Holographic traffic signal lamp system and method

Abstract

The invention relates to the technical field of traffic management infrastructures, and particularly discloses a holographic traffic signal lamp system and method, and the method comprises the steps: collecting vibration wave data in real time through a vibration detector, and remotely transmitting the collected vibration wave data to a third-party platform; the method comprises the following steps: extracting vibration wave archived data marked as abnormal from a three-party data source as a data set for comparative analysis, and introducing a deep learning algorithm to carry out waveform feature extraction on vibration wave data and each abnormal vibration wave archived data; the vibration wave data and each abnormal vibration wave archived data are subjected to dynamic query response analysis based on fluctuation characteristics, so that potential association between the real-time vibration wave data and the abnormal vibration wave archived data is mined, and intelligent earthquake early warning release is realized. In this way, the problem of false triggering caused by transient interference factors can be effectively solved, the accuracy and reliability of earthquake early warning are improved, and the situations of false alarm and missing alarm are reduced.

Description

Technical Field

[0001] This application relates to the technical field of traffic management infrastructure, and more specifically, to a holographic traffic signal system and method. Background Art

[0002] With the acceleration of urbanization and the development of technology, the concept of smart cities has gradually become an important direction for modern urban management. The construction of smart cities not only requires improving resource utilization efficiency and optimizing urban management, but also needs to have the ability to respond quickly in the face of emergencies such as natural disasters. As an important part of urban infrastructure, the functions of traffic signals have expanded from single traffic command to a comprehensive system integrating multiple intelligent detection and analysis technologies. For example, the invention patent with the publication number CN113724512A proposes a holographic traffic signal system, which not only has the traditional traffic command function, but also integrates functions such as road panoramic photography, microwave vehicle detection, air quality analysis and earthquake monitoring, providing a powerful basic traffic measure for the application of smart cities.

[0003] In this holographic traffic signal system, the earthquake monitoring function is realized through vibration wave analysis. When the vibration wave analysis data reaches the vibration wave threshold preset by the third-party platform, the system will trigger the earthquake warning mechanism and notify relevant departments and the public to take necessary protective measures in time, which helps to improve the earthquake warning ability and emergency response speed of the city. However, this earthquake warning mechanism based on simple threshold judgment may be interfered by various factors in actual applications, such as short-term strong winds, passing heavy trucks or nearby construction activities, resulting in abnormal fluctuations in vibration wave data, thus mis-triggering the earthquake warning mechanism.

[0004] Therefore, an optimized holographic traffic signal system and method are expected. Summary of the Invention

[0005] To solve the above technical problems, this application is proposed. The embodiments of this application provide a holographic traffic signal system and method, which use vibration detectors to collect vibration wave data in real time and remotely send the collected vibration wave data to a third-party platform. Then, the vibration wave archive data marked as abnormal is extracted from the three-party data source as a data set for comparison and analysis, and a deep learning algorithm is introduced to extract waveform features from the vibration wave data and each abnormal vibration wave archive data. Through dynamic query response analysis based on fluctuation characteristics of the vibration wave data and each abnormal vibration wave archive data, the potential association between the real-time vibration wave data and the abnormal vibration wave archive data is mined, so as to realize intelligent earthquake warning release. In this way, the problem of mis-triggering caused by transient interference factors can be effectively overcome, the accuracy and reliability of earthquake warning are improved, and the situation of false alarms and missed alarms is reduced.

[0006] Accordingly, according to one aspect of the present application, a holographic traffic signal lamp system is provided, which includes a signal lamp installed at an intersection, its signal lamp control box, a vibration detector, and a remote communication module. Among them, the vibration detector is used to collect vibration wave data in real time and send the vibration wave data to a third-party platform through the remote communication module. The third-party platform is used to compare and analyze the vibration wave data with three-party data sources to obtain a vibration wave analysis result, and the vibration wave analysis result is used to indicate whether to issue an earthquake early warning prompt. Among them, comparing and analyzing the vibration wave data with three-party data sources includes extracting fluctuation characteristics of the vibration wave data and archived vibration wave data marked as abnormal, and performing dynamic query response analysis based on the fluctuation characteristics to obtain the vibration wave analysis result.

[0007] According to another aspect of the present application, a control method for a holographic traffic signal lamp is provided, which includes:

[0008] Using a vibration detector to collect vibration wave data in real time and send the vibration wave data to a third-party platform through a remote communication module;

[0009] On the third-party platform, comparing and analyzing the vibration wave data with three-party data sources to obtain a vibration wave analysis result, and the vibration wave analysis result is used to indicate whether to issue an earthquake early warning prompt. Among them, comparing and analyzing the vibration wave data with three-party data sources includes extracting fluctuation characteristics of the vibration wave data and archived vibration wave data marked as abnormal, and performing dynamic query response analysis based on the fluctuation characteristics to obtain the vibration wave analysis result.

[0010] Compared with the prior art, the holographic traffic signal lamp system and method provided by the present application utilize a vibration detector to collect vibration wave data in real time and remotely send the collected vibration wave data to a third-party platform. Then, archived vibration wave data marked as abnormal is extracted from three-party data sources as a data set for comparison and analysis, and a deep learning algorithm is introduced to extract waveform characteristics of the vibration wave data and each abnormal archived vibration wave data. By performing dynamic query response analysis based on the fluctuation characteristics between the vibration wave data and each abnormal archived vibration wave data, the potential association between the real-time vibration wave data and the abnormal archived vibration wave data is mined, thereby realizing intelligent earthquake early warning release. In this way, the problem of false triggering caused by transient interference factors can be effectively overcome, the accuracy and reliability of earthquake early warning are improved, and the situations of false alarms and missed alarms are reduced. Description of the Drawings

[0011] The above and other objects, features, and advantages of the present application will become more apparent by describing the embodiments of the present application in more detail with reference to the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation to the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0012] Figure 1 It is a block diagram of a third-party platform in a holographic traffic signal lamp system according to an embodiment of the present application;

[0013] Figure 2 It is a schematic diagram of data flow of a third-party platform in a holographic traffic signal lamp system according to an embodiment of the present application;

[0014] Figure 3 It is a block diagram of a waveform feature query response coding module in a holographic traffic signal lamp system according to an embodiment of the present application;

[0015] Figure 4 It is a flowchart of a control method for a holographic traffic signal lamp according to an embodiment of the present application. Detailed implementation manners

[0016] Next, the embodiments of the present application will be described in more detail with reference to the accompanying drawings. The above and other objects, features, and advantages of the present application will become more apparent. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described herein.

[0017] As mentioned in the above background art, Patent CN113724512A proposed a holographic traffic signal lamp system, which includes signal lamps installed at intersections and their signal lamp control boxes, vibration detectors, and remote communication modules. In addition to the traditional traffic command function, it also integrates the function of earthquake monitoring, providing a powerful basic traffic measure for the application of smart cities.

[0018] In this holographic traffic signal lamp system, the earthquake monitoring function is realized through vibration wave analysis. When the vibration wave analysis data reaches the vibration wave threshold preset by the third-party platform, the system will trigger an earthquake warning mechanism, timely notify relevant departments and the public to take necessary protective measures, which helps to improve the city's earthquake disaster warning ability and emergency response speed. However, this earthquake warning mechanism based on simple threshold judgment may be interfered by various factors in actual applications, such as short-term strong winds, passing heavy trucks, or nearby construction activities, resulting in abnormal fluctuations in vibration wave data, thus mis-triggering the earthquake warning mechanism.

[0019] In view of the above technical problems, based on the above solutions, the present application proposes an optimized holographic traffic signal system, which uses a vibration detector to collect vibration wave data in real time and remotely transmits the collected vibration wave data to a third-party platform. Then, archived vibration wave data marked as abnormal is extracted from the three-party data source as a data set for comparison and analysis, and a deep learning algorithm is introduced to extract waveform features from the vibration wave data and each archived abnormal vibration wave data. By performing dynamic query response analysis based on wave characteristics on the vibration wave data and each archived abnormal vibration wave data, potential associations between the real-time vibration wave data and the archived abnormal vibration wave data are mined, thereby realizing intelligent earthquake early warning release. In this way, the problem of false triggering caused by transient interference factors can be effectively overcome, the accuracy and reliability of earthquake early warning can be improved, and the situations of false alarms and missed alarms can be reduced.

[0020] Specifically, in the technical solution of the present application, the vibration detector is used to collect vibration wave data in real time and send the vibration wave data to a third-party platform through a remote communication module. The third-party platform is used to compare and analyze the vibration wave data with a three-party data source to obtain a vibration wave analysis result, and the vibration wave analysis result is used to indicate whether to issue an earthquake early warning prompt. Among them, comparing and analyzing the vibration wave data with the three-party data source includes extracting wave characteristics and performing dynamic query response analysis based on wave characteristics on the vibration wave data and the archived vibration wave data marked as abnormal to obtain the vibration wave analysis result.

[0021] In specific implementation, the vibration detector, as the front-end device for data collection, is crucial in terms of its performance and selection. Currently, there are various types of vibration detectors on the market, such as accelerometers and displacement sensors based on the piezoelectric effect. In this system, a high-precision and wide-band accelerometer is selected as the vibration detector. This accelerometer can sensitively sense extremely subtle vibration changes on the earth's surface, and its working principle is based on the characteristic that when a piezoelectric material is deformed under an external force, a change in the electric charge amount will occur on its surface. When the accelerometer is fixed at a monitoring position such as the ground or the building foundation, the acceleration change caused by ground vibration will cause a corresponding change in the electric charge amount of the piezoelectric material, and this change in the electric charge amount is converted into an electrical signal proportional to the vibration acceleration through a specific circuit and output.

[0022] In order to achieve real-time acquisition of vibration wave data, it is also necessary to construct an efficient data acquisition mechanism. The electrical signals output by the vibration detector are first processed by a signal conditioning circuit. The main function of the signal conditioning circuit is to perform operations such as amplifying and filtering the original electrical signals. Since the electrical signals output by the vibration detector are usually relatively weak and may be mixed with various noise interferences, the amplifier circuit will amplify the signals to an amplitude range suitable for subsequent processing. At the same time, the filter circuit uses a band-pass filter. According to the frequency characteristics of seismic waves (the frequency range of general seismic waves is between 0.1 Hz and 10 Hz), it filters out interference signals such as high-frequency noise and low-frequency drift, and only retains the effective frequency components related to seismic waves. The electrical signals after signal conditioning enter the analog-to-digital converter (ADC). The role of the ADC is to convert continuous analog electrical signals into discrete digital signals so that the computer system can process and store them. To meet the requirements of real-time acquisition, a high-speed and high-precision ADC chip is selected, and its sampling frequency can be set according to actual needs, generally between several hundred Hz and several thousand Hz, to ensure that the dynamic changes of vibration waves can be accurately captured. For example, setting the sampling frequency to 1000 Hz means that 1000 vibration wave data points can be collected per second, which can more meticulously restore the waveform of vibration waves.

[0023] After being converted by the ADC, the digital signals enter the data acquisition controller. The data acquisition controller is usually implemented based on a microcontroller (MCU) or a digital signal processor (DSP). In this system, a high-performance DSP is selected as the data acquisition controller. The DSP has powerful digital signal processing capabilities and fast data operation speeds, and can efficiently process and store a large amount of vibration wave data. The data acquisition controller packs and stores the digital signals converted by the ADC according to the preset acquisition strategy, and monitors the status of data acquisition in real time. To ensure the integrity and accuracy of the data, the data acquisition controller also has data verification and error correction functions. For example, the cyclic redundancy check (CRC) algorithm is used to verify the collected data. Once data errors or losses are found, measures such as retransmission can be taken in a timely manner to correct them.

[0024] After the vibration detector collects vibration wave data in real time, it is necessary to send the data to a third-party platform through a remote communication module. When selecting a remote communication module, factors such as communication distance, data transmission rate, and stability need to be considered. In this system, a 5G wireless communication module is adopted as the main means of remote communication. 5G communication technology has a high data transmission rate and can meet the requirements of real-time and large-volume transmission of vibration wave data. The 5G wireless communication module establishes a connection with the mobile communication base station through an antenna, and encapsulates and transmits the vibration wave data packaged by the data acquisition controller according to the 5G communication protocol. During the data transmission process, in order to ensure the security and reliability of the data, encryption transmission technology, such as the SSL / TLS encryption protocol, is used to encrypt the transmitted data to prevent the data from being stolen or tampered with during transmission.

[0025] In terms of docking with the third-party platform, it is first necessary to communicate and negotiate fully with the third-party platform to clarify key information such as the interface specifications, data formats, and communication protocols for data transmission. According to the data interface document provided by the third-party platform, develop a dedicated data transmission program. The data transmission program runs on the data acquisition controller and is responsible for converting and encapsulating the locally collected vibration wave data into the format required by the third-party platform. For example, the third-party platform may require the data to be transmitted in JSON format. The data transmission program will convert the collected vibration wave data into a data structure that conforms to the JSON format specification and add necessary metadata information, such as the data acquisition time, acquisition location, device number, etc. Then, the data transmission program sends the encapsulated data to the third-party platform through the 5G wireless communication module. During the data sending process, the data transmission program will monitor the status of data transmission in real time, such as the sending progress and whether it is successful. If the data sending fails, the data transmission program will retransmit according to the preset retry strategy to ensure that the data can reach the third-party platform accurately and without error.

[0026] To ensure the stability and reliability of the entire data acquisition and transmission process, a complete set of monitoring and maintenance mechanisms also need to be established. In terms of hardware, regularly check and maintain devices such as vibration detectors, signal conditioning circuits, ADCs, and 5G wireless communication modules to ensure the normal operation of the devices. For example, check whether the installation of the vibration detector is firm and whether there are any loosening or damage; check whether the electronic components of the signal conditioning circuit are aging or damaged; regularly test the signal strength of the 5G wireless communication module to ensure the communication quality. In terms of software, monitor the running status of the data acquisition and transmission programs in real time, and record key events and error information during the program running process through log records. Once an abnormality in the program is found, it can be promptly investigated and repaired. At the same time, regularly conduct performance tests on the data acquisition and transmission system, such as indicators such as data acquisition accuracy, transmission rate, and packet loss rate, and optimize and adjust the system according to the test results.

[0027] Figure 1 It is a block diagram of a third - party platform in the holographic traffic signal light system according to an embodiment of the present application. Figure 2 It is a schematic diagram of data flow of a third - party platform in the holographic traffic signal light system according to an embodiment of the present application. As Figure 1 and Figure 2 shown, the third - party platform 100 includes: an abnormal vibration wave archive data extraction module 110, configured to extract vibration wave archive data marked as abnormal from the three - party data source to obtain a sequence of vibration wave archive data; a waveform feature extraction module 120, configured to extract waveform features from each vibration wave archive data in the vibration wave data and the sequence of vibration wave archive data to obtain a vibration wave detection data waveform feature coding vector and a sequence of archive vibration wave data waveform feature coding vectors; a waveform feature query response coding module 130, configured to perform feature search interaction response coding driven by dynamic weights on the vibration wave detection data waveform feature coding vector and the sequence of archive vibration wave data waveform feature coding vectors to obtain a vibration wave query response coding vector; an analysis result generation module 140, configured to determine the vibration wave analysis result based on the vibration wave query response coding vector.

[0028] In the third - party platform 100 of the above - mentioned holographic traffic signal light system, the abnormal vibration wave archive data extraction module 110 is configured to extract vibration wave archive data marked as abnormal from the three - party data source to obtain a sequence of vibration wave archive data. Specifically, the three - party data source is a data repository collected, sorted, and marked by a professional institution or system. Based on seismological principles, geological structure knowledge, and a large amount of past experience in monitoring geological activities such as earthquakes, the vibration wave data is marked, and the vibration wave data related to abnormal situations such as earthquakes is marked as abnormal. By using the vibration wave archive data marked as abnormal as a reference, the present application provides an important reference standard for the analysis of current real - time vibration wave data, which helps to more accurately identify the vibration wave characteristics related to earthquakes.

[0029] Specifically, before establishing a stable and reliable connection with a third-party data source, a comprehensive assessment of the data source is required, including its data quality, data update frequency, data storage format, and the type of data interfaces provided. For example, some data sources may store data in a relational database and provide data interfaces based on the SQL query language; while others may store data in the form of a file system and access data through specific APIs (Application Programming Interfaces). Based on the assessment results, select appropriate technical means to interface with the data source. If the data source provides an SQL interface, corresponding database connection tools can be used. For example, in Python, the SQLAlchemy library can be utilized to establish a connection with the data source database by configuring the correct database connection string. The connection string contains key information such as the database type (e.g., MySQL, PostgreSQL, etc.), server address, port number, database name, username, and password, ensuring that the data source can be accessed accurately and without error.

[0030] Since the data extraction process involves multiple links such as network communication, data source access, and data processing, a sound error handling and logging mechanism also needs to be established. For example, network connection interruption, data source server failure, data format mismatch, etc. To address these potential problems, the data extraction program needs to have strong error handling capabilities. When an error occurs, it can promptly capture the exception information and take corresponding recovery measures according to the error type. For example, if the network connection is interrupted, the program can attempt to re-establish the connection, and after a certain number of retries, if the connection still cannot be restored, record the error information and stop the data extraction process. At the same time, for subsequent problem troubleshooting and system optimization, it is necessary to record in detail every key operation and event during the data extraction process, including successful operations and errors that occur. The log record should contain detailed information such as the operation time, operation content, data involved, and error information. Through the analysis of the log file, the problem can be quickly located, and targeted improvements and optimizations can be made to the data extraction process.

[0031] In the third - party platform 100 of the above - mentioned holographic traffic signal light system, the waveform feature extraction module 120 is used to extract waveform features from each vibration wave archive data in the sequence of the vibration wave data and the vibration wave archive data to obtain a sequence of vibration wave detection data waveform feature coding vectors and archive vibration wave data waveform feature coding vectors. In a specific example of the present application, the waveform feature extraction module is used to: use a waveform feature extractor based on a CNN model to perform feature extraction on each vibration wave archive data in the sequence of the vibration wave data and the vibration wave archive data to obtain the sequence of the vibration wave detection data waveform feature coding vectors and the archive vibration wave data waveform feature coding vectors. It should be understood that since the vibration wave data is typical time - series data, there are certain structures and patterns in its time dimension. For example, before an earthquake occurs, the vibration waves will have different frequencies, amplitude changes, and waveform features. Therefore, in order to effectively capture the fluctuation patterns of the vibration wave data, the present application adopts a CNN (Convolutional Neural Network) model to extract the temporal fluctuation features of the vibration wave data and each vibration wave archive data. Specifically, the CNN model has excellent performance in processing data with time structures. In the present application, the CNN model slides a one - dimensional convolutional kernel in the time dimension of the vibration wave data and each vibration wave archive data to extract the fluctuation features of local time segments, so as to be able to capture the subtle changes and patterns of the vibration wave data in the time dimension. In addition, by setting one - dimensional convolutional kernels of different lengths, the fluctuation features of different time scales can be extracted, so as to more comprehensively understand the characteristics of the vibration wave data. For example, a smaller convolutional kernel can capture the high - frequency detailed features of the vibration wave, such as instantaneous vibration mutations; while a larger convolutional kernel can extract the low - frequency overall trend features, such as long - term vibration fluctuation conditions. In this way, the dynamic fluctuation patterns of the vibration wave data can be fully mined, and a sequence of vibration wave detection data waveform feature coding vectors and archive vibration wave data waveform feature coding vectors containing the temporal fluctuation information of the vibration wave can be obtained, which serves as an important basis for identifying earthquake signals.

[0032] In the third - party platform 100 of the above - mentioned holographic traffic signal light system, the waveform feature query response encoding module 130 is used to perform feature search interaction response encoding driven by dynamic weights on the sequence of the vibration wave detection data waveform feature coding vectors and the archive vibration wave data waveform feature coding vectors to obtain a vibration wave query response coding vector. That is, further perform semantic query interaction response analysis on the real - time vibration wave detection data waveform features and the waveform features of each vibration wave archive data marked as abnormal to explore the waveform similarity and correlation between the real - time vibration wave data and the abnormal vibration wave archive data, so as to determine whether the waveform change pattern of the currently detected vibration wave data matches the known earthquake wave waveform change pattern.

[0033] Figure 3 The block diagram of the waveform feature query response encoding module in the holographic traffic signal lamp system according to the embodiments of the present application. As Figure 3 shown, the waveform feature query response encoding module 130 includes: a deep implicit feature extraction unit 131, configured to perform deep implicit feature extraction on each archived shock wave data waveform feature encoding vector in the sequence of the shock wave detection data waveform feature encoding vector and the archived shock wave data waveform feature encoding vector to obtain a sequence of shock wave detection data waveform deep implicit feature encoding vectors and archived shock wave data waveform deep implicit feature encoding vectors; a semantic response interaction encoding unit 132, configured to perform semantic response decision anchoring encoding on the shock wave detection data waveform deep implicit feature encoding vector and each archived shock wave data waveform deep implicit feature encoding vector in the sequence of the archived shock wave data waveform deep implicit feature encoding vectors to obtain a sequence of shock wave waveform semantic feature query response anchoring encoding matrices; a dynamic aggregation encoding unit 133, configured to perform dynamic aggregation encoding on the sequence of the shock wave waveform semantic feature query response anchoring encoding matrices based on the feature contributions of each shock wave waveform semantic feature query response anchoring encoding matrix in the sequence of the shock wave waveform semantic feature query response anchoring encoding matrices to obtain the shock wave query response encoding vector.

[0034] In a specific example of the present application, the deep implicit feature extraction unit 131 is configured to: use a deep implicit feature extraction module based on a fully connected encoding network to process each archived shock wave data waveform feature encoding vector in the sequence of the shock wave detection data waveform feature encoding vector and the archived shock wave data waveform feature encoding vector to obtain the sequence of the shock wave detection data waveform deep implicit feature encoding vectors and the archived shock wave data waveform deep implicit feature encoding vectors, which is expressed by the formula:

[0035] V 2 ={v 21 ,v 22 ,...,v 2i ,...,v 2n}

[0036]

[0037] where V 2 represents the sequence of archived shock wave data waveform feature encoding vectors, v 21 , v 22 , v 2i and v 2nThe 1st, 2nd, i-th, and n-th archived shock wave data waveform feature coding vectors in the sequence representing the archived shock wave data waveform feature coding vectors respectively, where n is the number of vectors in the sequence of the archived shock wave data waveform feature coding vectors, V 1 represents the shock wave detection data waveform feature coding vector, W 1 represents the shock wave detection data waveform feature weight matrix, W 2 represents the archived shock wave data waveform feature weight matrix, b 1 represents the shock wave detection data waveform feature bias term, b 2 represents the archived shock wave data waveform feature bias term, sigmoid(·) represents the sigmoid activation function, V d1 represents the shock wave detection data waveform deep implicit feature coding vector, v d2i represents v 2i The corresponding archived shock wave data waveform deep implicit feature coding vector.

[0038] Here, in order to enhance the feature expression ability of real-time shock wave data and each piece of shock wave archived data marked as abnormal, the present application first uses a fully connected coding network to respectively perform deep implicit feature extraction on the shock wave detection data waveform feature coding vector and each archived shock wave data waveform feature coding vector, so as to utilize the powerful non-linear mapping ability of the fully connected coding network to learn the temporal non-linear correlation of the global waveform features of shock wave data, and more precisely depict the fluctuation characteristics of shock wave data, thereby obtaining a sequence of shock wave detection data waveform deep implicit feature coding vectors and archived shock wave data waveform deep implicit feature coding vectors.

[0039] In a specific example of the present application, the semantic response interaction coding unit 132 is expressed by the formula:

[0040]

[0041] where, (·) T represents the transpose of a vector, represents vector multiplication, S is the feature scale value of v d2i , M 12i represents V d1 and v d2i The shock wave waveform semantic feature query response anchor coding matrix between.

[0042] That is, further perform semantic response anchoring encoding on the depth implicit feature encoding vector of the vibration wave detection data waveform and the depth implicit feature encoding vectors of each archived vibration wave data waveform respectively, and perform feature interaction through the multiplication operation between vectors to capture the potential correlation pattern of waveform features between the real-time vibration wave data and each archived vibration wave data marked as abnormal, thereby generating a sequence of vibration wave waveform semantic feature query response anchoring encoding matrices.

[0043] In a specific example of the present application, the dynamic aggregation encoding unit 133 includes: a feature contribution quantum unit, which is used to calculate the decision anchor adaptive splicing factor of each vibration wave waveform semantic feature query response anchoring encoding matrix based on the feature distribution of each vibration wave waveform semantic feature query response anchoring encoding matrix in the sequence of vibration wave waveform semantic feature query response anchoring encoding matrices to obtain a sequence of vibration wave query response decision anchor adaptive splicing factors. More specifically, use the sum of the feature variance and the drift coefficient of the vibration wave waveform semantic feature query response anchoring encoding matrix as the numerator, and calculate the square of the difference between the maximum eigenvalue and the feature mean of the vibration wave waveform semantic feature query response anchoring encoding matrix multiplied by the number of its eigenvalues, plus the drift coefficient and twice the feature variance as the denominator to obtain the vibration wave query response decision anchor adaptive splicing factor, which is expressed by the formula:

[0044]

[0045] k = count(M 12i )

[0046] where count(·) represents calculating the number of elements of the matrix, k represents the difference amplification coefficient, that is, the number of eigenvalues of the vibration wave waveform semantic feature query response anchoring encoding matrix, σ 2 represents the feature variance of the vibration wave waveform semantic feature query response anchoring encoding matrix, ∈ represents the drift coefficient of the vibration wave waveform semantic feature query response anchoring encoding matrix, μ represents the feature mean of the vibration wave waveform semantic feature query response anchoring encoding matrix, max(·) is the maximum value function, E 12i represents the vibration wave query response decision anchor adaptive splicing factor corresponding to M 12i .

[0047] Here, in order to more precisely measure the correlation degree of waveform features between real-time vibration wave data and each vibration wave archive data, the present application further introduces a feature interaction response aggregation coding method driven by dynamic weights, and calculates its decision anchor adaptive splicing factor based on the feature distribution of the response anchor coding matrix queried by the semantic features of each vibration wave waveform, generating a sequence of vibration wave query response decision anchor adaptive splicing factors. Here, the vibration wave query response decision anchor adaptive splicing factor is used to reflect the semantic similarity degree of waveform features between each vibration wave archive data marked as abnormal and real-time vibration wave data, as well as the dominance of the semantic interaction information between the two in the global context, and is the weight basis for subsequent feature interaction response aggregation coding.

[0048] Specifically, here, the drift coefficient is used to smooth the fluctuation of the feature distribution of the vibration wave waveform semantic feature query response anchor coding matrix. In a preferred example of the present application, the calculation process of the drift coefficient can be expressed by the formula:

[0049]

[0050] where η is the intermediate transition representation value of the feature distribution equilibrium state of the vibration wave waveform semantic feature query response anchor coding matrix, m ij is the j-th eigenvalue of the matrix M 12i and e is the natural constant.

[0051] Preferably, for the drift coefficient ∈ in the vibration wave query response decision anchor adaptive splicing factor, for the state transition of the eigenvalue set distribution of the vibration wave waveform semantic feature query response anchor coding matrix from weak overall interpretability of the mean to strong local interpretability of the maximum value, the present application enhances the global dominance basis of the vibration wave waveform semantic feature query response anchor coding matrix by introducing the weak-to-strong interpretable generalization of the drift coefficient ∈.

[0052] Specifically, taking η as the intermediate state transition representation from weak interpretability to strong interpretability, for each eigenvalue m ij of the vibration wave waveform semantic feature query response anchor coding matrix, taking it as the importance score of the vibration wave waveform semantic feature query response anchor coding matrix for the global smoothing state transition, to perform global control of the importance score weight of the intermediate state transition η relative to the global state transition, so as to realize the interpretable generalization inference of the weight basis of the vibration wave query response decision anchor adaptive splicing factor.

[0053] In a specific example of the present application, the dynamic aggregation encoding unit 133 further includes: a weighting subunit, configured to perform weighting processing on the sequence of vibration wave query response decision anchor adaptive splicing factors based on the Softmax function to obtain a sequence of vibration wave query response decision anchor adaptive splicing weight factors; an aggregation encoding subunit, configured to perform weighted fusion and feature shape reshaping on the sequence of vibration wave waveform semantic feature query response anchor encoding matrices based on the sequence of vibration wave query response decision anchor adaptive splicing weight factors to obtain the vibration wave query response encoding vector, which is represented by the formula:

[0054] a 12i = softmax(E 12i )

[0055]

[0056] where softmax(·) represents the normalized exponential function, a 12i represents the vibration wave query response decision anchor adaptive splicing weight factor of matrix M 12i , M c represents the vibration wave waveform semantic feature query response anchor encoding fusion matrix, reshape(·) represents the feature shape reshaping function, and v c represents the vibration wave query response encoding vector.

[0057] That is, the sequence of vibration wave query response decision anchor adaptive splicing factors is normalized into a probability distribution form with a value range in the interval [0, 1] by using the Softmax function, as the weight coefficient for subsequent feature dynamic response aggregation encoding, to enhance the distinguishability and expression ability of features by amplifying the significant differences between each vibration wave waveform semantic feature query response anchor encoding matrix. Finally, weighted fusion is performed on the sequence of vibration wave waveform semantic feature query response anchor encoding matrices based on the generated weight coefficients, so as to adaptively focus on the vibration wave archive data waveform features that are strongly correlated with the waveform features of real-time vibration wave data, thereby generating the most responsive feature representation and obtaining the vibration wave query response encoding vector.

[0058] In the third-party platform 100 of the above holographic traffic signal light system, the analysis result generation module 140 is used to determine the vibration wave analysis result based on the vibration wave query response coding vector. In a specific example of the present application, the analysis result generation module 140 is used to: input the vibration wave query response coding vector into an anomaly diagnosis module based on a classifier to obtain the vibration wave analysis result. It should be understood that after the above processing, the vibration wave query response coding vector contains the waveform correlation response information between the real-time vibration wave data and each archived vibration wave data marked as abnormal. After receiving the vibration wave query response coding vector, the classifier determines whether to issue an earthquake warning prompt by learning its characteristic distribution pattern to judge whether the real-time vibration wave data conforms to the known earthquake signal characteristic pattern. Once the classifier determines that the current vibration wave data is an earthquake signal, an alarm will be immediately triggered, and an earthquake warning prompt will be sent to the surrounding area through the holographic traffic signal light so that refuge measures can be taken in time to minimize the losses caused by the earthquake disaster.

[0059] In summary, the holographic traffic signal light system according to the embodiment of the present application is clarified. It uses a vibration detector to collect vibration wave data in real time and remotely sends the collected vibration wave data to a third-party platform. Then, archived vibration wave data marked as abnormal is extracted from the three-party data source as a data set for comparison and analysis, and a deep learning algorithm is introduced to extract waveform features of the vibration wave data and each abnormal archived vibration wave data. Through dynamic query response analysis based on wave characteristics of the vibration wave data and each abnormal archived vibration wave data, the potential association between the real-time vibration wave data and the abnormal archived vibration wave data is mined, so as to realize the intelligent release of earthquake warnings. In this way, the problem of false triggering caused by transient interference factors can be effectively overcome, the accuracy and reliability of earthquake warnings can be improved, and the situations of false alarms and missed alarms can be reduced.

[0060] Furthermore, the present application also provides a control method for a holographic traffic signal light.

[0061] Figure 4 It is a flowchart of the control method for a holographic traffic signal light according to the embodiment of the present application. As Figure 4As shown, the control method of the holographic traffic signal lamp includes the steps of: S1, using a vibration detector to collect vibration wave data in real time and sending the vibration wave data to a third-party platform through a remote communication module; S2, at the third-party platform, comparing and analyzing the vibration wave data with a tripartite data source to obtain a vibration wave analysis result, where the vibration wave analysis result is used to indicate whether to issue an earthquake early warning prompt. Among them, comparing and analyzing the vibration wave data with the tripartite data source includes extracting the fluctuation characteristics of the vibration wave data and the archived vibration wave data marked as abnormal and performing dynamic query response analysis based on the fluctuation characteristics to obtain the vibration wave analysis result.

[0062] Here, those skilled in the art can understand that the specific operations of each step in the above control method of the holographic traffic signal lamp have been introduced in detail in the description of the holographic traffic signal lamp system above with reference to Figures 1 to 3 and therefore, the repeated description thereof will be omitted.

[0063] The basic principles of the present invention have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, advantages, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, advantages, effects, etc. are essential for each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and easy understanding, rather than limitations, and the above details do not limit the present invention to necessarily adopt the above specific details to be implemented.

Claims

1. A holographic traffic signal light system, comprising a signal light installed at an intersection and a signal light control box, a vibration detector and a remote communication module, characterized in that: The vibration detector is used to collect vibration wave data in real time, and send the vibration wave data to a third-party platform through the remote communication module. The third-party platform is used to compare and analyze the vibration wave data with the third-party data source to obtain a vibration wave analysis result. The vibration wave analysis result is used to indicate whether to issue an earthquake early warning prompt. The comparison and analysis of the vibration wave data with the third-party data source includes extracting fluctuation characteristics of the vibration wave data and the vibration wave archived data marked as abnormal, and performing dynamic query response analysis based on the fluctuation characteristics to obtain the vibration wave analysis result.

2. The holographic traffic signal light system according to claim 1, characterized in that: The third-party platform includes: An abnormal shock wave archive data extraction module is used to extract the shock wave archive data marked as abnormal from the three-party data source to obtain a sequence of shock wave archive data; A waveform feature extraction module, used to extract waveform features from the shock wave data and each shock wave archive data in the series of shock wave archive data to obtain a series of shock wave detection data waveform feature coding vectors and archive shock wave data waveform feature coding vectors; A waveform feature query response coding module, used for performing feature search interactive response coding based on dynamic weight driving on the series of the waveform feature coding vector of the vibration wave detection data and the waveform feature coding vector of the archived vibration wave data to obtain a vibration wave query response coding vector; The analysis result generating module is used to determine the shock wave analysis result based on the shock wave query response coding vector.

3. The holographic traffic signal light system according to claim 2, characterized in that: The waveform feature extraction module is used to: A waveform feature extractor based on a CNN model is used to extract features from the shock wave data and each shock wave archive data in the series of the shock wave archive data to obtain a series of the shock wave detection data waveform feature coding vectors and the archived shock wave data waveform feature coding vectors.

4. The holographic traffic signal light system according to claim 3, characterized in that: The waveform feature query response encoding module includes: A deep implicit feature extraction unit, used for performing deep implicit feature extraction on each archived shock wave data waveform feature coding vector in the sequence of the shock wave detection data waveform feature coding vector and the archived shock wave data waveform feature coding vector to obtain a sequence of the shock wave detection data waveform deep implicit feature coding vector and the archived shock wave data waveform deep implicit feature coding vector; A semantic response interaction coding unit, used for performing semantic response decision anchor coding on each archived shock wave data waveform depth implicit feature coding vector in the sequence of the shock wave detection data waveform depth implicit feature coding vector and the archived shock wave data waveform depth implicit feature coding vector to obtain a sequence of shock wave waveform semantic feature query response anchor coding matrices; A dynamic aggregation coding unit is used to dynamically aggregate the sequence of the vibration wave waveform semantic feature query response anchor coding matrices based on the feature contribution of each vibration wave waveform semantic feature query response anchor coding matrix in the sequence of the vibration wave waveform semantic feature query response anchor coding matrix to obtain the vibration wave query response coding vector.

5. The holographic traffic signal light system according to claim 4, characterized in that: The deep implicit feature extraction unit is used to: A deep implicit feature extraction module based on a fully connected coding network is used to process the vibration wave detection data waveform feature coding vector and each archived vibration wave data waveform feature coding vector in the series of the vibration wave detection data waveform feature coding vector to obtain the vibration wave detection data waveform deep implicit feature coding vector and the series of the archived vibration wave data waveform deep implicit feature coding vector.

6. The holographic traffic signal light system according to claim 5, characterized in that: The dynamic aggregation coding unit includes: A feature contribution measurement subunit is used to calculate the decision anchor adaptive splicing factors of each shock wave waveform semantic feature query response anchor coding matrix based on the feature distribution of each shock wave waveform semantic feature query response anchor coding matrix in the sequence of the shock wave waveform semantic feature query response anchor coding matrix to obtain a sequence of shock wave query response decision anchor adaptive splicing factors; A weighting subunit, used for performing weighting processing on the series of the shock wave query response decision anchor adaptive splicing factors based on the Softmax function to obtain a series of shock wave query response decision anchor adaptive splicing weight factors; The aggregation coding subunit is used to perform weighted fusion and feature reshape on the series of the shock wave waveform semantic feature query response anchor coding matrix based on the series of adaptive splicing weight factors of the shock wave query response decision anchor to obtain the shock wave query response coding vector.

7. The holographic traffic signal light system according to claim 6, characterized in that: The feature contribution measurement subunit is used to: The sum of the feature variance and the drift coefficient of the shock wave waveform semantic feature query response anchor coding matrix is ​​used as the numerator, and the square of the difference between the maximum eigenvalue and the feature mean of the shock wave waveform semantic feature query response anchor coding matrix is ​​calculated multiplied by the number of eigenvalues, and then the drift coefficient and twice the feature variance are added as the denominator to obtain the shock wave query response decision anchor adaptive splicing factor.

8. The holographic traffic signal light system according to claim 7, characterized in that: The drift coefficient is used to smooth the feature distribution fluctuation of the vibration wave waveform semantic feature query response anchor coding matrix.

9. The holographic traffic signal light system according to claim 8, characterized in that: The analysis result generating module is used for: The shock wave query response encoding vector is input into a classifier-based abnormality diagnosis module to obtain the shock wave analysis result.

10. A method for controlling a holographic traffic signal light, characterized in that: include: Using a vibration detector to collect vibration wave data in real time, and sending the vibration wave data to a third-party platform through a remote communication module; On the third-party platform, the shock wave data is compared and analyzed with the third-party data source to obtain a shock wave analysis result, and the shock wave analysis result is used to indicate whether to issue an earthquake warning prompt, wherein the comparison and analysis of the shock wave data with the third-party data source includes extracting fluctuation characteristics of the shock wave data and shock wave archived data marked as abnormal, and performing dynamic query response analysis based on the fluctuation characteristics to obtain the shock wave analysis result.

Images