An intelligent control system for tunnel lights

Abstract

The present invention discloses an intelligent control system for tunnel lights, which relates to the technical field of tunnel light control. Through six modules of data acquisition, light intensity test, brightness maintenance evaluation, accuracy analysis, accuracy classification, and dynamic adjustment, the accuracy and stability of the brightness adjustment of the system in case of emergencies and peak traffic flow are comprehensively improved. By collecting real-time data on light intensity and traffic flow changes, simulating abnormal scenarios to test the brightness response of the control system, identifying and evaluating the accuracy of the algorithm in emergencies, and then predictively optimizing the response parameters through the dynamic adjustment module, problems such as misjudgment, instability, and high energy consumption of existing algorithms under extreme conditions are solved. At the same time, the adaptability of the system to abnormal environments is significantly improved, the stable, safe, and energy-saving lighting in the tunnel is realized, and the driving environment is made more stable and reliable.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel light control, and particularly to an intelligent control system for tunnel lights. Background Art

[0002] The intelligent control of tunnel lights refers to the automatic adjustment of the lighting system in the tunnel through intelligent technology to improve safety and achieve energy conservation. The intelligent control system uses devices such as light sensors and traffic flow sensors to continuously monitor data such as the light intensity inside and outside the tunnel and traffic flow in real time, and automatically adjusts the brightness of the tunnel lights according to the actual situation. The intelligent control system can flexibly adjust the lighting brightness according to the change of traffic flow in the tunnel, increasing the brightness when the traffic flow is large and decreasing the brightness when the traffic flow is low to save energy.

[0003] The existing technology has the following deficiencies:

[0004] The tunnel intelligent control system may use complex algorithms to adjust the brightness. However, if the algorithm does not fully consider abnormal situations (such as sudden large-scale traffic flow or extreme weather conditions), the system may make misjudgments. In extreme cases, the failure of the algorithm may cause the brightness to be adjusted irregularly between low, medium, and high brightness, creating a chaotic visual environment for the driver and making the driving experience extremely unstable and dangerous. In addition, if the algorithm cannot resume normal control, it may also cause the system to maintain a very high brightness during non-peak periods, resulting in excessive energy consumption and system loss. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent control system for tunnel lights to solve the deficiencies in the background art.

[0006] To achieve the above purpose, the present invention provides the following technical solution: An intelligent control system for tunnel lights, including a data acquisition module, a light intensity test module, a brightness maintenance evaluation module, an accuracy analysis module, an accuracy division module, and a dynamic adjustment module;

[0007] Data acquisition module: By installing high-precision light sensors and high-speed traffic flow sensors at the tunnel entrance and exit, it continuously monitors and acquires light intensity change data under multiple different weather conditions and traffic flow mutation data within several time periods, and generates corresponding data sets;

[0008] Light intensity test module: Based on abnormal events in the tunnel, it sets test scenarios, records the brightness response stability of the control system to each light intensity change rate, and evaluates the real-time response ability of the algorithm under sudden light changes;

[0009] Brightness maintenance evaluation module: Using traffic flow mutation data to simulate a traffic flow surge scenario, it observes the brightness adjustment amplitude of the control system and evaluates the brightness maintenance ability of the control system under continuous peak traffic;

[0010] Accuracy analysis module: Comprehensively analyze the real-time response ability of the algorithm under sudden light changes and the brightness maintenance ability of the control system under continuous peak traffic, and determine the accuracy of the algorithm in adjusting the tunnel lamp brightness when dealing with emergencies according to the analysis results;

[0011] Accuracy classification module: Classify the accuracy of the algorithm in adjusting the tunnel lamp brightness when dealing with emergencies into different levels, classify it into accurate adjustment, incomplete accurate adjustment, and inaccurate adjustment, and perform corresponding processing;

[0012] Dynamic adjustment module: For incomplete accurate adjustment, predict the abnormal degree of the accuracy of the algorithm in adjusting the tunnel lamp brightness when dealing with emergencies within a fixed time period, and dynamically adjust the trigger threshold and adjustment rate of the brightness response in the algorithm according to the prediction results.

[0013] Preferably, in the light intensity test module, after analyzing the response time between when the algorithm in the control system detects a light intensity change and the start of brightness adjustment, generate a brightness response time abnormality index. The method for obtaining the brightness response time abnormality index is as follows:

[0014] Set the normal brightness response time T to follow a normal distribution , where: μ is the average value of the brightness response time, is the variance of the brightness response time. Based on historical data, calculate the average value and variance of the normal brightness response time, and obtain μ and respectively, for constructing the prior probability , and the expression is: ; During the actual monitoring process, record the actual brightness response time of the system each time, denoted as ; Define the collected brightness response time to fluctuate normally under normal conditions , so the likelihood function of the current observed value is The calculation expression is: ; Based on Bayes' theorem, calculate the posterior probability that the brightness response time belongs to the normal range , that is: ; Among them, P(normal) is the prior probability of the control system under normal conditions, and P(Tobs) is the marginal probability of the brightness response time in all possible cases: ; Among them represents the observation probability in the abnormal case, and P(abnormal) is the prior probability of the control system under abnormal conditions; Convert the posterior probability into a brightness response time abnormality index, and the expression is: ; In the formula, SQ is the brightness response time abnormality index.

[0015] Preferably, in the brightness maintenance evaluation module, after analyzing the fluctuation of the brightness value under peak traffic, a peak brightness fluctuation index is generated. The method for obtaining the peak brightness fluctuation index is as follows:

[0016] During the tunnel peak period, collect the time series data of brightness adjustment to form a sequence of brightness values L(t). Convert the brightness time series data L(t) to the frequency domain to obtain the frequency components. The expression is: ; where: F(f) is the Fourier transform coefficient corresponding to the frequency f, representing the signal intensity at the frequency f, N is the number of sampling points of the time series, L(t) is the brightness value at the t-th moment, is the kernel function of the Fourier transform, and i is the imaginary unit;

[0017] Calculate the amplitude of each frequency component , representing the intensity of brightness fluctuation at this frequency: ; where, is the real part of X(f), is the imaginary part High-frequency component extraction: Divide the spectrum into low-frequency and high-frequency components. The high-frequency components represent the rapid fluctuations of brightness. Select the high-frequency part above the frequency threshold and calculate its corresponding amplitude value. Accumulate the amplitude values of the high-frequency part to form the peak brightness fluctuation index. The expression is: where: is the total amplitude of the high-frequency part, is the total amplitude of the entire spectrum, and GH is the peak brightness fluctuation index.

[0018] Preferably, in the accuracy analysis module, convert the brightness response time anomaly index and the peak brightness fluctuation index into feature vectors, and use the feature vectors as the input of the machine learning model. The machine learning model takes the accuracy value label of predicting the brightness adjustment of tunnel lights by each group of feature vector prediction algorithms in emergencies as the prediction target, and takes minimizing the sum of prediction errors of the accuracy value labels of all algorithms for adjusting the brightness of tunnel lights in emergencies as the training target. Train the machine learning model until the sum of prediction errors reaches convergence and then stop the model training. Determine the accuracy value of the algorithm for adjusting the brightness of tunnel lights in emergencies according to the model output result. Among them, the machine learning model is a polynomial regression model.

[0019] Preferably, in the accuracy division module, compare the obtained accuracy value of the algorithm for adjusting the brightness of tunnel lights in emergencies with the gradient standard threshold. The gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold. Compare the accuracy value of the algorithm for adjusting the brightness of tunnel lights in emergencies with the first standard threshold and the second standard threshold respectively;

[0020] If the accuracy value of the algorithm for adjusting the brightness of tunnel lights in an emergency is greater than the second standard threshold, it indicates that the algorithm has high accuracy in adjusting the brightness of tunnel lights in an emergency. At this time, a high-accuracy adjustment signal is generated and classified into the accuracy adjustment category. The algorithm can accurately adjust the brightness without further adjustment;

[0021] If the accuracy value of the algorithm for adjusting the brightness of tunnel lights in an emergency is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it indicates that the algorithm has medium accuracy in adjusting the brightness of tunnel lights in an emergency. At this time, a medium-accuracy adjustment signal is generated and classified into the incomplete accuracy adjustment category. The algorithm can partially respond to sudden changes but accurately. Record and optimize the algorithm parameters to improve future response capabilities;

[0022] If the accuracy value of the algorithm for adjusting the brightness of tunnel lights in an emergency is less than the first standard threshold, it indicates that the algorithm has low accuracy in adjusting the brightness of tunnel lights in an emergency. At this time, a low-accuracy adjustment signal is generated and classified into the inaccuracy adjustment category. The algorithm does not respond accurately enough in an emergency, generates an alarm signal, and performs parameter optimization.

[0023] Preferably, in the dynamic adjustment module, for incomplete accuracy adjustment, predict the degree of abnormality of the accuracy of the algorithm for adjusting the brightness of tunnel lights in response to emergencies within a fixed time period, and dynamically adjust the trigger threshold and adjustment rate of brightness response in the algorithm according to the prediction results. Specifically:

[0024] Set a time period W. For incomplete accuracy adjustment, that is, the accuracy value of the algorithm for adjusting the brightness of tunnel lights in an emergency generated within a fixed time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold. Collect the accuracy values of the algorithm for adjusting the brightness of tunnel lights in an emergency generated in subsequent fixed time periods that are greater than or equal to the first standard threshold and less than or equal to the second standard threshold, and construct a corresponding data set, and calculate the standard deviation of the data set as the abnormality index of the accuracy of the algorithm for adjusting the brightness of tunnel lights in response to emergencies within a fixed time period, marked as 。

[0025] Preferably, according to the predicted degree of abnormality , dynamically adjust the trigger threshold of brightness response. The trigger threshold adjustment formula: ; where: is the trigger threshold of brightness response after dynamic adjustment, is the basic trigger threshold, that is, the brightness response trigger standard under normal circumstances. k is an adjustment coefficient used to control the influence of the degree of abnormality on the trigger threshold; Dynamically adjust the rate of brightness change according to the abnormal prediction value . The brightness adjustment rate formula: ; where: is the adjusted brightness change rate is the basic brightness adjustment rate, that is, the brightness adjustment speed under normal circumstances. m is the rate adjustment coefficient, which controls the influence of the predicted anomaly degree on the adjustment rate.

[0026] In the above technical solution, the technical effects and advantages provided by the present invention are as follows:

[0027] 1. By integrating data acquisition, light intensity testing, brightness maintenance evaluation, accuracy analysis, accuracy classification, and dynamic adjustment modules, the intelligent control system of tunnel lights can flexibly and accurately respond to emergencies and changes in peak traffic flow. With the help of high-precision sensors and simulated test scenarios, the system monitors and records the changes in light intensity and traffic flow in real time, generates the brightness response time anomaly index and the peak brightness fluctuation index, so as to detect and evaluate the reaction ability of the algorithm under different environmental conditions. Through the accuracy analysis and classification module, the system can automatically identify the performance of the algorithm in emergencies and classify and process it according to the preset threshold to ensure appropriate brightness adjustment effects.

[0028] 2. The present invention uses the dynamic adjustment module to optimize the brightness response trigger threshold and adjustment rate of the algorithm in real time for the situation of incomplete accuracy adjustment, improving the response sensitivity and stability of the system. This design not only effectively solves the problems of misjudgment, frequent adjustment, and excessive energy consumption of traditional algorithms under extreme conditions, but also significantly improves the driving safety and energy utilization efficiency in the tunnel, ensuring a stable lighting effect in various emergencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0030] Figure 1 is the system module diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0032] Embodiment, please refer to Figure 1As shown in the figure, an intelligent control system for tunnel lights in this embodiment includes a data acquisition module, a light intensity test module, a brightness maintenance evaluation module, an accuracy analysis module, an accuracy classification module, and a dynamic adjustment module;

[0033] Data acquisition module: By installing high-precision light sensors and high-speed traffic flow sensors at the tunnel entrance and exit, it monitors and collects light intensity change data under multiple different weather conditions and traffic flow mutation data within several time periods in real time, and generates corresponding data sets;

[0034] Light intensity test module: Based on abnormal events in the tunnel, it sets test scenarios, records the brightness response stability of the control system to the light intensity change rate, and evaluates the real-time response ability of the algorithm under sudden light changes;

[0035] Brightness maintenance evaluation module: Uses traffic flow mutation data to simulate a traffic flow surge scenario, observes the brightness adjustment range of the control system, and evaluates the brightness maintenance ability of the control system under continuous peak traffic flow;

[0036] Accuracy analysis module: Comprehensively analyzes the real-time response ability of the algorithm under sudden light changes and the brightness maintenance ability of the control system under continuous peak traffic flow, and determines the accuracy of the algorithm in adjusting the brightness of tunnel lights when dealing with emergencies according to the analysis results;

[0037] Accuracy classification module: Classifies the accuracy of the algorithm in adjusting the brightness of tunnel lights when dealing with emergencies into different levels, divides it into accurate adjustment, incomplete accurate adjustment, and inaccurate adjustment, and performs corresponding processing;

[0038] Dynamic adjustment module: For incomplete accurate adjustment, it predicts the abnormal degree of the accuracy of the algorithm in adjusting the brightness of tunnel lights when dealing with emergencies within a fixed time period, and dynamically adjusts the trigger threshold and adjustment rate of the brightness response in the algorithm according to the prediction results.

[0039] Among them, in the data acquisition module, high-precision light sensors suitable for the tunnel environment are selected, which are required to quickly and accurately detect light intensity changes under different weather and lighting conditions. High-speed traffic flow sensors are selected, which are required to accurately record parameters such as the number of vehicles per second and the driving speed, and are applicable to scenarios of sudden traffic flow surges. The light sensors are installed at key positions at the tunnel entrance and exit respectively to ensure that the light intensity changes inside and outside the tunnel can be comprehensively monitored. The traffic flow sensors are installed at the tunnel entrance (or at the exit when necessary) to ensure that traffic flow data can be collected in real time when vehicles enter and leave the tunnel.

[0040] Calibrate the light sensor and traffic flow sensor using a standard light source and known traffic flow data to ensure data accuracy. During the calibration process, record the response time and accuracy range of the sensors to ensure that fast-changing data can be accurately captured during actual use. Conduct preliminary tests under different time periods and weather conditions (such as sunny days, cloudy days, rainy days, nights) to check whether the sensors can continuously and stably collect data. Monitor the stability of data transmission to ensure that data can be reliably transmitted to the data acquisition module.

[0041] Set the light sensor and traffic flow sensor to high-frequency acquisition mode. The light intensity change is collected once per second, and the traffic flow data is updated according to the real-time data of vehicle passing to capture subtle changes. Set the data acquisition period, such as continuous acquisition for 24 hours or a week, to ensure that different weather conditions (sunny days, rainy days, foggy days) and traffic states (morning and evening rush hours, off-peak periods, etc.) can be covered. Transmit the collected light intensity change data and traffic flow data to the database of the central control system in real time. The database is backed up regularly to prevent data loss and ensure data integrity.

[0042] Monitor and record sudden changes in weather, such as a sunny day suddenly turning into a heavy rainstorm, day-night alternation, a foggy day with reduced visibility, etc. Mark the light intensity change rate during sudden weather events. For example, the light intensity change rate increases significantly during a heavy rainstorm, and evaluate the system's ability to handle drastic light intensity changes based on this data.

[0043] Record sudden changes in traffic flow, including a sharp increase or decrease in traffic flow. For example, when a traffic accident causes a sharp increase in diverted traffic flow, mark the time point, flow increase amplitude, and duration of the event. Mark the time periods with peak traffic flow and significant traffic flow fluctuations for subsequent algorithm testing and optimization.

[0044] Integrate the light intensity change data collected under different weather and time periods, and classify and store them according to weather types (such as sunny days, rainy days, foggy days, nights) and time periods (morning rush hour, evening rush hour, off-peak). Calculate the average light intensity change rate, maximum change value, and change trend in each case to form a complete light intensity change data set. Classify the collected traffic flow data according to time periods, such as morning rush hour, evening rush hour, flat peak period, and special situations such as sudden traffic flow changes (such as sudden increase or decrease). Calculate the average traffic flow, traffic flow peak, and volatility in each time period to form a detailed traffic flow change data set. Combine the light intensity change data and traffic flow data to generate a multi-dimensional data set for testing tunnel lighting control under different weather, time periods, and traffic flow states. Mark the sudden events separately to ensure that they can be used for simulation testing and accuracy evaluation of subsequent algorithms.

[0045] Filter out the outliers caused by signal interference or sensor errors in data collection to ensure the accuracy and continuity of the data set. Randomly select some data for verification to ensure consistency with the actual environmental conditions and confirm the validity and reliability of the data set.

[0046] In the light intensity test module, based on the actual environment of the tunnel, select common sudden light change conditions that affect the driver's line of sight as the test scenarios. Typical scenarios include: sudden heavy rain or thick fog: simulate the scenario of a sudden decrease in light intensity. Rapid conversion from sunny to cloudy: simulate the transition of light from bright to dim. Day-night alternation or contrast between light inside and outside the tunnel: simulate the situation of entering a dark area from a bright area.

[0047] According to the collected light intensity change data, set light change values at different rates. For example: slow change: the light intensity changes gradually within 1 minute. Medium rate change: the light intensity gradually decreases or increases within a few seconds. Sudden rapid change: the light intensity rapidly decreases within 1 second to simulate sudden weather changes. Establish rate levels: divide the light intensity change rate into three levels: "low speed", "medium speed", and "high speed" for precise evaluation of different change rates.

[0048] Gradually input the set light intensity rate change values into the control system to simulate a real light mutation scenario. At this time, the light sensor simulates the real light intensity change and transmits it to the control system, triggering the response of the system's brightness adjustment algorithm.

[0049] After analyzing the response time between the detection of light intensity change and the start of brightness adjustment in the control system algorithm, generate a brightness response time anomaly index. The method for obtaining the brightness response time anomaly index is as follows:

[0050] Set that the normal brightness response time T follows a normal distribution , where: μ is the average value of the brightness response time, is the variance of the brightness response time. Based on historical data, calculate the average value and variance of the normal brightness response time to obtain μ and , respectively, for constructing the prior probability , and the expression is: ; During the actual monitoring process, record the actual brightness response time of the system each time, denoted as ; Define that the fluctuation of the collected brightness response time follows a normal distribution under normal circumstances , so the likelihood function of the current observed value is and the calculation expression is: ; Based on Bayes' theorem, calculate the posterior probability that the brightness response time belongs to the normal range, that is: ; where P(normal) is the prior probability of the control system under normal conditions, which can usually be set to 1 (assuming the normal state is the default state), and P(Tobs) is the marginal probability of the brightness response time in all possible cases: ; where represents the observation probability in abnormal situations, which can generally be set to a certain fixed low probability value or estimated based on historical abnormal data, and P(abnormal) is the prior probability of the control system under abnormal conditions; the posterior probability is converted into a brightness response time anomaly index, and the expression is: ; in the formula, SQ is the brightness response time anomaly index.

[0051] The larger the brightness response time anomaly index, the more significantly the current brightness response time deviates from the normal range, indicating that the algorithm has a poor real-time response ability under sudden light changes. A high anomaly index means that the system fails to adjust the brightness in a timely manner, which may cause the driver to experience an uncomfortable visual environment under sudden light conditions and increase the traffic safety risk. Therefore, a higher anomaly index indicates that the algorithm has deficiencies in coping with sudden light changes and needs further optimization and adjustment.

[0052] On the contrary, the smaller the brightness response time anomaly index, the more the current brightness response time conforms to the normal range, indicating that the algorithm has a strong real-time response ability under sudden light changes. A low anomaly index means that the system can quickly and accurately adjust the lighting brightness to ensure the safety and comfort of the driver under different light conditions. At this time, the intelligent control ability of the system is effectively verified, and it can maintain a good lighting effect in case of emergencies, thus enhancing traffic safety.

[0053] In the brightness maintenance evaluation module, first, historical traffic flow data is obtained, including normal flow and sudden flow conditions. These data should cover traffic flow changes in different time periods and different weather conditions. Based on the historical data, multiple traffic flow surge scenarios are constructed for the evaluation of the control system. These scenarios should consider different sudden conditions, such as traffic accidents, special events, or weather changes.

[0054] Determine the start time, duration, and traffic flow increase rate of each simulation scenario. Set a threshold to distinguish between peak flow and normal flow states. Input the traffic flow mutation data into the control system and start real-time monitoring and adjustment. The system should automatically adjust the lighting brightness in the tunnel according to the input flow data. Observe the brightness adjustment amplitude of the control system in case of traffic flow surge in real time, and record the specific brightness value of each adjustment.

[0055] After analyzing the fluctuation of the brightness value under peak flow, a peak brightness fluctuation index is generated. The method for obtaining the peak brightness fluctuation index is:

[0056] During the tunnel peak period, the time series data of brightness adjustment is collected to form a set of brightness value sequences L(t). If there is an obvious trend item in the brightness data (such as continuous increase or decrease), the trend can be removed by difference or filtering methods to make the data more concentrated on the fluctuation component. The brightness time series data L(t) is converted to the frequency domain to obtain the frequency component, which is expressed as: ; Where: F(f) is the Fourier transform coefficient (complex form) corresponding to frequency f, representing the signal strength at frequency f, N is the number of sampling points in the time series, L(t) is the brightness value at time t, is the kernel function of Fourier transform, i is the imaginary unit;

[0057] Calculate the magnitude (amplitude) of each frequency component , which represents the intensity of brightness fluctuation at this frequency: ;in, is the real part of X(f), is the imaginary part of X(f). High-frequency component extraction: The spectrum is usually divided into low-frequency and high-frequency components, where the high-frequency components represent rapid fluctuations in brightness. Select a frequency threshold above The high-frequency part is calculated to calculate the corresponding amplitude value. The frequency threshold can be set according to the requirements of brightness fluctuation, such as the number of changes per second. The amplitude of the high-frequency part is accumulated to form a high peak brightness fluctuation index, which is expressed as: in: is the high frequency part (frequency above the threshold ) is the sum of the amplitudes of . is the total amplitude of the entire spectrum, and GH is the peak brightness fluctuation index.

[0058] The larger the peak brightness fluctuation index, the more dramatic the brightness fluctuation of the control system during peak traffic, and the more difficult it is to maintain the brightness level smoothly. This means that when the control system handles the brightness requirements of continuous peak traffic, frequent and excessive brightness adjustments may occur, resulting in unstable brightness. For drivers, such frequent brightness fluctuations may cause visual fatigue or interference, increasing driving risks. Therefore, a higher fluctuation index indicates that the system has a weaker ability to maintain brightness, and further optimization may be needed to improve the smoothness and consistency of brightness adjustment.

[0059] On the contrary, the smaller the peak brightness fluctuation index, the more stable the system can maintain the brightness level under peak traffic, with smaller adjustment amplitude and frequency. This means that the control system can quickly adapt and stably provide consistent lighting when facing continuous peak traffic, reducing unnecessary fluctuations. A low fluctuation index usually means that the control system has good brightness maintenance capabilities and can provide a stable and comfortable lighting environment during peak hours, which helps to improve the safety and comfort of tunnel driving.

[0060] Accuracy Analysis Module: Comprehensively analyze the real-time response ability of the algorithm under sudden light changes and the brightness maintenance ability of the control system under continuous peak traffic, and determine the accuracy of the algorithm in adjusting the tunnel light brightness during emergencies according to the analysis results.

[0061] Convert the brightness response time anomaly index and the peak brightness fluctuation index into feature vectors, use the feature vectors as the input of the machine learning model. The machine learning model takes predicting the accuracy value label of the algorithm in adjusting the tunnel light brightness during emergencies for each group of feature vectors as the prediction target, and takes minimizing the sum of the prediction errors of the accuracy value labels of all algorithms in adjusting the tunnel light brightness during emergencies as the training target, and train the machine learning model until the sum of the prediction errors reaches convergence and then stop the model training. Determine the accuracy value of the algorithm in adjusting the tunnel light brightness during emergencies according to the model output result, where the machine learning model is a polynomial regression model.

[0062] The method for obtaining the accuracy value of the algorithm in adjusting the tunnel light brightness during emergencies is: Obtain the corresponding function expression from the first feature vector training data of the trained machine learning model: ; In the formula, is the output function of the model, SQ is the brightness response time anomaly index, GH is the peak brightness fluctuation index, is the accuracy value of the algorithm in adjusting the tunnel light brightness during emergencies.

[0063] Accuracy Classification Module: Classify the accuracy of the algorithm in adjusting the tunnel light brightness during emergencies into different levels, classify it into accurate adjustment, incomplete accurate adjustment and inaccurate adjustment, and perform corresponding processing.

[0064] Compare the obtained accuracy value of the algorithm in adjusting the tunnel light brightness during emergencies with the gradient standard thresholds. The gradient standard thresholds include the first standard threshold and the second standard threshold, and the first standard threshold is less than the second standard threshold. Compare the accuracy value of the algorithm in adjusting the tunnel light brightness during emergencies with the first standard threshold and the second standard threshold respectively;

[0065] If the accuracy value of the algorithm in adjusting the tunnel light brightness during emergencies is greater than the second standard threshold, it indicates that the accuracy of the algorithm in adjusting the tunnel light brightness during emergencies is high. At this time, generate a high-accuracy adjustment signal and classify it into the accurate adjustment category. The algorithm can accurately adjust the brightness, and the system does not need further adjustment;

[0066] If the accuracy value of the algorithm for adjusting the brightness of tunnel lights during emergencies is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it indicates that the accuracy of the algorithm for adjusting the brightness of tunnel lights during emergencies is medium. At this time, a medium-accuracy adjustment signal is generated and classified into the incomplete-accuracy adjustment category. The algorithm can partially respond to sudden changes but is not accurate enough. The system can record and optimize the algorithm parameters to improve future response capabilities.

[0067] If the accuracy value of the algorithm for adjusting the brightness of tunnel lights during emergencies is less than the first standard threshold, it indicates that the accuracy of the algorithm for adjusting the brightness of tunnel lights during emergencies is low. At this time, a low-accuracy adjustment signal is generated and classified into the inaccuracy adjustment category. The algorithm does not respond accurately enough during emergencies. The system should generate an alarm signal and perform necessary parameter tuning, or enable a backup algorithm to ensure safety.

[0068] Dynamic adjustment module: For incomplete-accuracy adjustment, predict the degree of abnormality of the accuracy of the algorithm for adjusting the brightness of tunnel lights during emergencies within a fixed time period, and dynamically adjust the trigger threshold and adjustment rate of the brightness response in the algorithm according to the prediction result.

[0069] Set a time period W. For incomplete-accuracy adjustment, that is, the accuracy value of the algorithm for adjusting the brightness of tunnel lights during emergencies generated within a fixed time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold. Collect the accuracy values of the algorithm for adjusting the brightness of tunnel lights during emergencies that are greater than or equal to the first standard threshold and less than or equal to the second standard threshold generated in subsequent fixed time periods, and construct the corresponding data set, and calculate the standard deviation of the data set as the abnormality index of the accuracy of the algorithm for adjusting the brightness of tunnel lights during emergencies within a fixed time period, marked as 。

[0070] According to the predicted degree of abnormality , dynamically adjust the trigger threshold of the brightness response to improve the reaction sensitivity of the system or reduce unnecessary frequent adjustments. The trigger threshold adjustment formula: ; where: is the trigger threshold of the brightness response after dynamic adjustment, is the basic trigger threshold, that is, the brightness response trigger standard under normal circumstances. k is the adjustment coefficient used to control the influence of the degree of abnormality on the trigger threshold, usually set through experiments or historical data. When is higher, the predicted degree of abnormality of the system is large. To prevent instability caused by frequent brightness adjustments, the trigger threshold can be increased so that the system is not easily affected by small fluctuations; if is lower, the trigger threshold is close to the basic value to ensure normal response sensitivity.

[0071] According to the anomaly prediction value Dynamically adjust the rate of brightness change so that the system can adapt to the brightness change in emergencies more quickly or smoothly. The brightness adjustment rate formula is: Where: is the adjusted brightness change rate, which determines the speed of brightness adjustment. is the basic brightness adjustment rate, that is, the brightness adjustment speed under normal conditions. m is the rate adjustment coefficient, which controls the influence of the predicted anomaly degree on the adjustment rate and is set through historical data. When is high, it indicates that the light change brought by the emergency is more intense, and the brightness adjustment rate will be higher than the basic rate, enabling the system to respond more quickly to the light change. When is low, the brightness adjustment rate remains at the basic rate to ensure the stability of the adjustment.

[0072] In this embodiment, the data acquisition module installs high-precision sensors at the tunnel entrance and exit to monitor and collect light intensity change and traffic flow mutation data in real time, generating a data set. The light intensity test module records the brightness response of the system at different light intensity change rates by setting an abnormal event test scenario to evaluate the real-time response ability of the algorithm. The brightness maintenance evaluation module evaluates the brightness stability of the system under continuous peaks by simulating a sharp increase in traffic flow. The accuracy analysis module comprehensively analyzes the performance of the algorithm under sudden light changes and peak traffic to determine the accuracy of adjusting the brightness. The accuracy classification module classifies the accuracy into high, medium, and low categories and takes corresponding measures. The dynamic adjustment module dynamically adjusts the trigger threshold and adjustment rate of the brightness response according to the anomaly prediction of medium accuracy to optimize the sensitivity and stability of the system and ensure reliable lighting support under various emergency conditions.

[0073] The above formulas are all dimensionless and take their numerical calculations. The formula is obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0074] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, or a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0075] As described above, the specific implementation manners of the present application are only described, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all of them should be covered by the protection scope of the present application.

Claims

1. An intelligent control system for tunnel lights, characterized in that: It includes data acquisition module, light intensity test module, brightness maintenance evaluation module, accuracy analysis module, accuracy division module and dynamic adjustment module; Data acquisition module: by installing high-precision light sensors and high-speed traffic flow sensors at the entrance and exit of the tunnel, real-time monitoring and collection of light intensity change data under multiple different weather conditions and traffic flow mutation data in several time periods, and generating corresponding data sets; Light intensity test module: Set test scenarios based on abnormal events in the tunnel, record the brightness response stability of the control system to various light intensity change rates, and evaluate the algorithm's real-time response capabilities under sudden light changes; Brightness maintenance evaluation module: Use traffic flow mutation data to simulate traffic flow surge scenarios, observe the brightness adjustment range of the control system, and evaluate the control system's ability to maintain brightness under continuous peak traffic; Accuracy analysis module: Comprehensively analyze the algorithm's real-time response capability to sudden light changes and the control system's ability to maintain brightness under sustained peak traffic, and determine the accuracy of the algorithm in adjusting tunnel light brightness in response to emergencies based on the analysis results; Accuracy classification module: divides the accuracy of the algorithm in adjusting the brightness of tunnel lights in response to emergencies into different levels, including accuracy adjustment, incomplete accuracy adjustment and inaccuracy adjustment, and performs corresponding processing; Dynamic adjustment module: For incomplete accuracy adjustment, the abnormal degree of accuracy of the algorithm in adjusting the brightness of tunnel lights in response to emergencies within a fixed time period is predicted, and the trigger threshold and adjustment rate of the brightness response in the algorithm are dynamically adjusted according to the prediction results.

2. The intelligent control system for tunnel lights according to claim 1, characterized in that: In the light intensity test module, the response time between the detection of light intensity change and the start of brightness adjustment in the control system algorithm is analyzed to generate a brightness response time abnormality index. The brightness response time abnormality index is obtained as follows: Set the normal brightness response time T to follow the normal distribution , where: μ is the average value of the brightness response time, is the variance of the brightness response time. Based on historical data, the average and variance of the normal brightness response time are calculated to obtain μ and , used to construct the prior probability , the expression is: ; In the actual monitoring process, record the actual brightness response time of the system each time, denoted as ; Define the acquired brightness response time Under normal circumstances, fluctuations follow a normal distribution , so the current observation The likelihood function The calculation expression is: ; Based on Bayesian theorem, calculate the posterior probability that the brightness response time is within the normal range ,Right now: ; Where P(Normal) is the prior probability of the control system under normal conditions, and P(Tobs) is the marginal probability of the brightness response time under all possible conditions: ;in represents the observation probability under abnormal conditions, P(abnormal) is the prior probability of the control system under abnormal conditions; the posterior probability Converted into brightness response time abnormality index, the expression is: ; Where SQ is the brightness response time anomaly index.

3. The intelligent control system for tunnel lights according to claim 2, characterized in that: In the brightness maintenance evaluation module, the peak brightness fluctuation index is generated after analyzing the fluctuation of the brightness value under the peak flow rate. The method for obtaining the peak brightness fluctuation index is: During the tunnel peak period, the time series data of brightness adjustment is collected to form a set of brightness value sequences L(t). The brightness time series data L(t) is converted to the frequency domain to obtain the frequency component, which is expressed as: ; Where: X(f) is the Fourier transform coefficient corresponding to frequency f, representing the signal strength at frequency f, N is the number of sampling points in the time series, L(t) is the brightness value at time t, is the kernel function of Fourier transform, i is the imaginary unit; Calculate the amplitude of each frequency component , which represents the intensity of brightness fluctuation at this frequency: ;in, is the real part of X(f), is the imaginary part of X(f), high-frequency component extraction: the spectrum is divided into low-frequency and high-frequency components, where the high-frequency components represent rapid fluctuations in brightness, and high-frequency components above the frequency threshold are selected. The high-frequency part of , calculates its corresponding amplitude value, and accumulates the amplitude value of the high-frequency part to form the high peak brightness fluctuation index, which is expressed as: in: is the sum of the amplitudes of the high frequency parts, is the total amplitude of the entire spectrum, and GH is the peak brightness fluctuation index.

4. The intelligent control system for tunnel lights according to claim 3, characterized in that: In the accuracy analysis module, the brightness response time anomaly index and the peak brightness fluctuation index are converted into feature vectors, and the feature vectors are used as the input of the machine learning model. The machine learning model uses each set of feature vectors to predict the accuracy value label of the algorithm for adjusting the brightness of tunnel lights in emergencies as the prediction target, and takes minimizing the sum of the prediction errors of the accuracy value labels of all algorithms for adjusting the brightness of tunnel lights in emergencies as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The accuracy value of the algorithm for adjusting the brightness of tunnel lights in emergencies is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

5. The intelligent control system for tunnel lights according to claim 4, characterized in that: In the accuracy division module, the obtained accuracy value of the algorithm for adjusting the brightness of the tunnel light in an emergency is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the accuracy value of the algorithm for adjusting the brightness of the tunnel light in an emergency is compared with the first standard threshold and the second standard threshold respectively; If the accuracy value of the algorithm in adjusting the brightness of the tunnel lights in the emergency is greater than the second standard threshold, it means that the algorithm has high accuracy in adjusting the brightness of the tunnel lights in the emergency, and a high accuracy adjustment signal is generated and classified into an accuracy adjustment category, indicating that the algorithm can accurately adjust the brightness without further adjustment; If the accuracy value of the algorithm in adjusting the brightness of the tunnel lights in an emergency is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it means that the accuracy of the algorithm in adjusting the brightness of the tunnel lights in an emergency is moderate. At this time, a moderate accuracy adjustment signal is generated and classified as an incomplete accuracy adjustment category. The algorithm can partially respond to the sudden change, and the algorithm parameters are recorded and optimized to improve the future response capability. If the accuracy value of the algorithm in adjusting the brightness of the tunnel lights in an emergency is less than the first standard threshold, it means that the algorithm has low accuracy in adjusting the brightness of the tunnel lights in an emergency. At this time, a low-accuracy adjustment signal is generated and classified into an inaccuracy adjustment category. The algorithm does not respond accurately in the emergency, generates an alarm signal, and performs parameter tuning.

6. The intelligent control system for tunnel lights according to claim 1, characterized in that: In the dynamic adjustment module, for incomplete accuracy adjustment, the abnormal degree of accuracy of the algorithm in adjusting the brightness of tunnel lights in response to emergencies within a fixed time period is predicted, and the trigger threshold and adjustment rate of the brightness response in the algorithm are dynamically adjusted according to the prediction results, specifically: Set a time period W, for incomplete accuracy adjustment, that is, the accuracy value of the algorithm for adjusting the brightness of the tunnel lights in an emergency generated within a fixed time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, collect the accuracy values ​​of the algorithm for adjusting the brightness of the tunnel lights in an emergency generated within a subsequent fixed time period that are greater than or equal to the first standard threshold and less than or equal to the second standard threshold, and construct a corresponding data set, and calculate the standard deviation of the data set as the abnormal index of the accuracy of the algorithm for adjusting the brightness of the tunnel lights in a fixed time period when responding to emergencies, marked as .

7. The intelligent control system for tunnel lights according to claim 6, characterized in that: According to the predicted abnormality , dynamically adjust the trigger threshold of the brightness response, the trigger threshold adjustment formula is: ;in: is the brightness response trigger threshold after dynamic adjustment, is the basic trigger threshold, i.e., the brightness response trigger standard under normal circumstances, and k is the adjustment coefficient, which is used to control the influence of the abnormal degree on the trigger threshold; according to the abnormal prediction value Dynamically adjust the rate of brightness change. The brightness adjustment rate formula is: ;in: is the adjusted brightness change rate is the basic brightness adjustment rate, i.e., the brightness adjustment speed under normal circumstances, and m is the rate adjustment coefficient, which controls the impact of the predicted abnormality on the adjustment rate.

Images