Fog lamp induction warning control system

Abstract

A fog lamp induction warning control system comprises a data acquisition module, the data acquisition module comprises a vibration sensor, an acoustic sensor and a light sensor, and the vibration sensor is used for capturing vibration waveforms generated during vehicle collision and acquiring vibration data in real time; the signal processing module is used for processing signals of the data acquisition module, and the data processing module performs fusion by using weighted average; setting a threshold value to judge whether the fused features belong to abnormal conditions or not; the threshold value is determined through statistical analysis or experiment of historical data, and the output result exceeding the threshold value is output; the warning module comprises a fog lamp and a sound alarm arranged on the edge of the fog lamp; the invention aims to solve the problem that a secondary collision accident is easily caused because a traffic accident warning cannot be sent to a driver behind in time after a vehicle is collided.

Description

Technical Field

[0001] The invention belongs to the technical field of automation and relates to a fog lamp induction warning control system. Background Art

[0002] In the construction of modern transportation infrastructure, the fog lamp system is a key link to ensure road driving safety, especially on highways, mountain bends and other sections where low visibility is prone to occur. Fog lamps are indispensable. Existing fog lamps are usually arranged linearly along the left and right sides of the lane. This layout is designed to cover the driving area in all directions and provide sufficient light guidance for passing vehicles. In addition, each fog lamp is given an independent coding mark. These codes are connected to the road management system, which can accurately record the location of the fog lamp and the distance from the surrounding reference points, providing convenience for traffic scheduling, road maintenance and other tasks, and helping to improve the overall intelligent level of traffic operation.

[0003] However, with the increase in actual usage scenarios, the traditional fog lamp system has exposed serious defects when facing sudden traffic accidents. When a vehicle collides, the existing fog lamps only focus on static lighting and positioning functions, and cannot respond quickly and actively to warn the vehicle coming from behind. The driver behind approaches the accident scene at high speed without warning, and it is very easy to hit the vehicle in front because there is no time to brake or avoid measures, causing a secondary collision accident, resulting in more serious casualties and property losses, which greatly threatens road traffic safety.

[0004] In response to this urgent problem, the patent number is CN202010900399.4, and the name of the patent is road warning system. The core is to install impact sensing devices at intervals on the road guardrails. Once an impact force exceeding the threshold is detected, the nearby warning lights are activated to light up continuously to alert the vehicles behind. The advantage of applying this method to the production process is that it utilizes existing road facilities for accident monitoring, is relatively easy to install, and can provide intuitive warnings within a certain area; however, its shortcomings cannot be ignored. On the one hand, the warning lights and fog light systems are independent of each other, which can easily cause visual confusion to the drivers behind and distract their attention from dangerous areas; on the other hand, the simple continuous lighting method has low warning recognition in complex lighting environments such as strong light, backlight or foggy days, and it is difficult to ensure that the drivers behind receive danger signals in a timely and accurate manner. Comprehensive considerations are taken to meet the safety needs in production practice. Summary of the invention

[0005] The present invention provides a fog lamp induction warning control system, which effectively solves the problem that a traffic accident warning cannot be issued to the rear driver in time after a vehicle collision, which easily leads to a secondary collision accident.

[0006] In order to solve the above problems, the technical solution adopted by the invention is:

[0007] A fog lamp induction warning control system, comprising:

[0008] Data acquisition module: The acquisition module includes a vibration sensor, an acoustic sensor, and a light sensor. The vibration sensor is used to capture the vibration waveform generated when the vehicle collides and collect vibration data in real time;

[0009] Acoustic sensors are used to collect sound signals and capture characteristic audio of metal tearing and parts collision;

[0010] Light sensor, used to monitor the intensity changes around the fog lamps at a sampling frequency, and to capture data on sudden changes in light caused by sparks caused by collisions, lamp damage, or vehicle body obstruction in real time;

[0011] Signal processing module: used to process the signal of the data acquisition module. The specific processing method is as follows:

[0012] S01 simultaneously receives data collected by vibration sensors, acoustic sensors, and light sensors.

[0013] The raw data is filtered and denoised to remove possible interference and noise.

[0014] S02 For vibration data, calculate the frequency, amplitude, energy and other characteristic quantities of the vibration signal, convert the time domain vibration data into the frequency domain through fast Fourier transform, and extract the main frequency components and corresponding amplitudes;

[0015] S03 uses audio processing algorithms to extract the frequency, intensity, and duration characteristics of the sound.

[0016] S04 analyzes the light intensity data, including the slope, mutation amplitude, and duration of the light intensity change.

[0017] S05 assigns weights according to the importance and relevance of different sensor data.

[0018] The data processing module uses weighted averaging for fusion, setting the weight of vibration data to W1, the weight of acoustic data to W2, and the weight of light data to W3. The fused eigenvalue F is expressed as:

[0019] F=W1×F1+W2×F2+W3×F3

[0020] Among them, F1, F2, and F3 are the characteristic values ​​of vibration, acoustic, and light data, respectively;

[0021] S06 sets a threshold to determine whether the fused feature is an abnormal situation; the threshold is determined through statistical analysis of historical data or experiments, and a collision exceeding the threshold by 0% to 20% is defined as a mild collision, and A1 is output; when the threshold is exceeded by 20% to 50%, it is determined as a moderate collision, and A2 is output; if the threshold is exceeded by 50% or more, it is a severe collision, and A3 is output.

[0022] Warning module: The warning module includes a fog lamp and a sound alarm arranged next to the fog lamp. When the warning module receives the signal output by the receiving and processing module and alarms, when the warning module receives the output structure A1, the warning module triggers the fog lamp strobe frequency to be L1; when the warning module receives the output result A2, the warning module triggers the fog lamp strobe frequency to be L2, and also emits an intermittent alarm sound with a sound frequency of M1; when the warning module receives the output result A3, the warning module triggers the fog lamp strobe frequency to be L3, and also emits a continuous high-pitched alarm with a sound frequency of M2.

[0023] The principle of this scheme is:

[0024] This fog light induction warning control system collects vehicle collision related information in multiple dimensions through vibration, acoustic and light sensors. After filtering, denoising and feature extraction, it fuses data according to weights, judges the severity of the collision based on the threshold, and provides graded warnings through different fog light strobe frequencies and sound alarm responses. The entire system achieves comprehensive and accurate collision detection and effective graded warnings to ensure driving safety.

[0025] The beneficial effects of this program:

[0026] By integrating vibration sensors, acoustic sensors and light sensors for multi-dimensional data collection, it is possible to capture the characteristics of a collision from the perspective of different physical phenomena, avoiding the blind spots and misjudgment risks that may exist in a single sensor, thereby improving the comprehensiveness and accuracy of collision detection.

[0027] During signal processing, the raw data is filtered and denoised to effectively remove interference and noise, providing reliable data for subsequent feature extraction and analysis, making the system's understanding of the collision more in-depth and accurate.

[0028] Weights are assigned according to the importance and relevance of data from different sensors and weighted average fusion is performed, which fully considers the value of various data sources, achieves optimized data integration, and improves the reliability of collision judgment.

[0029] Reasonable thresholds are set to determine whether the fused features are abnormal, and different levels of mild, moderate and severe collisions are divided, so that the system can accurately assess the severity of the collision.

[0030] In the warning module, different signals are output according to the severity of the collision, triggering different fog light strobe frequencies and sound alarm responses. It can provide appropriate reminders in minor collisions to avoid excessive interference; enhance the warning effect in moderate collisions to attract more attention; and emit a strong and continuous high-pitched alarm in severe collisions to attract the attention of surrounding vehicles and pedestrians to the greatest extent, buying precious time for timely rescue and risk avoidance measures.

[0031] Furthermore, the data acquisition module is arranged in fog lamps arrayed on both sides of the road, and the fog lamps are located on both sides of the road.

[0032] Furthermore, in the signal processing module, wavelet transform is used to analyze the time-frequency characteristics of the vibration signal for calculating the characteristic quantity of the vibration data.

[0033] Furthermore, in the signal processing module, for processing acoustic data, a convolutional neural network is used to extract features of the input acoustic data, capture local patterns and frequency features in the audio signal, and then the features extracted by the convolutional neural network are input into the gated recurrent unit.

[0034] Furthermore, in S04, a classification algorithm combining random forest and extreme gradient boosting is used to distinguish light intensity changes caused by different reasons.

[0035] Furthermore, in the data processing module, an adaptive threshold setting method is adopted to automatically adjust the threshold according to the statistical characteristics of recent historical data. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flow chart of the present invention; DETAILED DESCRIPTION

[0037] Embodiment 1 is basically as attached Figure 1 As shown, a fog lamp induction warning control system includes:

[0038] Data acquisition module: The acquisition module includes a vibration sensor, an acoustic sensor, and a light sensor. The vibration sensor is used to capture the vibration waveform generated when the vehicle collides and collect vibration data in real time;

[0039] Acoustic sensors are used to collect sound signals and capture characteristic audio of metal tearing and parts collision;

[0040] Light sensor, used to monitor the intensity changes around the fog lamps at a sampling frequency, and to capture data on sudden changes in light caused by sparks caused by collisions, lamp damage, or vehicle body obstruction in real time;

[0041] Signal processing module: used to process the signal of the data acquisition module. The specific processing method is as follows:

[0042] S01 simultaneously receives data collected by vibration sensors, acoustic sensors, and light sensors.

[0043] The raw data is filtered and denoised to remove possible interference and noise.

[0044] S02 For vibration data, calculate the frequency, amplitude, energy and other characteristic quantities of the vibration signal, convert the time domain vibration data into the frequency domain through fast Fourier transform, and extract the main frequency components and corresponding amplitudes;

[0045] S03 uses audio processing algorithms to extract the frequency, intensity, and duration characteristics of the sound.

[0046] S04 analyzes the light intensity data, including the slope, mutation amplitude, and duration of the light intensity change.

[0047] S05 assigns weights according to the importance and relevance of different sensor data.

[0048] The data processing module uses weighted average for fusion. Since collision judgment is performed in foggy weather, weights are assigned according to the concentration of fog. When there is heavy fog, the reliability of light data is reduced, so the weight of light data is reduced, while the weights of other data are increased. The weight of vibration data is set to W1, the weight of acoustic data is set to W2, and the weight of light data is set to W3. The fused eigenvalue F is expressed as:

[0049] F=W1×F1+W2×F2+W3×F3

[0050] Among them, F1, F2, and F3 are the characteristic values ​​of vibration, acoustic, and light data, respectively;

[0051] S06 sets a threshold to determine whether the fused feature is an abnormal situation; the threshold is determined through statistical analysis of historical data or experiments, and a collision exceeding the threshold by 0% to 20% is defined as a mild collision, and A1 is output; when the threshold is exceeded by 20% to 50%, it is determined as a moderate collision, and A2 is output; if the threshold is exceeded by 50% or more, it is a severe collision, and A3 is output.

[0052] Warning module: The warning module includes a fog lamp and a sound alarm arranged next to the fog lamp. When the warning module receives the signal output by the receiving and processing module and alarms, when the warning module receives the output structure A1, the warning module triggers the fog lamp strobe frequency to be L1; when the warning module receives the output result A2, the warning module triggers the fog lamp strobe frequency to be L2, and also emits an intermittent alarm sound with a sound frequency of M1; when the warning module receives the output result A3, the warning module triggers the fog lamp strobe frequency to be L3, and also emits a continuous high-pitched alarm with a sound frequency of M2.

[0053] This fog light induction warning control system collects vehicle collision related information in multiple dimensions through vibration, acoustic and light sensors. After filtering, denoising and feature extraction, it fuses data according to weights, judges the severity of the collision based on the threshold, and provides graded warnings through different fog light strobe frequencies and sound alarm responses. The entire system achieves comprehensive and accurate collision detection and effective graded warnings to ensure driving safety.

[0054] By integrating vibration sensors, acoustic sensors and light sensors for multi-dimensional data collection, it is possible to capture the characteristics of a collision from the perspective of different physical phenomena, avoiding the blind spots and misjudgment risks that may exist in a single sensor, thereby improving the comprehensiveness and accuracy of collision detection.

[0055] During signal processing, the raw data is filtered and denoised to effectively remove interference and noise, providing reliable data for subsequent feature extraction and analysis, making the system's understanding of the collision more in-depth and accurate.

[0056] Weights are assigned according to the importance and relevance of data from different sensors and weighted average fusion is performed, which fully considers the value of various data sources, achieves optimized data integration, and improves the reliability of collision judgment.

[0057] Reasonable thresholds are set to determine whether the fused features are abnormal, and different levels of mild, moderate and severe collisions are divided, so that the system can accurately assess the severity of the collision.

[0058] In the warning module, different signals are output according to the severity of the collision, triggering different fog light strobe frequencies and sound alarm responses. It can provide appropriate reminders in minor collisions to avoid excessive interference; enhance the warning effect in moderate collisions to attract more attention; and emit a strong and continuous high-pitched alarm in severe collisions to attract the attention of surrounding vehicles and pedestrians to the greatest extent, buying precious time for timely rescue and risk avoidance measures.

[0059] The data acquisition module is arranged in the fog lights arrayed on both sides of the highway. The fog lights are located on both sides of the highway, close to the edge of the road, and can more quickly capture data such as vibration, sound and light changes related to vehicle collisions, thereby reducing data transmission delays, improving the timeliness and accuracy of data acquisition, and enabling the system to respond more quickly.

[0060] In the signal processing module, for the calculation of the characteristic quantity of vibration data, wavelet transform is used to analyze the time-frequency characteristics of the vibration signal. Wavelet transform can provide more accurate time-frequency localization information than traditional methods. In the scenario of vehicle collision, the vibration signal is often transient and non-stationary. Wavelet transform clearly reveals the changes in the frequency components of vibration at different time points, which helps to more accurately capture the subtle features of the collision moment. The vibration generated by vehicle collision may have instantaneous strong impact and short-term high-frequency vibration components. Wavelet transform can keenly identify these special vibration modes and improve the detection sensitivity and accuracy of collision events.

[0061] In the signal processing module, for the processing of acoustic data, a convolutional neural network is used to extract features of the input acoustic data to capture local patterns and frequency features in the audio signal, and then the features extracted by the convolutional neural network are input into a gated recurrent unit. Inputting the features extracted by the convolutional neural network into the gated recurrent unit can capture the temporal dynamic information in the acoustic signal and improve the ability to recognize and judge the acoustic features of vehicle collisions.

[0062] In S04, a classification algorithm combining random forest and extreme gradient boosting is used to distinguish light intensity changes caused by different reasons. Random forest has excellent noise resistance and tolerance to outliers. When processing data collected by the light sensor that may contain noise and anomalies, it can maintain good stability and accuracy, and reduce the impact of individual erroneous data points on the classification results; and the extreme gradient boosting algorithm, with its powerful learning ability and ability to fit complex relationships, can dig out deep nonlinear relationships and complex patterns in light intensity change data, thereby improving classification accuracy, and can improve classification accuracy, stability, generalization ability and computational efficiency, and provide more reliable and timely light intensity change analysis results for the fog light induction warning control system.

[0063] In the data processing module, an adaptive threshold setting method is adopted to automatically adjust the threshold according to the statistical characteristics of recent historical data. The adaptive threshold setting method can better adapt to the dynamic changes in the actual scene. Due to the complexity and uncertainty of the vehicle driving environment, the performance of the collision characteristics may be different. By automatically adjusting the threshold according to the statistical characteristics of recent historical data, the current actual situation is reflected in real time, and the accuracy and reliability of collision judgment are improved. At the same time, different vehicle types, driving speeds and road conditions will lead to differences in collision data. The adaptive threshold is adjusted according to the specific driving conditions, so that the system can judge whether a collision occurs in various complex traffic scenes, avoiding misjudgment or missed judgment that may be caused by fixed thresholds. Specific embodiment 2

[0065] Data acquisition modules are installed in an array inside the fog lamps on both sides of the road, and the vibration sensors, acoustic sensors and light sensors in them are always in working condition.

[0066] When a car collides with a truck, the vibration sensor quickly captures the vibration waveform generated by the collision and collects vibration data in real time. At the same time, the acoustic sensor collects the characteristic audio of metal tearing and parts collision, and the light sensor monitors the sudden change of light caused by the spark caused by the collision.

[0067] These raw data are simultaneously transmitted to the signal processing module, where they are first filtered and denoised to remove possible interference and noise. For vibration data, the time-domain vibration data is converted into the frequency domain through fast Fourier transform, and the characteristic quantities of the vibration signal, such as frequency, amplitude, and energy, are calculated. The time-frequency characteristics of the vibration signal are analyzed by wavelet transform to more accurately capture the subtle features of the collision moment. The frequency, intensity, and duration characteristics of the acoustic data are extracted using an audio processing algorithm, and a convolutional neural network is used to extract features from the input acoustic data. The extracted features are then input into a gated recurrent unit to capture the temporal dynamic information in the acoustic signal. For light data, the slope, mutation amplitude, and duration of the light intensity change are analyzed, and a classification algorithm combining random forest and extreme gradient boosting is used to distinguish the causes of the light intensity change.

[0068] According to the importance and relevance of different sensor data, the weight of vibration data is set to 0.4, the weight of acoustic data is set to 0.3, and the weight of light data is set to 0.3, and weighted average fusion is performed. The fused feature value is compared with the threshold automatically adjusted by the statistical characteristics of recent historical data. The feature value after fusion of this collision exceeds 35% of the threshold and is judged as a moderate collision.

[0069] The warning module receives a signal of a moderate collision, triggering the fog lights to flash at a higher strobe frequency and issuing an intermittent alarm sound with a moderate frequency, which can attract the attention of surrounding vehicles and pedestrians without being too harsh.

[0070] The above are only embodiments of the present invention. Common knowledge such as the known specific structures and characteristics in the scheme is not described in detail here. Ordinary technicians in the relevant field are aware of all the common technical knowledge in the technical field to which the invention belongs before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply conventional experimental means before that date. Under the enlightenment given by this application, ordinary technicians in the relevant field improve and implement this scheme in combination with their own abilities. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements are made, which should also be regarded as the scope of protection of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification are used to explain the content of the claims.

Claims

1. A fog lamp induction warning control system, characterized in that: A data acquisition module, the acquisition module includes a vibration sensor, an acoustic sensor, and a light sensor, the vibration sensor is used to capture the vibration waveform generated when the vehicle collides, and collect vibration data in real time; Acoustic sensors are used to collect sound signals and capture characteristic audio of metal tearing and parts collision; Light sensor, used to monitor the intensity changes around the fog lamps at a sampling frequency, and to capture data on sudden changes in light caused by sparks caused by collisions, lamp damage, or vehicle body obstruction in real time; Signal processing module: used to process the signal of the data acquisition module. The specific processing method is as follows: S01 simultaneously receives data collected by vibration sensors, acoustic sensors, and light sensors. The raw data is filtered and denoised to remove possible interference and noise. S02 For vibration data, calculate the frequency, amplitude, energy and other characteristic quantities of the vibration signal, convert the time domain vibration data into the frequency domain through fast Fourier transform, and extract the main frequency components and corresponding amplitudes; S03 uses audio processing algorithms to extract the frequency, intensity, and duration characteristics of the sound. S04 analyzes the light intensity data, including the slope, mutation amplitude, and duration of the light intensity change. S05 assigns weights according to the importance and relevance of different sensor data. The data processing module uses weighted averaging for fusion, setting the weight of vibration data to W1, the weight of acoustic data to W2, and the weight of light data to W3. The fused eigenvalue F is expressed as: F=W1×F1+W2×F2+W3×F3 Where F1, F2, and F3 are the characteristic values ​​of vibration, acoustic, and optical data, respectively; S06 sets a threshold to determine whether the fused feature is abnormal; the threshold is determined by statistical analysis of historical data or experiments, and a collision exceeding the threshold of 0% to 20% is defined as a mild collision, and A1 is output; when the threshold is exceeded by 20% to 50%, it is determined as a moderate collision, and A2 is output; if the threshold is exceeded by 50% or more, it is a severe collision, and A3 is output; The warning module includes a fog lamp and a sound alarm arranged next to the fog lamp. When the warning module receives a signal output by the receiving and processing module and sounds an alarm, when the warning module receives the output structure A1, the warning module triggers the fog lamp to have a strobe frequency of L1; when the warning module receives the output result A2, the warning module triggers the fog lamp to have a strobe frequency of L2, and also issues an intermittent alarm sound with a sound frequency of M1; when the warning module receives the output result A3, the warning module triggers the fog lamp to have a strobe frequency of L3, and also issues a continuous high-pitched alarm with a sound frequency of M2.

2. A fog lamp induction warning control system according to claim 1, characterized in that: The data acquisition module is arranged in the fog lamps on both sides of the road.

3. A fog lamp induction warning control system according to claim 1, characterized in that: In the signal processing module, wavelet transform is used to analyze the time-frequency characteristics of the vibration signal for calculating the characteristic quantity of the vibration data.

4. A fog lamp induction warning control system according to claim 3, characterized in that: In the signal processing module, for the processing of acoustic data, a convolutional neural network is used to extract features of the input acoustic data, capture local patterns and frequency features in the audio signal, and then the features extracted by the convolutional neural network are input into the gated recurrent unit.

5. The fog lamp induction warning control system according to claim 1, characterized in that: In S04, a classification algorithm combining random forest and extreme gradient boosting is used to distinguish light intensity changes caused by different reasons.

6. The fog lamp induction warning control system according to claim 1, characterized in that: In the data processing module, an adaptive threshold setting method is adopted to automatically adjust the threshold according to the statistical characteristics of recent historical data.

Images