An early warning classification system and method based on intelligent recognition

Abstract

The present invention relates to an early warning classification system and method based on intelligent recognition, which includes: a first data processing unit for grouping pre-entered reference data and establishing a change trend reference function of specific environmental parameters at a specified construction location during construction based on the continuity of time points for the grouped reference data; a second data processing unit for analyzing construction risks by establishing change trend functions of multiple groups of single environmental parameters; a third data processing unit for importing multi-source environmental parameters collected by a number of data monitoring units into a construction risk identification model, thereby determining construction risks based on the index data output by the model, and simultaneously verifying the construction risk information analyzed by the second data processing unit. The third data processing unit calibrates the level of construction risks according to the classification early warning threshold range in which the output index data is located, and performs a classification early warning prompt based on the calibration result.

Description

Technical Field

[0001] The present invention relates to the technical field of construction risk control, and particularly to an early warning grading system and method based on intelligent recognition. Background Art

[0002] With the continuous development of computer software and the Internet, it has become an industry trend to evaluate the risk management of tunnel construction quality and safety in the form of computer software. There are many uncertain factors in the construction risk of large tunnel projects, and the construction risk is characterized by diversity and randomness. Construction projects are restricted by various conditions and have a great impact on the environment and society. At present, there is no effective risk assessment system in the field of engineering construction. Therefore, there is an urgent need in the field of engineering construction to build an online risk assessment system for the tunnel construction quality and safety in construction projects. This system can analyze and evaluate the technical risks of large tunnel projects, so as to integrate reasonable and efficient risk control measures, change the passive emergency risk management mode into an active early warning risk management mode, and effectively reduce the engineering technical risks. In addition, although image acquisition devices have been introduced at the construction site to monitor, identify and evaluate the risks of tunnel construction, the existing camera devices can only be used as terminal data acquisition devices for remote manual monitoring by construction workers. Image acquisition devices such as cameras can only automatically identify and mark abnormal phenomena with obvious features according to pre-set reference samples. They cannot automatically learn based on monitoring information and manual certification results, and they cannot automatically identify abnormalities and analyze risks for abnormal images without reference samples through image processing.

[0003] The patent document with the publication number CN113738448A discloses a multi-source grading information intelligent monitoring and early warning method for water inrush from surrounding rock mass during mining. During the mining process, a microseismic monitoring system, a mine geological radar monitoring system, and an infrared radiation monitoring system are jointly used to monitor and grade and early warn the water inrush situation at three spatial positions, near, far and adjacent, of the surrounding rock mass during mining. The multi-source information grading early warning indexes and corresponding discrimination thresholds are determined. At the same time, the water inrush in the monitored area on the surface of the roof and floor is collected and sampled, and the water inrush source is analyzed to assist in judging whether the water-conducting fissures in the aquifers of the roof and floor are connected, so as to form a multi-source information real-time grading information intelligent monitoring and early warning system for water inrush from surrounding rock mass during mining. Although the early warning method involved in this patent improves the accuracy of judging the water inrush phenomenon of the surrounding rock mass by judging the thresholds of multiple monitoring information, the comparison method still uses the data at a single time point for comparison, and the comparison result is one-sided and single.

[0004] Therefore, in order to improve the accuracy and effectiveness of construction risk identification during tunnel construction, especially for the effectiveness of identification results and construction risk prediction, this application uses multiple sets of continuous data on the time axis to establish a change trend curve, and judges whether there is a construction risk at the construction site by comparing the change trend curves, eliminating the misguidance of the judgment result caused by single abnormal data unrelated to construction risks, so that the early warning classification system provided by this application can intelligently identify construction risks and update risk level data according to real-time data without manual operation.

[0005] In addition, on the one hand, there are differences in the understanding of those skilled in the art; on the other hand, although the inventor studied a large number of documents and patents when making this invention, all details and content are not listed in detail due to space limitations. However, this does not mean that this invention does not possess the features of these prior arts. On the contrary, this invention already possesses all the features of the prior arts, and the applicant reserves the right to add relevant prior arts in the background art. Summary of the Invention

[0006] Currently, the prior art usually uses the environmental parameters at a single time point collected in real time to analyze and judge the engineering state of the construction site. However, the environmental parameters at a single time point are easily affected by factors such as personnel, machinery and equipment, environment, and structure, resulting in monitoring errors at a certain moment or abnormal fluctuations in single data values due to changes in the external environment. For example, when a heavy vehicle passes over the ground above the tunnel, the tunnel will experience abnormal vibrations following the passing of the vehicle, and it may also cause a sudden increase in the settlement data at a single time point collected. However, such abnormalities do not necessarily lead to construction risks, and whether there are potential construction risk hazards in the future also needs to be verified by a continuous set of time data with an extended time line. At this time, the data at a single time point no longer has the universality of risk identification. In addition, the change in the existing environmental parameters may also be caused by a temporary mutation in the external environment of the tunnel. For example, when there is a short-term heavy rain weather at the construction site, the air humidity at the construction site and the humidity of the tunnel wall will increase rapidly in a short time. At this time, the abnormality of the environmental parameters cannot be used as the basis for the construction risk of water seepage in the tunnel. Therefore, in order to eliminate the defect that the environmental parameters at the construction site are identified and judged to have construction risks or may cause construction risks due to data mutations at a single time point caused by external factors, the present invention uses multiple sets of continuous data to establish a change trend curve to judge construction risks. In particular, this application also verifies the construction risks determined by the change trend curve by establishing a construction risk identification model, thereby improving the identification accuracy of the system. In addition, the established construction risk identification model can directly output an index data, and calibrate the level of the identified construction risk through the grading early warning threshold range where the index data is located.

[0007] In view of the deficiencies of the prior art, the technical solution of the present invention provides an early warning classification system based on intelligent recognition, which includes a first data processing unit for grouping pre-entered reference data and establishing a change trend reference function of specific environmental parameters at a specified construction location during construction according to the continuity of time points for the grouped reference data; a second data processing unit for analyzing construction risks by establishing change trend functions of multiple groups of single environmental parameters; and a third data processing unit for importing multi-source environmental parameters collected by several data monitoring units into a construction risk identification model, thereby determining construction risks based on the index data output by the model and synchronously verifying the construction risk information analyzed by the second data processing unit. The third data processing unit also calibrates the level of construction risks according to the classification early warning threshold range in which the output index data is located, and thus gives a classification early warning prompt according to the calibration result. Its advantages are as follows. The present invention changes the existing point data comparison method for construction risk analysis to a curve data comparison method, uses multiple groups of environmental parameters continuous on the time axis to establish corresponding change trend functions, thereby generating corresponding change trend curves, and determines whether there are abnormal fluctuations or changes in environmental parameters by comparing the change trend curve generated by the monitored environmental parameters with the change trend reference curve. Thus, in the case of eliminating individual abnormally monitored environmental parameters, the system can effectively analyze whether there are construction risks at the construction site. In addition, in order to improve the reliability of the analysis results, the system also verifies the above construction risk results by combining and analyzing multi-source environmental parameters. At the same time, the system calibrates the construction risk level by setting index weights, and updates the calibration information for dividing risk levels according to the verification results, thereby improving and correcting the risk level standard established in advance according to engineering data and expert opinions.

[0008] According to a preferred embodiment, the second data processing unit classifies and fits the environmental parameters corresponding to multiple consecutive time points collected by the data monitoring unit, thereby obtaining real-time change trend functions of multiple single environmental parameters within this time period. Furthermore, it predicts the change situation of environmental parameters in the next time period based on independent change trend functions established for different environmental parameters and establishes a change trend prediction function using the predicted environmental parameters, and analyzes whether there are construction risks at the construction site by comparing the real-time change trend function, the change trend prediction function and the change trend reference function synchronously.

[0009] According to a preferred embodiment, the change trend reference function, the change trend real-time function, and the change trend prediction function can all be characterized by curve images; the curve image corresponding to the change trend reference function includes a minimum value curve image, a maximum value curve image, and a median value curve image respectively established by the first data processing unit using reference data for multiple groups at the same time point within a time period.

[0010] According to a preferred embodiment, the construction risk identification model is constructed by combining multi-source environmental parameters according to their respective different index weights, and the third data processing unit generates a risk prediction value by summing the index data corresponding to the multi-source environmental parameters, so as to calibrate the level of construction risk by judging the grading warning range where the risk prediction value is located.

[0011] According to a preferred embodiment, the comparison between the change trend real-time function, the change trend prediction function and the change trend reference function is a comparison between multiple curve images, so as to analyze the construction risk at the construction site by comparing the slope changes of the curves and the positional relationship between multiple curves.

[0012] According to a preferred embodiment, when the change trend real-time curve and / or the change trend prediction curve intersect with the minimum value curve or the maximum value curve, the second data processing unit determines that there is a construction risk at the construction site, and traces and manually verifies the risk source according to the data monitoring unit where the abnormal environmental parameter appears.

[0013] According to a preferred embodiment, when the change trend real-time curve and / or the change trend prediction curve remains between the minimum value curve and the maximum value curve, the second data processing unit analyzes whether there is a potential construction risk by comparing the change in the slope of the curve, where

[0014] When the slope of the change trend prediction curve develops in a continuously increasing or continuously decreasing direction relative to the slopes of the minimum value curve and the maximum value curve, the second data processing unit determines that there is a potential construction risk at the construction site.

[0015] According to a preferred embodiment, the third data processing unit adjusts the index weights corresponding to different environmental parameters following the construction risk information output by the system, the construction risks confirmed manually at the construction site, and the changes in the construction environment during the construction process.

[0016] The technical solution of the present invention also provides an early warning grading method based on intelligent identification, which at least includes the following steps:

[0017] Group the pre - entered reference data, and establish a reference function for the change trend of specific environmental parameters at the specified construction location during the construction period based on the continuity of time points for the grouped reference data;

[0018] Analyze the construction risks at the specified construction location by establishing change trend functions for multiple single environmental parameters;

[0019] Import the multi - source environmental parameters collected by several data monitoring units into the construction risk identification model, thereby determining the construction risks based on the index data output by the model, and synchronously verifying the obtained construction risk information;

[0020] Calibrate the level of construction risks according to the grading warning threshold range where the output index data is located, and thus give a grading warning prompt according to the calibration result.

[0021] According to a preferred embodiment, the change trend function is obtained by classifying and fitting the environmental parameters corresponding to multiple consecutive time points collected by the data monitoring unit, so as to obtain the real - time change trend function of multiple single environmental parameters within this time period. Furthermore, the change situation of environmental parameters in the next time period is predicted based on the independent change trend functions established for different environmental parameters, and whether there are construction risks at the construction site is analyzed by comparing the real - time change trend function, the change trend prediction function and the change trend reference function in the same period. Brief Description of the Drawings

[0022] Figure 1 is a topology diagram of a preferred early warning grading system and method based on intelligent recognition proposed by the present invention;

[0023] Figure 2 is a schematic work flow diagram of a preferred early warning grading system and method based on intelligent recognition proposed by the present invention.

[0024] List of Reference Numerals

[0025] 1: First data processing unit; 2: Second data processing unit; 3: Third data processing unit; 4: Data monitoring unit. Detailed Description of the Preferred Embodiment

[0026] The following is a detailed description with reference to the drawings.

[0027] Embodiment 1

[0028] This application provides an early warning grading system based on intelligent recognition, which may include a first data processing unit 1, a second data processing unit 2, a third data processing unit 3 and a data monitoring unit 4.

[0029] According to Figure 1In a specific embodiment shown, the first data processing unit 1 classifies the pre-collected engineering data and expert opinions, and stores different data in different storage spaces in groups, so as to establish a change trend reference function based on the parameter data that changes orderly along the time axis stored in a single storage space. The second data processing unit 2 uses the environmental parameters collected by the data monitoring unit 4 as the data basis, and thus establishes change trend functions for multiple groups of single environmental parameters, and analyzes the construction risks that may exist or may evolve at the construction site by comparing with the change trend reference function established by the first data processing unit 1. The third data processing unit 3 integrates and configures weights for multi-source environmental parameters by establishing a construction risk identification model, and thus outputs a risk identification result based on multi-source environmental parameters, and then uses this risk identification result to verify the construction risks analyzed by the second data processing unit 2, so as to improve the accuracy of the system in identifying construction risks. In addition, the third data processing unit 3 can also use weight indicators to calculate the index data for grading and calibration of different environmental parameters for construction risks, and thus obtain the level of construction risks by integrating multi-source environmental parameters, which is convenient for construction personnel to make different emergency responses according to the risk level.

[0030] Preferably, the first data processing unit 1 arranges the data corresponding to different time points in an orderly manner according to the time continuity of the same type of reference data collected by the same monitoring device, so as to obtain multiple groups of reference data that are continuous in time points, and thus uses the grouped reference data to establish a change trend curve function of specific environmental parameters monitored by the monitoring device set at the specified construction location during construction. Preferably, the specific environmental parameters refer to the same type of environmental parameters collected by the same device at different time points, for example, the deformation parameters, displacement parameters or cracking parameters of tunnel support, etc.

[0031] Preferably, the second data processing unit 2 performs classification fitting on the environmental parameters corresponding to multiple consecutive time points collected by the data monitoring unit 4, so as to obtain a real-time function of the change trend of multiple single environmental parameters within this time period. The second data processing unit 2 can classify and organize the real-time environmental parameters collected by multiple data monitoring units 4 over time, so as to form environmental parameter sets for multiple time periods of different data monitoring units 4. Preferably, the second data processing unit 2 constructs a real-time function of the change trend of the environmental parameter over time by fitting multiple environmental parameters that change with time of at least one time period of a single data monitoring unit 4. Preferably, the fitting is to establish the relationship between the environmental parameter and time according to the change trends of multiple groups of environmental parameters that are continuous in time, so as to facilitate obtaining the specific values of the environmental parameter at different time points. Preferably, the second data processing unit 2 can predict the change of the environmental parameter in the next time period according to the independent change trend functions established for different environmental parameters, so as to generate corresponding change trend prediction functions, and analyze whether there is construction risk at the construction site by comparing the change trend real-time function, the change trend prediction function and the change trend reference function synchronously. Preferably, the change trend reference function, the change trend real-time function and the change trend prediction function can all be represented by curve images. Synchronous comparison means comparing the change trend curve of the change trend real-time function or the change trend prediction function in a specified time period with the change trend curve of the change trend reference function formulated in advance within the above-mentioned specified time period, so as to analyze the differences between the development trends of the change trend real-time curve and the change trend prediction curve and the prediction curve and the sources of the differences, so as to better determine the construction risks that may exist or may evolve at the construction site. Preferably, the curve image corresponding to the change trend reference function includes the minimum value curve image, the maximum value curve image and the median curve image established by the first data processing unit 1 using the minimum value, the maximum value and the median value of multiple groups of the same time point of the environmental parameter within a time period.

[0032] Preferably, the comparison between the real-time change trend function, the predicted change trend function and the reference change trend function is the comparison of curves between multiple curve images, so as to analyze the construction risks at the construction site by comparing the slope changes of the curves and the positional relationship between the curves. Specifically, when there are intersections between the real-time change trend curve and / or the predicted change trend curve and the minimum value curve or the maximum value curve, the second data processing unit 2 determines that there are construction risks at the construction site, and traces and manually verifies the risk sources according to the data monitoring unit 4 where the abnormal environmental parameters appear. When the real-time change trend curve and / or the predicted change trend curve remains between the minimum value curve and the maximum value curve, the second data processing unit 2 analyzes whether there are potential construction risks by comparing the changes in the slopes of the curves. When the slope of the predicted change trend curve develops in a continuously increasing or continuously decreasing direction relative to the slopes of the minimum value curve and the maximum value curve, the second data processing unit 2 determines that there are potential construction risks at the construction site. For example, the slopes of the maximum value curve and the minimum value curve representing the tunnel settlement show a gradually decreasing trend, and the decreasing trend is that the slope decreases by 0.01 per hour. When the slopes of the real-time change trend curve and the predicted change trend curve also decrease with a reduction value of about 0.01, it is determined that the actual settlement is normal. When the slopes of the real-time change trend curve and the predicted change trend curve continuously decrease with a reduction value of about 0.02 or 0.05, it can be foreseen that the predicted change trend curve will intersect with the minimum value curve or the maximum value curve at a certain moment in the future. At this time, it can be judged that the change of the environmental parameter is abnormal, and there are construction risks at the construction site at a certain moment in the future. Moreover, the curve records a continuous change process, eliminating the influence of individual invalid data on the recognition result and increasing the accuracy of system recognition.

[0033] Preferably, the construction risk identification model is constructed by combining multi-source environmental parameters according to their respective different index weights. The third data processing unit 3 generates a risk prediction value by summing the index data corresponding to the multi-source environmental parameters, and then calibrates the level of construction risk by judging the classification warning range where the risk prediction value is located. Preferably, the framework of the construction risk identification model is the index weights corresponding to different environmental parameters. The system constructs a complete construction risk identification model by combining multi-source environmental parameters with their respective corresponding index weights. The result output by the construction risk identification model is the index data obtained by adding the products of the multi-source environmental parameters and their respective index weights. The third data processing unit 3 analyzes the result output by the construction risk identification model with the pre-established risk level standard, so as to judge whether the calculated index data is within the numerical range of the risk level, and calibrates the level of the actually monitored construction risk according to the risk warning threshold range where the index data is located, and then makes a classification warning prompt according to the calibration result. The third data processing unit 3 can also adjust the index weights corresponding to different environmental parameters according to the construction risk information output by the system, the construction risks confirmed manually at the construction site, and the changes in the construction environment during the construction process, so as to change the correlation between different environmental parameters in different construction environments and whether the construction site is safe.

[0034] The present invention uses multiple sets of continuous environmental parameters to establish a change trend curve for judging construction risks. In particular, the present application also verifies the construction risks judged by the change trend curve by establishing a construction risk identification model, thereby improving the identification accuracy of the system. In addition, the established construction risk identification model can directly output an index data, and calibrate the level of the identified construction risk through the grading early warning threshold range where the index data is located. For example: The construction risk level threshold range established by the system according to engineering data and expert opinions can be multi-segment graded according to the size of the index data. For example, when the index data is less than 100, it is determined that there is temporarily no construction risk; when the index data is between 101 and 103, it is determined that there is a first-level construction risk; when the index data is between 104 and 106, it is determined that there is a second-level construction risk; when the index data is between 107 and 109, it is determined that there is a second-level construction risk; The system establishes a complete construction risk level standard and threshold range according to the above segmentation setting method. Therefore, when the system performs identification and calibration, it uses the threshold range where the index data is located to calibrate the risk level corresponding to the construction risk. Preferably, the third data processing unit 3 can also adjust the index weights of different environmental parameters according to changes in human factors, machine equipment factors, environmental factors, and structural factors. For example, there are heavy vehicles driving on the road directly above the tunnel during a certain period of time, and the settlement data, support deformation data, and vibration conditions of the tunnel wall of the tunnel during this period will all change. However, such changes do not necessarily cause construction risks, and they only change the fluctuation or change trend of a single environmental parameter. Therefore, the index weights of such environmental parameters at this time need to be modified according to the actual situation. In addition, when mechanical equipment is damaged, the continued use state of the mechanical equipment may also change to a certain extent, and the abnormal jitter generated during its operation may change in frequency and amplitude. Such changes may also cause changes in environmental parameters, but the short-term changes in environmental parameters cannot be used as an effective judgment basis. The prior art usually uses the environmental parameters at a time point to identify construction risks, and it cannot effectively eliminate such invalid data. The generation of actual construction risks is a gradually changing process, usually caused by qualitative changes leading to quantitative changes. Therefore, the present application judges whether there are construction risks at the construction site by eliminating single invalid data and using the continuous changes of periodic data, which has data accuracy and can also effectively judge which stage the existing construction risks are in, so as to facilitate construction personnel to make different treatment operations according to the actual situation.

[0035] Preferably, the environmental parameters collected by the data monitoring unit 4 correspond to the disaster-causing factors that may cause construction risks at the tunnel construction site. The data monitoring unit 4 can monitor the settlement, displacement, deformation of the tunnel support, the cracking and collapse of the tunnel wall, etc.

[0036] Embodiment 2

[0037] This embodiment is a further improvement of Embodiment 1, and repeated content will not be elaborated.

[0038] Abundant monitoring devices, instruments, and numerous monitoring professionals are spread throughout the construction impact area. There are a large number of monitoring devices inside the tunnel construction. The data collection frequency of the devices is high, usually collecting monitoring data 24 hours a day, all-weather. However, as the construction progresses, the application of more and more monitoring devices makes the amount of monitoring data gradually huge. The traditional way of using monitoring data is relatively single. Generally, computer technology is used to analyze the existing monitoring data and then handed over to experts with rich engineering experience for judgment to obtain the entire evaluation result. Such a relatively single index and the result of human decision-making have certain subjectivity and ambiguity. The work experience and capabilities of different construction personnel vary, the working states of different mechanical equipment are inconsistent, and the impacts of different surrounding environments are different. More importantly, the structure inside the tunnel is constantly changing. The changes in these factors will all lead to errors in the judgment of monitoring information, possibly resulting in misjudgment and false alarms. The efficiency of safety evaluation at the engineering construction site is also particularly important. The existing technology has the defect of judging safety only from the standard of whether a single index exceeds the threshold, ignoring the mutual connection between multiple indicators. Such existing technology often affects the accuracy of the evaluation, thus leaving potential safety hazards for the construction process.

[0039] Based on the current engineering monitoring status, a comprehensive and integrated evaluation of multi-index, multi-level, and heterogeneous data in tunnel construction can be carried out through the method of multi-source data fusion. Through the analysis of existing engineering monitoring data and the research of relevant theories, a data fusion model is proposed. Based on this model, data with different dimensions can be unified and integrated to achieve a comprehensive and integrated evaluation of tunnel construction safety. For this purpose, the system needs to select refined monitoring data, comprehensively analyze the safety factors involved in tunnel construction, predict its impact on construction safety, improve the accuracy of safety evaluation, and further adopt appropriate data to improve the existing multi-source data fusion model to improve the efficiency and accuracy of safety warning.

[0040] Data fusion is usually defined as a technology that can process information collected from multiple levels, angles, and aspects and transform it into information that can effectively provide support for assisting people or making automatic decisions. The sources of data can be diverse, including not only data collected by sensors, database data, numerical simulation data, but also various other forms, such as numbers, texts, images, knowledge and experience, video monitoring, etc. Data fusion has the advantages of improving detection efficiency, reliability, credibility, and reducing data ambiguity errors, and data fusion has played an excellent role in the fields of automation, intelligence, etc. The identification of multi-valent monitoring indicators is the basis for multi-source data fusion in tunnel construction safety assessment. As the information source for tunnel construction safety assessment, the monitoring data must be accurate data, which not only affects the selection of the multi-source data fusion model but also determines the accuracy of the evaluation results to a certain extent.

[0041] Preferably, the monitoring of tunnel construction safety is spatially divided into the surface of the tunnel support structure to the surrounding environment, buildings, and pipelines, all of which may affect the progress of tunnel excavation support and construction safety; in terms of construction procedures, from drilling and installing pipe sheds, setting rock bolts, to gradually excavating the tunnel, primary support, and then pouring concrete for the secondary lining will cause stress settlement and deformation in the tunnel. This application takes the human-machine-environment-structure in the tunnel safety assessment project as the monitoring information indicators, and constructs the evaluation information indicators of multi-source data fusion through hierarchical analysis. Information acquisition, identification, and fusion are carried out on the monitoring object information in each link, so as to more accurately reflect the safety state information of key positions during the tunnel construction process.

[0042] Preferably, the disaster-causing factors of risks can be divided into human factors, machine equipment factors, environmental factors, and structural factors, etc.

[0043] (1) Human factors

[0044] ① Fatigue state

[0045] The unsafe behaviors of construction workers have been regarded as one of the main causes of construction accidents, and the psychological state of construction workers during the construction process directly affects their behaviors. According to the actual situation of the project construction and the availability of data, this application selects the heart rate index as the data for evaluating the fatigue state of construction workers. The heart rate data of construction workers can be collected by a wearable smart bracelet, and the change of heart rate data can be read in real time on a smartphone application program, so as to use the fatigue degree grading method with dynamic heart rate as the index to determine the fatigue state of on-site construction workers during operation.

[0046] ② Distance from the hazard source

[0047] Video monitoring technology is adopted at the construction site to record the safety status of on-site operations and conduct real-time visual positioning of construction workers, preventing construction workers from getting too close to hazard sources and causing dangerous accidents. During the tunnel construction process, managers can use video monitoring to grasp the behavior status of construction workers and ensure the personal safety of construction workers and prevent unsafe behaviors through observation.

[0048] ③ Relevant engineering experience

[0049] The technical level, professional quality, experience, etc. of the management personnel, special operation personnel and construction operation personnel at the construction site can be quantitatively reflected through their own engineering experience.

[0050] (2) Mechanical equipment

[0051] Generally, tunnel construction mechanical equipment includes pneumatic rock drills, double-fluid grouting pumps, excavators, concrete pumps, etc. Before the tunnel soil excavation, a rock drill is used to drill holes, and a grouting pump is used for grouting reinforcement. Then, the soil excavation is carried out, and finally, concrete is poured to ensure the stability of the surrounding rock after excavation. Mechanical equipment is an important factor in ensuring the stability of tunnel excavation. To identify the monitoring information sources for mechanical equipment, it is necessary to focus on whether the parameters of mechanical equipment are within the corresponding safety thresholds during operation and pay attention to the maintenance status of mechanical equipment. Therefore, the corresponding monitoring information sources are mainly as follows:

[0052] ① Working air pressure. The working air pressure affects the working performance of the machine and the service life of parts. During the actual operation of the equipment, too high air pressure of the rock drill will cause excessive mechanical vibration, resulting in increased workpiece wear, and too low pressure will cause the machine to not reach the best working state. To measure the air pressure status of the rock drill, it is necessary to regularly record the air pressure value during machine operation; the grouting pump pressure is an important factor affecting construction grouting. Too high grouting pressure may be caused by blocked pipelines or mixers, and too low pressure may be due to slurry leakage or the slurry flowing away through certain underground pipelines. The grouting pump pressure status is also a key object for monitoring construction equipment;

[0053] The concrete pump has 3 main systems: a mixing system, a reversing system, and a main pumping system. The main pump uses constant power regulation, which means that if the pressure in the pipeline increases, the main pump will automatically reduce the displacement to ensure a constant power value, prevent the motor from overheating, and thus improve the power utilization rate. Therefore, an appropriate pressure value is an important factor in providing engineering efficiency.

[0054] ②Mechanical wear. Tunnel construction is different from other industries. It is mainly underground operation with poor on-site conditions and is greatly affected by rain conditions. Shenzhen is an area with abundant rainfall. Due to the influence of natural climate, engineering machinery and equipment are extremely prone to wear and rust, which not only reduces the service life of engineering machinery and equipment, but also is more likely to cause dangerous accidents. Mechanical wear needs to be strictly controlled within the maximum limit of wear, especially for vulnerable parts, and those showing wear should be replaced in a timely manner.

[0055] ③Mechanical failures. Tunnel construction machinery is often used under harsh conditions. If equipment maintenance cannot keep up, it will lead to a high equipment failure rate, hinder the construction progress, and shorten the service life of the equipment. Equipment management depends on system guarantee. In order to ensure the safe construction of equipment in the best state and reduce the equipment failure rate, highly qualified professional and technical personnel are often required for operation and maintenance.

[0056] (3) Environmental factors

[0057] Excessive changes in the surrounding environment will lead to safety warnings, suspension of work, and maintenance in tunnel construction. Therefore, for the surrounding environment information of tunnel construction, it is also necessary to analyze and identify the main safety risk factors.

[0058] ①Surrounding ground surface.

[0059] The excavation of soil mass in tunnel construction will unload the surrounding soil mass, causing ground settlement, ground cracking, and even ground collapse. Whether the surrounding ground surface of the tunnel is safe indicates the safety situation of the construction inside the tunnel. Therefore, the identification of the safety monitoring information of the surrounding bottom edge is mainly studied from the following aspects:

[0060] Surrounding ground settlement. According to the geological survey report, the geology within the scope of this research project is complex, and tunnel construction will surely have a certain impact on the nearby ground surface and the surrounding environment. And the surrounding ground settlement is an important factor in judging the safety of the construction inside the tunnel. Timely and rigorous monitoring of the ground settlement information during the construction process to provide feedback for guiding the construction not only ensures the safe and stable progress of tunnel construction, but also effectively controls the surrounding environment, reduces the impact caused by construction, and keeps the construction within the safe range. Similarly, the monitoring of the surrounding ground settlement of the tunnel also needs to be carried out from two aspects: the cumulative value of the ground settlement displacement and the change rate.

[0061] Surrounding road surface cracking. Tunnel construction is located in the center or on both sides of the existing urban roads. During the tunnel construction process, it will surely have a certain impact on the surrounding road surface, and there are risks such as large road deformation and affecting safe use. Therefore, the engineering monitoring should conduct a detailed investigation and evidence collection of the surrounding road surface, and focus on the cracking situation of the surrounding road surface in a targeted manner to ensure the smooth progress of tunnel construction.

[0062] Groundwater level. According to the engineering investigation report, the groundwater is abundant around the tunnel construction. Some of the rock and soil are silt layers, which are likely to cause sand gushing during tunnel construction, leading to ground settlement around and even affecting the construction safety. The dynamic change of the groundwater level is closely related to the atmospheric rainfall. The peak value of the water level is basically consistent with the peak and trough of the rainfall. When the rainy season begins, the water level rises, and when the dry season begins, the water level drops. Therefore, the monitoring data is closely linked to the rainfall during construction to timely understand the water level change of the surrounding environment of the tunnel. According to the rainfall situation, the focus of the monitoring work is adjusted appropriately, closely paying attention to the change of the groundwater, providing the water level monitoring parameters, and escorting the safe progress of the project.

[0063] ② Important pipelines. Since the tunnel construction is located below the ground surface, it often passes through various important pipelines, especially the important pipelines within the excavation range near the tunnel, such as cables, natural gas pipelines, water pipes, etc. During the tunnel excavation process, the displacement and unloading of the soil body will cause the settlement and cracking of the pipelines, which will not only cause great economic losses but also affect the safety of the tunnel construction workers. Therefore, the identification of the safety monitoring information sources of the underground pipelines around the tunnel is mainly carried out from the following aspects:

[0064] Pipeline settlement. The tunnel construction excavation will cause the displacement and unloading of the surrounding soil body, and then the balance of the stress state of the pipeline is damaged, resulting in pipeline settlement. According to the different materials of the important pipelines around the tunnel construction, they are divided into rigid pipelines (sewage, gas) and flexible pipelines (cables, information pipelines). Due to the inconsistent stiffness of the pipelines, their settlement warning values are also different. Therefore, the research on pipeline settlement mainly focuses on two aspects: the cumulative control value of settlement and the change rate.

[0065] Pipeline leakage. The common underground pipelines in the city include cables, natural gas pipelines, and water pipes. The damage of these three types of pipelines will have a significant impact on the construction workers and the lives of the surrounding urban residents. For example, the damage of the cable will cause the lack of electricity in the construction area, the machinery cannot work normally, affecting the construction efficiency, and in more serious cases, it will cause construction safety problems; the leakage of the natural gas pipeline will pose a major danger to the safety of the construction workers and the surrounding residents; and the rupture of the water supply and drainage pipeline is extremely likely to cause the instability of the surrounding rock of the tunnel and even the occurrence of a collapse accident. Therefore, during the daily work inspection, attention should be paid to whether there are abnormal smells or abnormal water leakage in the tunnel construction. Once an abnormal situation is found, it should be reported and disposed of in a timely manner to ensure the safety of the construction and the surrounding environment.

[0066] (4) Structural factors

[0067] The identification of the monitoring information of the tunnel construction support structure needs to consider the possible situations of the net convergence, crown settlement, and the cracking and leakage of the tunnel structure that may occur in the surrounding rock support structure. Combining with the monitoring of the engineering project, the monitoring information indexes of the tunnel construction structure are as follows:

[0068] ①Convergence of the clearance. The wireless laser sensor is used to monitor the convergence of the tunnel clearance in real time to detect the change of the surrounding rock structure of the tunnel. After excavation, the measuring points are installed and numbered quickly as required. The initial readings are taken in time after excavation. The measuring points are firm and reliable, easy to identify and properly protected. Therefore, the monitoring indexes need to consider two aspects: the cumulative value and the change rate.

[0069] ②Settlement of the crown. The settlement of the tunnel crown can be jointly characterized by the laser sensor and the inclinometer sensor during the tunnel construction excavation. There is a certain unloading situation, and the tunnel settlement will also occur. The tunnel settlement greatly reflects the safety status of the in-tunnel construction. Therefore, it is also necessary to monitor the cumulative value of its deformation and the change rate during the construction.

[0070] ③Structural cracking. The cracking of the tunnel support structure is related to various factors, including structural materials, construction techniques or excessive structural stress, which will all cause certain cracks. Therefore, the observation of the cracks should be carefully inspected and excluded, and the observation records should be well done.

[0071] ④Structural leakage. The groundwater level in the tunnel construction affected area is high and the soil property is soft. If there is too much rainfall, the continuous development of rainwater will cause the tunnel structure to leak. If not disposed of, the leakage will increase continuously, and then may evolve into water gushing, the tunnel structure will be damaged, affecting the construction safety, and more seriously, it will cause casualties.

[0072] Through the identification and analysis of the above monitoring indexes, the above monitoring indexes can be further normalized to make them dimensionless numbers that are more comparable. When the index has a positive effect on the construction risk, that is, the larger the value, the more likely a risk accident will occur. Based on the data fusion theory, the basic probability assignment of the monitoring indexes is constructed with the fuzzy matter element to form an improved D-S evidence fusion model. In the evaluation process of this model, data with different dimensions, qualitative, quantitative and heterogeneous can be fused. The safety state of the tunnel construction is evaluated through the multi-measuring-point-one-monitoring information-three-level index-two-level index-integral data fusion process. In this process, not only the overall evaluation of the tunnel excavation can be obtained, but also the safety state of a specific measuring point can be located to evaluate, making the evaluation more targeted. This model synthesizes almost all relevant information of the on-site construction monitoring, which can avoid the inaccuracy of the single information evaluation. The inaccuracy of the evaluation result is reduced after the hierarchical data fusion, which is more in line with the engineering practice.

[0073] Example 3

[0074] This embodiment provides an early warning grading method based on intelligent recognition, which at least includes the following steps:

[0075] Group the pre - entered reference data, and establish a reference function for the change trend of specific environmental parameters at a specified construction location during construction according to the continuity of time points for the grouped reference data;

[0076] Analyze the construction risks at the specified construction location by establishing change trend functions for multiple single environmental parameters. Among them,

[0077] The change trend function is obtained by classifying and fitting the environmental parameters corresponding to multiple consecutive time points collected by the data monitoring unit 4, so as to obtain the real - time change trend function of multiple single environmental parameters within this time period. Furthermore, the change situation of environmental parameters in the next time period is predicted based on the independent change trend functions established for different environmental parameters. And the construction risks at the construction site are analyzed by comparing the real - time change trend function, the change trend prediction function and the reference function of the change trend synchronously;

[0078] Import the multi - source environmental parameters collected by several data monitoring units 4 into the construction risk identification model, so as to judge the construction risks according to the index data output by the model, and verify the obtained construction risk information synchronously;

[0079] Judge the level of construction risks according to the grading warning threshold range where the output index data is located.

[0080] It should be noted that the above - mentioned specific embodiments are exemplary. Those skilled in the art can come up with various solutions inspired by the disclosed content of the present invention, and these solutions also belong to the disclosed scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the description of the present invention and its drawings are illustrative and do not constitute a limitation on the claims. The protection scope of the present invention is defined by the claims and their equivalents. Throughout the text, the features guided by "preferably" are only an optional way and should not be understood as must - be - set. Therefore, the applicant reserves the right to abandon or delete relevant preferred features at any time.

Claims

1. An early warning grading system based on intelligent recognition, characterized in that, Comprising: A first data processing unit (1) for establishing a risk level standard according to engineering data and expert opinions, grouping pre-entered reference data, and establishing a change trend reference function of specific environmental parameters at a specified construction location during construction based on the continuity of time points for the grouped reference data. The specific environmental parameters refer to the same type of environmental parameters collected by the same device at different time points; A second data processing unit (2) for analyzing construction risks by establishing change trend functions of multiple groups of single specific environmental parameters; A third data processing unit (3) for importing multi-source specific environmental parameters collected by a number of data monitoring units (4) into a construction risk identification model, thereby determining construction risks based on the index data output by the model, synchronously verifying the construction risk information analyzed by the second data processing unit (2), and updating the calibration information for dividing risk levels according to the verification results, so as to improve and correct the risk level standard established in advance according to engineering data and expert opinions; The third data processing unit (3) also calibrates the level of construction risks according to the risk level standard of the index data it outputs, and issues a classified early warning prompt according to the calibration results; The second data processing unit (2) performs classification fitting on the specific environmental parameters corresponding to multiple consecutive time points collected by the data monitoring unit (4) to obtain real-time change trend functions of multiple single specific environmental parameters within this time period, and then predicts the change of specific environmental parameters in the next time period based on the independent change trend functions established for different specific environmental parameters and establishes a change trend prediction function using the predicted specific environmental parameters. And by comparing the real-time change trend function, the change trend prediction function with the change trend reference function in the same period to analyze whether there are construction risks at the construction site; The change trend reference function, the real-time change trend function, and the change trend prediction function can all be represented by curve images; The curve image corresponding to the change trend reference function includes a minimum value curve image, a maximum value curve image, and a median value curve image established by the first data processing unit (1) using the minimum values, maximum values, and median values of multiple groups at the same time point within a time period for the reference data; The third data processing unit (3) adjusts the index weights corresponding to different specific environmental parameters according to the construction risk information output by the system, the construction risks confirmed manually at this construction site, and the changes in the construction environment during construction.

2. The early warning classification system based on intelligent recognition according to claim 1, wherein The construction risk identification model is constructed by combining multi-source specific environmental parameters according to their respective different index weights. The third data processing unit (3) generates a risk prediction value by adding up the index data corresponding to the multi-source specific environmental parameters, and thereby calibrates the level of construction risks by judging the risk level standard of the risk prediction value.

3. The early warning classification system based on intelligent recognition according to claim 2, characterized in that, The comparison between the real-time change trend function, the change trend prediction function and the change trend reference function is a comparison between multiple curve images, so as to analyze the construction risks at the construction site by comparing the slope changes of the curves and the positional relationships between multiple curves.

4. The early warning classification system based on intelligent recognition according to claim 3, wherein, When the real-time change trend curve and / or the change trend prediction curve intersect with the minimum value curve or the maximum value curve, the second data processing unit (2) determines that there are construction risks at the construction site, and traces and manually verifies the risk sources according to the data monitoring unit (4) where abnormal specific environmental parameters occur.

5. The early warning grading system based on intelligent recognition according to claim 4, wherein, When the real-time change trend curve and / or the change trend prediction curve remains between the minimum value curve and the maximum value curve, the second data processing unit (2) analyzes whether there are potential construction risk hazards by comparing the changes in the slopes of the curves, where when the slope of the change trend prediction curve develops in a continuously increasing or continuously decreasing direction relative to the slopes of the minimum value curve and the maximum value curve, the second data processing unit (2) determines that there are potential construction risk hazards at the construction site.

6. A warning classification method based on intelligent recognition, the method uses the warning classification system based on intelligent recognition according to any one of claims 1 to 5, characterized in that, The method at least includes the following steps: Group the pre-entered reference data, and establish a change trend reference function of the specific environmental parameters at the specified construction location during the construction period based on the continuity of time points for the grouped reference data; Analyze the construction risks at the specified construction location by establishing change trend functions for multiple single specific environmental parameters; Import the multi-source specific environmental parameters collected by several data monitoring units (4) into the construction risk identification model, so as to determine the construction risks according to the index data output by the model, and synchronously verify the obtained construction risk information; Calibrate the level of the construction risk according to the risk level standard of the output index data, and thus conduct a classified early warning prompt according to the calibration result.

7. The method according to claim 6, wherein The change trend function is obtained by classifying and fitting the specific environmental parameters corresponding to multiple consecutive time points collected by the data monitoring unit (4), so as to obtain the real-time change trend function of multiple single specific environmental parameters within this time period. Furthermore, the change situation of the specific environmental parameters in the next time period is predicted based on the independent change trend functions established for different specific environmental parameters, and whether there are construction risks at the construction site is analyzed by comparing the real-time change trend function, the change trend prediction function and the change trend reference function synchronously.

Images