changes in the lighting environment, and driving behavior. Within the data-driven model, the initial visual distance is dynamically corrected through multi-layer feature transformation, thereby obtaining a corrected visual distance estimate that is closer to actual working conditions. After obtaining the corrected visual distance estimate, the driver's visual reaction time from target recognition to braking operation can be calculated based on the current vehicle speed and the driving behavior mechanism. The reaction time is corrected for different traffic flow conditions by introducing a correlation function between reaction time and traffic complexity. Furthermore, based on the principles of vehicle kinematics, the corrected visual distance is coupled with vehicle speed for analysis to calculate the remaining distance margin corresponding to the driver's ability to complete safe braking under the current conditions, i.e., the safe braking margin. This safe braking margin is used to measure the redundancy of driving safety under the current visual conditions. Then, a visual comfort evaluation function can be introduced. This function is based on the mechanism of human visual fatigue, using brightness level, color temperature variation amplitude, and rate of change as core variables. By establishing a coupling relationship between brightness adaptation deviation and color temperature adaptation deviation, the adaptive pressure on the driver's visual system under the current lighting environment is calculated. When brightness changes excessively or color temperature changes abruptly, the visual load value output by the visual comfort evaluation function increases accordingly. Based on this, the visual load value is compared with a preset comfort threshold, and the impact of visual stimuli on fatigue during driving is accumulated through continuous-time integration, thus obtaining a quantitative index reflecting the cumulative effect of visual fatigue. This quantitative index is normalized and mapped to a standardized visual comfort index, which characterizes the driver's visual comfort level under the current brightness and color temperature combination. Finally, the feature set consisting of the corrected visual distance estimate, the visual comfort index, visual reaction time, and safe braking margin can be used as the driving visual characteristics in the current tunnel environment.

[0029] It should be noted that by fusing and analyzing multi-source data and further mapping it to driving visual characteristics, the originally scattered and single-dimensional environmental and traffic information can be transformed into a comprehensive evaluation result centered on the driver's visual perception. This makes tunnel lighting control no longer rely on experience thresholds or static standards, but rather on a dynamic judgment mechanism based on real driving visual ability, thus accurately reflecting the driver's essential visual needs under different working conditions.

[0030] In step S103, the relationship between the driving visual features and the vehicle speed, as well as the safe parking distance constraint, is used to perform inversion to obtain a set of lighting parameter thresholds that meet the requirements of safe driving.

[0031] In some embodiments, the set of lighting parameter thresholds that meet safe driving requirements is obtained by inversion based on the relationship between the driving visual features and the vehicle speed, as well as the safe stopping distance constraint. Specifically, this can be achieved in the following manner: The visual reaction time and vehicle speed in the driving visual recognition features are coupled and analyzed to obtain the safe stopping distance in the current tunnel environment; Based on the safe parking distance and the corrected visual distance estimate, a safety constraint criterion is constructed, and a visual comfort index is introduced as an adjustment factor, thereby establishing a reverse mapping mechanism from visual distance to lighting parameters. In the inversion process of the reverse mapping mechanism, a safety braking margin is introduced as a redundancy constraint, and then the initial lighting parameter threshold set of each lighting segment of the tunnel is determined according to the reverse mapping mechanism. The initial lighting parameter threshold set for each lighting segment is corrected by using traffic flow changes, thereby obtaining a lighting parameter threshold set that meets the requirements of safe driving.

[0032] In practical implementation, firstly, the visual reaction time and vehicle speed in driving visual characteristics can be coupled and analyzed. That is, based on the vehicle speed, the reaction distance generated by the driver within the visual reaction time is calculated according to the vehicle kinematics principle. This is the distance the vehicle continues to travel before the driver recognizes the target and makes a braking decision. The reaction distance is obtained by multiplying the vehicle speed by the visual reaction time. Then, a braking distance calculation model is constructed based on the vehicle braking performance parameters. By setting or estimating the braking deceleration, the distance required for the vehicle to come to a complete stop from the start of braking is calculated. Thus, the reaction distance and the braking distance can be superimposed to obtain the safe stopping distance under the current operating conditions. This safe stopping distance reflects the minimum space range required for the driver to complete a safe stop under the current lighting and traffic conditions. Then, a safety constraint criterion can be constructed based on the safe parking distance and the corrected visual distance estimate. This involves correlation analysis between the safe parking distance and the corrected visual distance estimate to determine whether the current visual capability meets safety requirements. When the corrected visual distance estimate is less than the safe parking distance, it indicates that the current lighting conditions are insufficient to support safe driving, requiring compensation through lighting parameter adjustments. A visual comfort index is introduced as an adjustment factor and coupled with the safety constraint criterion. By establishing a correlation between comfort and changes in brightness and color temperature, the adjustment process, which is solely safety-oriented, is constrained and corrected, ensuring that the lighting... Adjusting the parameters can improve visibility while avoiding visual fatigue and glare caused by excessive brightness or unreasonable color temperature. Furthermore, a reverse mapping mechanism can be constructed to deduce lighting parameters from the visibility distance. This reverse mapping mechanism focuses on improving the contrast between the target object and the background. By adjusting the lighting brightness level and optimizing the color temperature combination of the light source, the visibility distance is gradually increased and approaches or exceeds the safe parking distance. The reverse mapping process can be achieved through iterative calculation. In each round of calculation, the lighting parameters are incrementally adjusted based on the difference between the current visibility distance and the target safe distance, and the visibility effect is re-evaluated until the safety constraints are met.

[0033] In addition, in practical implementation, a safety braking margin can be introduced as a redundant constraint in the inversion process of the reverse mapping mechanism. That is, the original critical criterion that only satisfies the safe stopping distance is extended. On the basis of determining that the corrected recognition distance should not be less than the safe stopping distance, it is further required to be greater than the sum of the safe stopping distance and the safety braking margin, thus forming a more stringent safety boundary to cope with the uncertainties brought about by fluctuations in driving behavior, changes in road surface adhesion conditions, and sudden traffic disturbances. The safety braking margin can be embedded as an additional constraint in the reverse mapping calculation process, so that the inversion of lighting parameters not only meets the basic safety requirements, but also meets the requirements of redundant safety space. In practical implementation, by constraining and optimizing the brightness increase and color temperature adjustment range, the contrast between the target object and the background is further improved on the basis of meeting the minimum distinguishable threshold, thereby expanding the recognition distance. Then, the initial lighting parameter threshold set for each lighting segment of the tunnel can be determined based on the inverse mapping mechanism. That is, combined with the characteristics of the tunnel spatial segmentation, inverse mapping calculations are performed on the entrance segment, transition segment, middle segment, and exit segment respectively. The entrance segment focuses on increasing brightness to counteract the visual adaptation problem caused by strong light outside the tunnel. The transition segment uses a graded decreasing strategy to smooth brightness changes. The middle segment optimizes energy consumption while meeting redundancy safety constraints. The exit segment controls the difference in brightness between the inside and outside to avoid visual abrupt changes. Furthermore, in the inversion process of each segment, the combination of brightness and color temperature is continuously adjusted through iterative calculation so that the corrected recognition distance gradually approaches or exceeds the target sight distance including the safety braking margin. The range of lighting parameters corresponding to meeting the constraints is recorded, which is the initial lighting parameter threshold set for each lighting segment of the tunnel. Finally, traffic flow changes can be used to modify the initial lighting parameter threshold set for each lighting segment. That is, by establishing a mapping relationship between traffic density and risk level, the initial lighting parameter threshold set for each lighting segment can be dynamically adjusted. When traffic flow is high or operational complexity increases, the lighting threshold can be appropriately increased, thereby obtaining an optimized lighting parameter threshold set that takes into account both safety and comfort. Thus, the dataset composed of the optimized lighting parameter threshold sets of all lighting segments can be used as the lighting parameter threshold set that meets the requirements of safe driving.

[0034] It should be noted that by coupling driving visual characteristics with vehicle speed and introducing a safe stopping distance as a rigid constraint, the safety margin under the current lighting environment can be accurately determined, and the optimal range of brightness, color temperature and uniformity required to meet this safety margin can be derived in reverse. This avoids safety hazards caused by insufficient lighting, while reducing energy waste caused by excessive lighting. Furthermore, the lighting parameters under different working conditions can be updated in real time through a dynamic inversion mechanism, enabling the tunnel entrance, transition, middle and exit sections to automatically adjust lighting strategies according to environmental changes and traffic conditions. This achieves segmented differentiated control in space and continuous adaptive adjustment in time.

[0035] In step S104, lighting control strategies for each lighting segment in the tunnel are generated based on the set of lighting parameter thresholds and the current tunnel operating state vector, and the lighting fixtures for each lighting segment in the tunnel are controlled according to the corresponding lighting control strategies.

[0036] In some embodiments, generating lighting control strategies for each lighting segment in the tunnel based on the lighting parameter threshold set and the current tunnel operating state vector can be achieved in the following manner: The set of lighting parameter thresholds and the current tunnel operating state vector are correlated and mapped to generate a rigid constraint framework in the current tunnel environment. Based on the rigid constraint framework, a multi-objective optimization decision model with a three-layer hierarchical structure is constructed. The first layer of the multi-objective optimization decision model is the safety layer objective function, the second layer is the comfort layer objective function, and the third layer is the energy-saving layer objective function. The multi-objective optimization decision model is solved using the lexicographical order optimization method, and then outputs the lighting control strategy for each lighting segment in the tunnel.

[0037] In practical implementation, firstly, the set of lighting parameter thresholds can be correlated and mapped with the current tunnel operation state vector. This involves matching state variables such as external ambient brightness, ambient color temperature, weather level, traffic flow, and vehicle speed in the operation state vector with the corresponding brightness threshold, brightness difference threshold, and uniformity threshold for each lighting segment. This establishes a mapping relationship between state parameters and constraints, forming a rigid constraint framework under the current tunnel environment. This rigid constraint framework defines the minimum brightness level, allowable brightness variation range, and uniformity boundary that each lighting segment must meet under safety requirements. Then, using this rigid constraint framework as constraints, a three-layer hierarchical multi-objective optimization decision model can be constructed. The first layer is the safety layer objective function, which prioritizes ensuring the visual distance meets the safe stopping distance and maximizes the safe braking margin among all feasible solutions. The second layer is the comfort layer objective function, which minimizes brightness abrupt changes and color temperature deviations while satisfying the safety layer constraints, thereby reducing visual fatigue and improving visual continuity. The third layer is the energy-saving layer objective function, which further optimizes energy consumption while satisfying the objectives of the first two layers, achieving energy-saving operation by reducing unnecessary brightness output. Furthermore, the lexicographical optimization method can be used to solve the multi-objective optimization decision model. That is, the optimal solution is selected layer by layer in the order of safety first, comfort second, and energy saving last. In the optimization process of each layer, the optimal solution set of the previous layer is used as the constraint condition to ensure that the high-priority objective is not destroyed by the low-priority objective. After completing the multi-layer optimization solution, the optimal brightness value, color temperature value and its change gradient of each lighting segment under the current working conditions are obtained, and they are transformed into lighting control strategies for each lighting segment.

[0038] In some embodiments, the lighting fixtures in each lighting section of the tunnel can be controlled according to the corresponding lighting control strategy in the following ways: For each lighting segment in the tunnel, a longitudinal lighting distribution curve is generated based on the corresponding lighting control strategy; The longitudinal lighting distribution curve is mapped to each specific physical lamp in the corresponding lighting segment, thereby generating digital control instructions for each lighting segment that can be directly identified and executed by the lamp driver power supply. The digital control instructions include entry segment control instructions, transition segment control instructions, intermediate segment control instructions and exit segment control instructions. The estimated time when the vehicle arrives at the tunnel entrance is recorded as the baseline time, and the estimated time when the vehicle arrives at the starting point of each lighting section is determined based on the starting mileage marker of each lighting section in the tunnel. At a preset lead time before the arrival of the reference time, the entrance section control command is issued to the entrance section and the approach lighting fixtures in front of the entrance section, and the vehicle tracking triggering mechanism is used to issue control commands to the transition section, intermediate section and exit section respectively, thereby realizing the control of the lighting fixtures in each lighting section.

[0039] In practical implementation, firstly, for each lighting segment in the tunnel, a longitudinal lighting distribution curve can be generated based on the corresponding lighting control strategy. This is achieved by spatially interpolating and smoothing the target brightness, color temperature, and their gradient changes in the lighting control strategy corresponding to each lighting segment, ensuring that the lighting parameters change continuously along the vehicle's direction of travel. This generates the longitudinal lighting distribution curve for that lighting segment, avoiding visual discomfort caused by sudden changes in brightness and color temperature. Then, the longitudinal lighting distribution curve can be mapped one-to-one with the spatial location of the actual lamps in the tunnel. Based on the mileage marker location of each lamp, its corresponding target brightness and color temperature values ​​are extracted and converted into control parameters recognizable by the lamp driver power supply. For example, brightness control is achieved by adjusting the output power, and color temperature adjustment is achieved by matching different color temperature light source channels. This generates digital control commands for that lighting segment that can be directly recognized and executed by the lamp driver power supply. Through this method, digital control commands covering the entrance, transition, middle, and exit segments can be generated. These digital control commands include entrance segment control commands, transition segment control commands, middle segment control commands, and exit segment control commands. Furthermore, the estimated arrival time of the vehicle at the tunnel entrance can be recorded as the baseline time. The estimated arrival time of the vehicle at the starting point of each lighting section can be determined based on the mileage markers of each lighting segment within the tunnel, thus establishing a temporal and spatial correspondence. A predetermined lead time before the baseline time is reached, entrance section control commands are prioritized for the lighting fixtures of the entrance section and its leading ramps, ensuring the entrance area reaches the target lighting state in advance, guaranteeing visual adaptation for the vehicle upon entering the tunnel. Finally, a vehicle-tracking triggering mechanism can be employed to dynamically determine the vehicle's approach time to each lighting segment based on its real-time position. Before the vehicle enters the corresponding lighting segment, corresponding control commands are sequentially issued to the transition, intermediate, and exit section lights, allowing the lighting state to be dynamically updated as the vehicle moves. This enables precise, continuous, and adaptive control of the lighting fixtures in each tunnel lighting segment, ensuring visual safety and comfort during driving.

[0040] It should be noted that by coupling the set of lighting parameter thresholds with the operating state vector, the target brightness, color temperature, and uniformity range of each lighting segment under the current operating conditions can be accurately determined, and further transformed into an executable control strategy. This avoids the visual risks caused by insufficient lighting and the energy waste caused by excessive lighting. At the same time, the segmented control mechanism can implement differentiated control for different visual needs of the entrance, transition, middle, and exit segments, effectively mitigating the black hole and white hole effects, improving the continuity of driver visual adaptation, and thus achieving synergistic optimization of tunnel lighting in terms of safety, comfort, and energy saving.

[0041] Therefore, this invention firstly, by fusing and analyzing multi-source data and further mapping it to driving visual characteristics, the originally scattered and single-dimensional environmental and traffic information can be transformed into a comprehensive evaluation result centered on driver visual perception, thus accurately reflecting the driver's essential visual needs under different operating conditions. Secondly, by coupling driving visual characteristics with vehicle speed and introducing a safe stopping distance as a rigid constraint, the safety margin under the current lighting environment can be accurately determined, and the optimal brightness, color temperature, and uniformity range required to meet this safety margin can be derived in reverse. Finally, by coupling the set of lighting parameter thresholds with the operating state vector, the target brightness, color temperature, and uniformity range of each lighting segment under the current operating conditions can be accurately determined, and further transformed into an executable control strategy. Through a segmented control mechanism, differentiated control is implemented for the different visual needs of the entrance segment, transition segment, middle segment, and exit segment, effectively mitigating the black hole effect and white hole effect, improving the continuity of driver visual adaptation, and thus achieving synergistic optimization of tunnel lighting in terms of safety, comfort, and energy saving.

[0042] In summary, the technical solution adopted in this invention can perform dynamic inversion based on multi-source data and driving visual safety to generate a segmented adaptive control strategy, thereby achieving both safety and comfort in tunnel lighting.

[0043] Furthermore, in another aspect of the present invention, in some embodiments, the present invention provides an adaptive control system for tunnel lighting based on multi-source data fusion, referring to... Figure 4 The figure is a schematic diagram of a tunnel lighting adaptive control system based on multi-source data fusion according to some embodiments of the present invention. The tunnel lighting adaptive control system based on multi-source data fusion includes: The multi-source sensing module 201 is used to perform multi-source sensing of the environment inside and outside the tunnel and generate the original multi-source dataset of the current tunnel environment. Feature extraction module 202 is used to perform driving vision assessment based on the original multi-source dataset to obtain driving vision features in the current tunnel environment; The parameter inversion module 203 is used to perform inversion based on the relationship between the driving visual characteristics and the vehicle operating speed, as well as the safe parking distance constraint, to obtain a set of lighting parameter thresholds that meet the requirements of safe driving. The lighting control module 204 is used to generate lighting control strategies for each lighting segment in the tunnel based on the lighting parameter threshold set and the current tunnel operation state vector, and to control the lighting fixtures of each lighting segment in the tunnel according to the corresponding lighting control strategies.

[0044] In addition, the present invention also provides a computer device, the computer device including a memory and a processor, the memory storing code, the processor being configured to acquire the code and execute the above-described adaptive control method for tunnel lighting based on multi-source data fusion.

[0045] In some embodiments, reference Figure 5 The figure is a schematic diagram of the structure of a computer device implementing a tunnel lighting adaptive control method based on multi-source data fusion, according to some embodiments of the present invention. The tunnel lighting adaptive control method based on multi-source data fusion in the above embodiments can be implemented through... Figure 5 The computer device shown is used to implement this, and the computer device includes at least one processor 301, a communication bus 302, a memory 303, and at least one communication interface 304.

[0046] The processor 301 can be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more devices used to control the execution of the tunnel lighting adaptive control method based on multi-source data fusion in this invention.

[0047] The communication bus 302 can be used to transmit information between the aforementioned components.

[0048] The memory 303 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 303 may exist independently and be connected to the processor 301 via the communication bus 302. The memory 303 may also be integrated with the processor 301.

[0049] The memory 303 stores program code for executing the present invention, and its execution is controlled by the processor 301. The processor 301 executes the program code stored in the memory 303. The program code may include one or more software modules. In the above embodiments, the determination of the tunnel lighting adaptive control method based on multi-source data fusion can be achieved by the processor 301 and one or more software modules in the program code in the memory 303.

[0050] Communication interface 304 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0051] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0052] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This embodiment of the invention does not limit the type of computer device.

[0053] In addition, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described adaptive control method for tunnel lighting based on multi-source data fusion.

[0054] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0055] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for adaptive regulation of tunnel lighting based on multi-source data fusion, characterized in that, Includes the following steps: Multi-source sensing of the environment inside and outside the tunnel is performed to generate the original multi-source dataset of the current tunnel environment; Based on the original multi-source dataset, driving vision assessment is performed to obtain driving vision features in the current tunnel environment; Based on the relationship between the driving visual characteristics and the vehicle speed, and the safety parking distance constraint, an inversion is performed to obtain a set of lighting parameter thresholds that meet the requirements of safe driving. Based on the set of lighting parameter thresholds and the current tunnel operating state vector, a lighting control strategy for each lighting segment in the tunnel is generated, and the lighting fixtures for each lighting segment in the tunnel are controlled according to the corresponding lighting control strategy. Specifically, the driving visual recognition features obtained from the driving visual recognition evaluation based on the original multi-source dataset in the current tunnel environment include: Multidimensional feature extraction is performed on the original multi-source dataset to obtain the current tunnel operation status vector. The operation status vector includes the brightness and color temperature of the environment inside and outside the tunnel, weather level, traffic flow and vehicle speed features. The operation status vector is input into a pre-constructed driving vision performance model to output the driving vision features in the current tunnel environment. Specifically, inputting the operating state vector into a pre-constructed driving vision performance model, and then outputting the driving vision features in the current tunnel environment, includes: The components of the operating state vector are structurally analyzed to generate an input feature set for visual recognition analysis. This input feature set is then input into a pre-constructed driving visual recognition performance model, which is a coupled physical model based on human visual perception mechanisms and a data-driven model trained on experimental data. Based on the physical model, the contrast between the target object and the background is determined according to the principle of brightness and contrast, and the minimum discernible contrast threshold under the current lighting environment is determined by combining the contrast sensitivity characteristics of the human eye, thereby obtaining an initial visual recognition distance estimate. This initial visual recognition distance estimate, along with the operating state vector, is then used to generate an input feature set for visual recognition analysis. Traffic flow and vehicle speed features from the state vector are input into the data-driven model. The initial visual distance estimate is corrected through a nonlinear mapping relationship. Based on the corrected visual distance estimate and vehicle speed features, the driver's visual reaction time and safe braking margin are determined. By introducing a visual comfort evaluation function, the degree of visual fatigue under different brightness and color temperature combinations is quantitatively evaluated to obtain a visual comfort index. The corrected visual distance estimate, the visual comfort index, the visual reaction time, and the safe braking margin are integrated to output the driving visual characteristics in the current tunnel environment. Specifically, the set of lighting parameter thresholds that meet safe driving requirements is obtained by inversion based on the relationship between the driving visual characteristics and the vehicle speed, as well as the safe parking distance constraint. This set includes: A coupled analysis is performed on the visual reaction time and vehicle speed in the driving visual characteristics to obtain the safe stopping distance in the current tunnel environment. A safety constraint criterion is constructed based on the safe stopping distance and the corrected visual distance estimate, and a visual comfort index is introduced as an adjustment factor to establish a reverse mapping mechanism from visual distance to lighting parameters. A safe braking margin is introduced as a redundancy constraint during the inversion process of the reverse mapping mechanism, and the initial lighting parameter threshold set for each lighting segment of the tunnel is determined according to the reverse mapping mechanism. Traffic flow changes are used to correct the initial lighting parameter threshold set for each lighting segment, thereby obtaining a lighting parameter threshold set that meets the requirements of safe driving.

2. The adaptive control method for tunnel lighting based on multi-source data fusion according to claim 1, characterized in that, Multi-source sensing of the environment inside and outside the tunnel to generate the original multi-source dataset of the current tunnel environment specifically includes: Ambient light sensing units and meteorological monitoring units are deployed in typical impact areas outside the tunnel to acquire ambient light data and meteorological data outside the tunnel. The ambient light data outside the tunnel includes ambient brightness and ambient color temperature, and the meteorological data includes weather type and visibility. The tunnel lighting sensing unit is synchronously triggered, and multi-point brightness sampling devices are deployed in the entrance section, transition section, middle section and exit section to collect the ambient light data inside the tunnel. The ambient light data inside the tunnel includes the brightness information of each lighting section in the tunnel. The traffic operation detection unit is invoked synchronously to acquire traffic operation data reflecting the traffic operation status through video recognition and radar detection. The traffic operation data includes traffic flow and vehicle speed. The original multi-source dataset of the current tunnel environment consists of ambient light data outside the tunnel, meteorological data, ambient light data inside the tunnel, and traffic operation data.

3. The adaptive control method for tunnel lighting based on multi-source data fusion according to claim 1, characterized in that, The specific methods for generating lighting control strategies for each lighting segment in the tunnel based on the set of lighting parameter thresholds and the current tunnel operating state vector include: The set of lighting parameter thresholds and the current tunnel operating state vector are correlated and mapped to generate a rigid constraint framework in the current tunnel environment. Based on the rigid constraint framework, a multi-objective optimization decision model with a three-layer hierarchical structure is constructed. The first layer of the multi-objective optimization decision model is the safety layer objective function, the second layer is the comfort layer objective function, and the third layer is the energy-saving layer objective function. The multi-objective optimization decision model is solved using the lexicographical order optimization method, and then outputs the lighting control strategy for each lighting segment in the tunnel.

4. The adaptive control method for tunnel lighting based on multi-source data fusion according to claim 1, characterized in that, The lighting fixtures in each lighting section of the tunnel are adjusted according to the corresponding lighting control strategy, specifically including: For each lighting segment in the tunnel, a longitudinal lighting distribution curve is generated based on the corresponding lighting control strategy; The longitudinal lighting distribution curve is mapped to each specific physical lamp in the corresponding lighting segment, thereby generating digital control instructions for each lighting segment that can be directly identified and executed by the lamp driver power supply. The digital control instructions include entry segment control instructions, transition segment control instructions, intermediate segment control instructions and exit segment control instructions. The estimated time when the vehicle arrives at the tunnel entrance is recorded as the baseline time, and the estimated time when the vehicle arrives at the starting point of each lighting section is determined based on the starting mileage marker of each lighting section in the tunnel. At a preset lead time before the arrival of the reference time, the entrance section control command is issued to the entrance section and the approach lighting fixtures in front of the entrance section, and the vehicle tracking triggering mechanism is used to issue control commands to the transition section, intermediate section and exit section respectively, thereby realizing the control of the lighting fixtures in each lighting section.

5. A tunnel lighting adaptive control system based on multi-source data fusion, used to execute the tunnel lighting adaptive control method based on multi-source data fusion as described in any one of claims 1 to 4, characterized in that, include: The multi-source sensing module is used to perform multi-source sensing of the environment inside and outside the tunnel and generate the original multi-source dataset of the current tunnel environment. The feature extraction module is used to perform driving vision assessment based on the original multi-source dataset to obtain driving vision features in the current tunnel environment; The parameter inversion module is used to perform inversion based on the relationship between the driving visual characteristics and the vehicle operating speed, as well as the safe parking distance constraint, to obtain a set of lighting parameter thresholds that meet the requirements of safe driving. The lighting control module is used to generate lighting control strategies for each lighting segment in the tunnel based on the set of lighting parameter thresholds and the current tunnel operating state vector, and to control the lighting fixtures of each lighting segment in the tunnel according to the corresponding lighting control strategies.

6. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing code, and the processor being configured to retrieve the code and execute the tunnel lighting adaptive control method based on multi-source data fusion as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the tunnel lighting adaptive control method based on multi-source data fusion as described in any one of claims 1 to 4.

Images