Tunnel engineering construction procedure data automatic extraction method and system and electronic equipment

Abstract

The invention provides a tunnel engineering construction process data automatic extraction method and system and electronic equipment. The method comprises the following steps: acquiring a digital audio signal of a construction site acquired by an audio sensor; preprocessing the digital audio signal to extract audio features of the digital audio signal; inputting the audio features into a pre-trained construction scene classification model to obtain a construction scene classification result, and recording the starting and ending time of recognition of each construction scene; wherein each construction scene corresponds to one working procedure; and based on the construction scene classification result and the starting and ending time of each construction scene recognition, determining a process type and the starting and ending time of each process. According to the method, the accuracy of process identification is greatly improved, the process is automatically identified through the pre-trained construction scene classification model, and the efficiency of process identification is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, system and electronic equipment for automatically extracting tunnel engineering construction process data. Background Art

[0002] Tunnel construction is a complex and critical task, requiring consideration not only of geological conditions, groundwater levels, and geotechnical stability, but also of how to improve construction efficiency, ensure safety, and minimize environmental impact. Therefore, to improve efficiency and safety, it is crucial to accurately understand and predict data on tunnel construction processes.

[0003] Currently, data collection during tunnel construction processes mainly relies on manual recording or video recognition. These methods have some problems, such as manual recording is prone to errors, and video recognition is significantly affected by lighting conditions.

[0004] Therefore, a more efficient and accurate method is needed to collect and process tunnel construction process data. Summary of the Invention

[0005] The purpose of the embodiments of the present application is to provide a method, system and electronic equipment for automatically extracting tunnel engineering construction process data, so as to improve the accuracy and efficiency of tunnel construction process data identification.

[0006] In a first aspect, a method for automatically extracting tunnel engineering construction process data is provided, the method comprising: Acquire digital audio signals from the construction site collected by audio sensors; Preprocessing the digital audio signal to extract audio features of the digital audio signal; The audio features are input into the pre-trained construction scene classification model to obtain the construction scene classification results, and the start and end time of each construction scene recognition are recorded. Each construction scene corresponds to a process. Based on the construction scene classification results and the start and end time of each construction scene identification, the process type and the start and end time of each process are determined.

[0007] Optionally, preprocessing the digital audio signal to extract audio features of the digital audio signal includes: Restore the digital audio signal to obtain the original analog audio signal; Detect the zero-crossing rate and speech energy of analog audio signals; Extracting the spectrum characteristics of the digital audio signal using a preset spectrum conversion method; The zero-crossing rate, speech energy and spectrum characteristics are jointly determined as audio features.

[0008] Optionally, the extracting the spectral feature of the digital audio signal by using a preset spectral conversion method comprises: converting the digital audio signal into spectral data by using a short-time fast Fourier transform or a mel-frequency spectral conversion method; extracting a spectral centroid of the spectral data.

[0009] Optionally, the method further comprises: performing frequency domain filtering processing on the spectral feature of the audio to obtain a filtered spectral feature; inputting the filtered spectral feature and other audio features into a pre-trained construction machinery classification model to perform machinery equipment category identification, to obtain a machinery equipment classification result, and to record a voiceprint feature start time of the machinery equipment; taking the voiceprint feature start time of the machinery equipment as a time reference, recording a time when the voiceprint feature appears after the time reference, and obtaining a maximum time interval, a minimum time interval, and an average time interval; if the voiceprint feature of the machinery equipment is not detected again within a preset multiple time length of the maximum time interval, determining a time when the voiceprint feature last recorded appears as a voiceprint feature stop time of the machinery equipment, and taking the voiceprint feature start time and the voiceprint feature stop time as work start and end times of the machinery equipment; determining a work procedure identification result based on the machinery equipment classification result and a preset mapping relationship between the machinery equipment and the work procedure; relabeling a work procedure type and work procedure start and end times based on the work procedure identification result and the work start and end times of each machinery equipment, to obtain relabeled work procedure types and work procedure start and end times.

[0010] Optionally, the performing frequency domain filtering processing on the spectral feature of the audio to obtain a filtered spectral feature comprises: collecting background noise audio data during tunnel construction; converting the background noise audio data into a frequency domain to obtain frequency components of the background noise; subtracting the frequency components of the background noise from the spectral feature of the audio to be filtered to obtain the filtered spectral feature.

[0011] Optionally, the method further comprises: merging the work procedure type and the work procedure start and end times to form work procedure data; outputting the merged work procedure data in a preset format, the preset format being at least one of the following: a table and a text.

[0012] Optionally, the method further comprises: acquiring the digital audio signal of the construction site collected by the audio sensor; acquiring the digital audio signal of the construction site collected by the audio sensor at a preset interval.

[0013] In a second aspect, a tunnel engineering construction procedure data automatic extraction system is provided, and the system comprises: an audio sensor, a mounting position configuration unit, and a procedure identification platform; The mounting position configuration unit is configured to configure a construction site where the audio sensor is mounted. The audio sensor comprises a sound pickup device, an AD conversion unit, and a communication unit. The sound pickup device is mounted at a corresponding position of a construction site according to the configured construction site, and is configured to pick up an analog sound signal of the construction site. The AD conversion unit is configured to convert the analog sound signal into a digital audio signal. The communication unit is configured to transmit the digital audio signal to the procedure identification platform. The procedure identification platform is configured to execute the tunnel engineering construction procedure data automatic extraction method according to any one of the first aspect.

[0014] In a third aspect, an electronic device is provided, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory are in communication with each other through the communication bus. The memory is configured to store a computer program. The processor is configured to execute the program stored in the memory, and implement the method steps of any one of the first aspect.

[0015] In a fourth aspect, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method steps of any one of the first aspect.

[0016] The tunnel engineering construction procedure data automatic extraction method, system, and electronic device provided by the embodiments of the present application can obtain a digital audio signal collected by an audio sensor at a construction site, pre-process the digital audio signal to extract an audio feature of the digital audio signal, input the audio feature into a pre-trained construction scene classification model to obtain a construction scene classification result, and record a start time and an end time of each construction scene identification, wherein each construction scene corresponds to a procedure; and determine a procedure type and a start time and an end time of each procedure based on the construction scene classification result and the start time and the end time of each construction scene identification. The present application can record the start time and the end time of the construction by taking the scene sound of the construction equipment as a criterion, and will not cause the procedure data deviation caused by the arrival of the team personnel and the arrival of the construction machinery, and the accuracy of the procedure identification is greatly improved compared with the method of identifying the procedure by the video, and the pre-trained construction scene classification model is used to automatically identify the procedure, and the efficiency of the procedure identification is improved.

[0017] In order to make the above objectives, characteristics and advantages of the present application more apparent, more comprehensible, the following preferred embodiments are specifically described in detail below, together with the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0019] Figure 1 A flowchart of a tunnel engineering construction process data automatic extraction method provided by an embodiment of the present application; Figure 2 A structural schematic diagram of a tunnel engineering construction process data automatic extraction system provided by an embodiment of the present application; Figure 3 A structural schematic diagram of a tunnel engineering construction process data automatic extraction system provided by another embodiment of the present application; Figure 4 A structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the objectives, technical solutions and advantages of the embodiments of the present application more apparent, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0021] At present, tunnel construction process data collection mainly relies on manual recording or video recognition technology. These methods have some problems, for example, manual recording is prone to errors, and video recognition is greatly affected by light conditions.

[0022] Based on this, the embodiments of the present application provide a tunnel engineering construction process data automatic extraction method and system. The following will be described by embodiments.

[0023] Before the embodiments are described, first of all, the related knowledge of tunnel engineering is explained, so that the person skilled in the art can better understand the embodiments of the present application.

[0024] Tunnel engineering generally adopts open cut method and subsurface excavation method for construction, and the open cut method is commonly used in urban municipal tunnel construction, and the construction conditions are limited. In the case of complex and harsh construction environment and complex construction geology, shield method (including TBM) and mine method (also known as New Austrian Tunneling Method, i.e. blasting method) are commonly used for construction.

[0025] When the mine method is used for construction, the large cycle of tunnel construction is the construction of the working face, the construction of the secondary lining and the construction of the inverted arch, among which the construction of the working face is the basis and the key link affecting the efficiency of tunnel construction. Due to the influence of construction blasting and adverse gases after blasting, the construction of the secondary lining and the construction of the inverted arch will be restricted by the construction of the working face. The construction of the working face is a cyclic construction process composed of drilling, blasting, ventilation, risk removal, spoil removal, shotcreting and erecting, etc., and the last process is closely connected with the previous process. In order to not affect the progress of the project, multiple construction teams and multiple equipment need to be used for high-efficiency cooperation.

[0026] In the traditional construction operation, the alternation between the teams of the previous and subsequent processes is realized by the construction staff according to the on-site construction conditions to inform the next team, which often leads to the difficulty in realizing the high-efficiency connection between the teams, increases the time consumption of the construction cycle and results in the low efficiency of the construction cycle.

[0027] The following explains the various professional terms involved: Open cut method: This is a method of constructing a tunnel by excavating a large pit on the ground surface. It is usually suitable for conditions where the geological conditions are good and the surrounding environment allows large-scale ground operations. This method is commonly used in urban municipal tunnel construction.

[0028] Subsurface excavation method: This is a method of excavating underground, which is opposite to the open cut method and is suitable for conditions where large-scale ground excavation cannot be performed or is not suitable.

[0029] Shield method: This is a method of tunneling using a specially designed machine, the shield machine (Tunnel Boring Machine, TBM). The shield machine can support the tunnel wall while tunneling, and is particularly suitable for long-distance and large-diameter tunnel construction.

[0030] Mine method: This is a method of excavating a tunnel using blasting technology, which emphasizes controlling blasting parameters to reduce disturbance to the surrounding rock and supporting the surrounding rock in time through methods such as shotcrete to maintain its stability.

[0031] Working face construction: The working face refers to the working face that directly faces the rock or soil for drilling, blasting and cleaning during tunnel construction. It is the basic link of tunnel construction and affects the progress of the entire project.

[0032] Secondary lining construction: namely secondary lining construction, refers to the initial support structure after the stability, in order to further enhance the safety and durability of the tunnel, inside the primary support pouring concrete or other materials formed by the permanent lining.

[0033] Inverted arch construction: inverted arch is a special form of lining structure located at the bottom of the tunnel, which can effectively disperse the upper load and enhance the stability of the overall structure of the tunnel.

[0034] The embodiment of the application is for the automatic identification of each construction process of each construction link (working face construction, secondary lining construction, inverted arch construction, etc.) in the mine method construction.

[0035] The embodiment of the application provides a tunnel engineering construction process data automatic extraction method, which is applied to a process identification terminal, as shown in the figure, and the method comprises the following steps: Figure 1 Step S101: acquiring the digital audio signal of the construction site collected by the audio sensor.

[0036] In this step, a plurality of audio sensors are respectively installed at different positions of the construction site, for example, some are installed at the working face position, some are installed at the secondary lining construction area and the inverted arch construction area.

[0037] Each audio sensor is assigned a unique ID, and the ID is uploaded at the same time when the audio signal is collected on the audio sensor. In this way, when the process identification terminal receives the digital audio signal, the corresponding ID of the digital audio signal is obtained by analyzing the digital audio signal, so that the construction position of the digital audio signal can be determined.

[0038] In a feasible embodiment, in order to reduce the data acquisition and transmission amount, the digital audio signal of the construction site collected by the audio sensor is acquired at a preset interval. For example, it is collected once every 2 seconds.

[0039] The interval acquisition method can greatly reduce the data amount without affecting the capture of key information. Less data amount means that the amount of information that the back-end processing system needs to process is also reduced, which helps to improve the speed and efficiency of data analysis and processing.

[0040] Step S102: pre-processing the digital audio signal to extract the audio features of the digital audio signal.

[0041] By extracting some representative features of the digital audio signal as input data of the subsequent model, the process can be quickly identified, and the efficiency of identification is improved.

[0042] The specific pre-processing process will be described in the following embodiments, and will not be described here.​

[0043] Step S103: input the audio features into the pre-trained construction scene classification model to obtain a construction scene classification result, and record the start and end time of each construction scene identification; wherein each construction scene corresponds to a construction procedure.

[0044] In this step, the construction scene classification model may be, for example, a YAMnet model, a Tranformer model, etc.

[0045] Taking the YAMnet model as an example, the process of training the construction scene classification model includes: First step: obtain some original audio materials about various construction scenes in construction sites; Second step: pre-process the original audio materials, which includes extracting audio features; Third step: label the pre-processed audio features with construction scene labels; Fourth step: input the pre-processed audio features and construction scene labels into the YAMnet model for iterative training until a preset number of iterations is reached.

[0046] Step S104: determine the type of construction procedure and the start and end time of each construction procedure based on the construction scene classification result and the start and end time of each construction scene identification.

[0047] In this step, the construction scene classification result may be, for example, drilling, blasting, ventilation, risk removal, slag removal, spray mixing, and stand erection, etc.

[0048] Each scene corresponds to a construction procedure, so after determining the construction scene, the construction procedure and the start and end time of the procedure can be determined.

[0049] In this step, the start and end time of the procedure can be processed by time alignment, taking the first appearance time of the identified / labelled construction procedure classification result as the start time of the procedure, and according to the prior knowledge of tunnel construction procedures (such as procedure connection relationship, connection interval time threshold, etc.), the time when the procedure scene no longer appears is marked as the end time of the procedure operation.

[0050] In addition, after determining the procedure classification result, the sound features can be saved as the "scene voiceprint" of various construction procedures in combination with the classification result.

[0051] Through the above embodiment, it can be understood that the construction start and end time of the construction equipment is recorded by taking the scene sound when the construction equipment is working as a criterion, without process data deviation caused by the arrival of team personnel and construction machinery, and compared with the method of identifying the process through video, the accuracy of process identification is greatly improved, and the process identification efficiency is improved by automatically identifying the process through the pre-trained construction scene classification model.

[0052] On the basis of the above embodiment, the digital audio signal is preprocessed to extract the audio features of the digital audio signal, including: Step S102A: restore the digital audio signal to obtain the original analog audio signal.

[0053] In this step, the digital audio signal can be restored to the original mode audio signal through the DAC converter.

[0054] Step S102B: detect the zero-crossing rate and speech energy of the analog audio signal.

[0055] In this step, the zero-crossing rate refers to the number of times the audio signal crosses the zero point per unit time, which is commonly used to distinguish different tones or noise levels. It can assist in determining whether a certain audio segment contains speech.

[0056] In a specific example, the calculation method of the zero-crossing rate is to traverse the analog audio signal and count the number of sign changes between adjacent samples. For each sample point, check whether it and the previous point cross the 0 axis (i.e. one positive and one negative), and count.

[0057] The speech energy reflects the average value of the intensity or amplitude square of the audio signal within a period of time. High energy area corresponds to important work events. Combined with the zero-crossing rate, the speech segment can be more accurately located and segmented.

[0058] In a specific example, the calculation method of the speech energy is to square the amplitude of each frame of the signal and then take the average. This can be achieved through the sliding window technique, which moves a fixed number of sample points each time and calculates the square sum of all samples in the window.

[0059] Step S102C: extract the frequency spectrum features of the digital audio signal using a preset frequency spectrum conversion method.

[0060] Different types of audio signals (such as slagging and blasting) have different frequency spectrum features, which make them key features in classification and identification tasks.

[0061] In one possible implementation, the digital audio signal can be converted into spectral data using Short-Time Fourier Transform (STFT) or Mel-frequency cepstral transform. Then the spectral centroid of the spectral data is extracted.

[0062] In one specific implementation, the steps of STFT are as follows: Step 1: Framing and windowing: The continuous audio signal is divided into multiple overlapping segments (frames), and a window function (e.g., Hamming window) is applied to each frame to reduce edge effects.

[0063] Step 2: Fast Fourier Transform (FFT): FFT is performed on each frame to convert the time-domain signal into a frequency-domain representation. This results in a spectrogram for each frame.

[0064] Step 3: Extracting spectral centroid: The spectral centroid can be regarded as the "center of gravity" of the frequency, providing information about the "brightness" of the sound. The calculation formula is as follows:

[0065] where, represents the nth frequency value; represents the amplitude spectrum value corresponding to the frequency; N represents the number of frequency points.

[0066] In another possible implementation, the steps of extracting spectral features using Mel-frequency cepstral transform are as follows: Pre-emphasis: To compensate for the loss of high-frequency components, pre-emphasis processing is performed on the original signal.

[0067] Framing and windowing: Similar to STFT, the signal is first framed and a window function is applied.

[0068] Mel filter bank: A set of bandpass filters is used to map the frequency spectrum on a linear scale to a Mel scale, simulating the perceptual characteristics of the human auditory system.

[0069] Discrete cosine transform (DCT): DCT is performed on the energy after the Mel filter bank to obtain MFCC coefficients.

[0070] However, in this scenario, we are mainly interested in the spectral data rather than the final MFCC.

[0071] Extracting spectral centroid: Similarly, based on the results of the Mel-frequency spectrum, we can calculate the spectral centroid according to the above formula, except that the frequency axis is defined on the Mel scale.

[0072] Step S102D: jointly determine the zero-crossing rate, the speech energy and the spectral feature as the audio feature.

[0073] A single feature can only reflect a certain aspect of the audio signal, while the comprehensive use of multiple features can comprehensively describe the audio from multiple dimensions such as the time domain and the frequency domain. Different features can complement each other and reduce the possibility of misjudgment. For example, in a low signal-to-noise ratio environment, it may be difficult to distinguish speech and background noise relying on energy alone, but the recognition accuracy will be improved after combining the zero-crossing rate and the spectral feature. In the face of different types of construction scenes, this multi-feature method shows stronger adaptability and flexibility.

[0074] Therefore, by adopting the combination of the zero-crossing rate, the speech energy and the spectral feature, the performance of the audio processing system can be effectively improved, and the system is more accurate, stable and efficient.

[0075] On the basis of the above embodiment, the method further comprises: Step S105: performing frequency domain filtering processing on the spectral feature of the audio to obtain a filtered spectral feature.

[0076] In one example, a band-pass filter, a high-pass filter or a low-pass filter can be used, and appropriate filter parameters can be selected according to different construction positions.

[0077] Through the frequency domain filtering processing, unnecessary frequency components are removed, and spectral features useful for mechanical equipment classification are retained, reducing the sound interference of working equipment in different regions at the same time.

[0078] Step S106: jointly inputting the filtered spectral feature and other audio features into a pre-trained construction machinery classification model to perform mechanical equipment category recognition, obtaining a mechanical equipment classification result, and recording a voiceprint feature start time of the mechanical equipment; Taking the voiceprint feature start time of the mechanical equipment as a time reference, recording the time when the voiceprint feature appears after the time reference, and obtaining a maximum time interval, a minimum time interval and an average time interval; If the voiceprint feature of the mechanical equipment is not detected again within a preset multiple time length of the maximum time interval, the time when the last recorded voiceprint feature appears is determined as the voiceprint feature stop time of the mechanical equipment; and the voiceprint feature start time and the voiceprint feature stop time are taken as the working start and stop time of the mechanical equipment.

[0079] In the embodiment of the application, the construction machinery classification model can be based on a support vector machine (SVM), a convolutional neural network CNN or a CNN+RNN hybrid neural network.

[0080] The training steps of the construction machinery classification model can refer to the training steps of the construction scene classification model described above. The difference is that the training samples can be the original audio materials picked up by the audio sensor during the single device operation, and the labeled label is the category label of the mechanical equipment. In this way, the training effect of the mechanical equipment classification model can be improved.

[0081] Step S107: Based on the mechanical equipment classification result and the preset mapping relationship between the mechanical equipment and the process, the process recognition result is determined.

[0082] A database or mapping table is established in advance, which contains the relationship between different mechanical equipment and its corresponding process. For example, the slag car is associated with the slagging process.

[0083] Step S108: Based on the process recognition result, the work start and end time of each mechanical equipment, the process type and the process start and end time are re-calibrated respectively, and the re-calibrated process type and the process start and end time are obtained.

[0084] Combined with the construction scene (such as ground excavation, concrete pouring, etc.) and the specific mechanical equipment usage (such as excavator, pump truck, etc.), the specific process being performed can be more accurately identified. For example, it may be difficult to distinguish between excavation work and earthwork transportation work based on scene audio features alone, but the presence of excavators and dump trucks can more accurately determine the specific process.

[0085] Taking into account the type of mechanical equipment and its application in the construction scene, the accuracy and reliability of process recognition can be greatly improved.

[0086] In a feasible implementation, the frequency domain filtering processing of the frequency spectrum features of the audio to obtain the filtered frequency spectrum features includes: Collect background noise audio data during tunnel construction; Convert the background noise audio data to the frequency domain to obtain the frequency components of the background noise; Subtract the frequency components of the background noise from the frequency spectrum features of the audio to be filtered to obtain the filtered frequency spectrum features.

[0087] Based on the above embodiment, the method further includes: Merge the process type and the process start and end time to form process data; Output the merged process data in a preset format, and the preset format is at least one of the following: table, text.

[0088] Based on the same inventive concept, a tunnel engineering construction process data automatic extraction system is provided, as shown in Figure 2 The system includes: An audio sensor, an installation position configuration unit, and a process recognition platform 204; An installation position configuration unit is configured to configure a construction site for audio sensor installation; The audio sensor comprises a pickup 201, an AD conversion unit 202, and a communication unit 203. The pickup is installed at a corresponding position of the construction site according to the configured construction site, and is configured to pick up an analog sound signal of the construction site. The AD conversion unit is configured to convert the analog sound signal into a digital audio signal. The communication unit is configured to transmit the digital audio signal to a process identification platform. The communication unit can be wireless communication or wired communication.

[0089] The process identification platform is configured to execute the tunnel engineering construction process data automatic extraction method of any one of the above embodiments.

[0090] As shown in Figure 3 In a feasible implementation, in the process identification platform 204, a sensor management unit 2041, a data access unit 2042, a voice restoration unit 2043, a frequency domain filtering unit 2044, a data preprocessing unit 2045, a data management unit 2046, a construction machinery classification model 2047, a construction scene classification model 2048, and a time alignment unit 2049 are included.

[0091] The data access unit is configured to receive the digital audio signal uploaded by the communication unit.

[0092] The sensor management unit is configured to manage the ID of each pickup.

[0093] The voice restoration unit is configured to restore the digital audio signal to an analog audio signal.

[0094] The frequency domain filtering unit is configured to filter the digital audio signal.

[0095] The data preprocessing unit is configured to preprocess the digital audio signal, and the specific preprocessing process is the same as the preprocessing process in the above method embodiments.

[0096] The construction machinery classification model is configured to classify and identify the mechanical equipment.

[0097] The construction scene classification model is configured to classify and identify the construction scene.

[0098] The time alignment unit is configured to record and align the start and end times of each process and the start and end times of each mechanical equipment operation.

[0099] The data management unit is configured to store and manage the mapping relationship between the working process in the tunnel engineering and the main construction machinery.

[0100] Based on the same technical concept, the embodiment of the present application also provides an electronic device, such as Figure 4 As shown, the electronic device comprises a processor 401, a communication interface 402, a memory 403 and a communication bus 404, wherein the processor 401, the communication interface 402 and the memory 403 complete communication with each other through the communication bus 404.

[0101] The memory 403 is used for storing a computer program. The processor 401 is used for executing the program stored in the memory 403, so as to realize the steps of the tunnel engineering construction process data automatic extraction method.

[0102] The communication bus of the electronic device mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0103] The communication interface is used for communication between the electronic device and other devices.

[0104] The memory can comprise a Random Access Memory (RAM) and can also comprise a Non-Volatile Memory (NVM), for example at least one disk memory. Optionally, the memory can also be at least one storage device located away from the aforementioned processor.

[0105] The processor mentioned above can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; can also be a Digital Signal Processing (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0106] In the embodiments of the present application, it should be understood that the disclosed system and method can be implemented in other manners. The embodiments described above are merely exemplary. For example, the division of the units is only a logical function division, and there can be another division manner in actual implementation; for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, and electrical, mechanical or other forms.

[0107] In addition, the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. In actual implementation, a part or all of the units can be selected according to the actual needs to achieve the purposes of the embodiments of the present application.

[0108] In addition, the functional modules in the various embodiments of the present application can be integrated together to form a separate part, or each module can exist independently, or two or more modules can be integrated to form a separate part.

[0109] It should be noted that if the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the parts that make contributions to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and various media that can store program codes.

[0110] In this document, the terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations.

[0111] The above merely provides an example of the present application, and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for automatically extracting tunnel construction process data, characterized in that: The method comprises: Acquire digital audio signals from the construction site collected by audio sensors; Preprocessing the digital audio signal to extract audio features of the digital audio signal; Inputting the audio features into a pre-trained construction scene classification model to obtain a construction scene classification result, while recording the start and end time of each construction scene identification; wherein each construction scene corresponds to a process; The process type and the start and end time of each process are determined based on the construction scene classification result and the start and end time of each construction scene identification.

2. The method according to claim 1, characterized in that The preprocessing of the digital audio signal to extract the audio features of the digital audio signal includes: Restoring the digital audio signal to obtain the original analog audio signal; Detecting the zero-crossing rate and speech energy of the analog audio signal; Extracting the spectrum characteristics of the digital audio signal using a preset spectrum conversion method; The zero-crossing rate, speech energy, and the spectrum feature are jointly determined as audio features.

3. The method according to claim 2, characterized in that The extracting of the spectrum characteristics of the digital audio signal by using a preset spectrum conversion method includes: Converting the digital audio signal into spectrum data using short-time fast Fourier transform or Mel spectrum conversion method; Extract the spectrum centroid of the spectrum data.

4. The method according to claim 2, characterized in that The method further comprises: Performing frequency domain filtering on the spectral features of the audio to obtain filtered spectral features; Input the filtered spectral features and other audio features into a pre-trained construction machinery classification model to identify the type of machinery and equipment, obtain a classification result for the machinery and equipment, and simultaneously record the start time of the voiceprint features of the machinery and equipment; Taking the start time of the voiceprint feature of the mechanical device as the time reference, recording the time when the voiceprint feature appears after the time reference, and obtaining the maximum time interval, the minimum time interval, and the average time interval; If the voiceprint feature of the mechanical device is not detected again within a preset multiple of the maximum time interval, the last recorded voiceprint feature occurrence time is determined as the voiceprint feature stop time of the mechanical device; and the voiceprint feature start time and voiceprint feature stop time are used as the operation start and end time of the mechanical device; Determining a process identification result based on the mechanical equipment classification result and a preset mapping relationship between the mechanical equipment and the process; Based on the process identification result and the operation start and end time of each mechanical equipment, the process type and process start and end time are recalibrated respectively to obtain the recalibrated process type and process start and end time.

5. The method according to claim 4, characterized in that The performing frequency domain filtering on the spectral features of the audio to obtain filtered spectral features includes: Collect background noise audio data during tunnel construction; Converting the background noise audio data into a frequency domain to obtain a frequency component of the background noise; The frequency component of the background noise is subtracted from the frequency spectrum feature of the audio to be filtered to obtain a filtered frequency spectrum feature.

6. The method according to claim 4, characterized in that The method further comprises: Combine the process type and process start and end time to form process data; The merged process data is output in a preset format, where the preset format is at least one of the following: table and text.

7. The method according to claim 1, characterized in that The digital audio signal of the construction site collected by the audio sensor is obtained; The digital audio signal of the construction site collected by the audio sensor is obtained at a preset interval.

8. A system for automatically extracting tunnel construction process data, characterized in that: The system includes: an audio sensor, an installation position configuration unit and a process identification platform; The installation position configuration unit is used to configure the construction position for installing the audio sensor; The audio sensor includes a microphone, an AD conversion unit and a communication unit; The microphone is installed at a corresponding position on the construction site according to the configured construction location, and is used to pick up the analog sound signal on the construction site; The AD conversion unit is used to convert the analog sound signal into a digital audio signal; The communication unit is used to transmit the digital audio signal to the process identification platform; The process identification platform is used to execute the method for automatically extracting tunnel engineering construction process data as described in any one of claims 1-7.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; The memory is used to store computer programs; The processor is configured to implement the method steps described in any one of claims 1 to 7 when executing the program stored in the memory.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps according to any one of claims 1 to 7 are implemented.

Images