Real-time monitoring and intelligent early warning system for manhole cover based on internet of things

Abstract

The application discloses a well lid state real-time monitoring and intelligent early warning system based on an Internet of Things, and belongs to the technical field of monitoring and early warning, and specifically comprises the following steps: integrating various sensors in a well lid monitoring terminal, and synchronously collecting equipment operation state data; marking the location of a well lid on a map according to the equipment operation state, and determining a well lid with a normal equipment state; classifying the well lid with the normal equipment state, and determining the inclination angle threshold, the vibration frequency threshold and the vibration noise threshold of a well lid vibration noise risk, a well lid subsidence risk or a road surface flatness risk according to different well lid types; determining whether a well lid with a failure precursor anomaly appears the well lid vibration noise risk, the well lid subsidence risk or the road surface flatness risk; simultaneously monitoring the abnormal displacement and water immersion conditions of the well lid; grading the intelligent well lid according to the determination result, and correspondingly triggering different level early warning signals; and pushing early warning information and operation and maintenance work orders according to the different level early warning signals.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of monitoring alarm, in particular to a real-time monitoring and intelligent early warning system for the state of a manhole cover based on the Internet of Things. BACKGROUND

[0002] In the field of urban infrastructure operation and maintenance, manhole covers, as key nodes of underground pipe network systems, have the characteristics of large quantity, wide distribution and complex working conditions. Their operating state is directly related to road traffic safety and residents' living environment. With the penetration of Internet of Things technology in the field of infrastructure monitoring, existing manhole cover monitoring systems based on the Internet of Things gradually replace the traditional mode. The core architecture includes three modules: front-end monitoring terminal, low-power data transmission network and back-end management platform. The front-end monitoring terminal integrates sensing devices such as vibration sensors, tilt sensors and noise sensors, which can collect real-time operating state data such as vibration frequency, tilt angle and vibration noise of the manhole cover. The data transmission link uses NB-IoT, LoRa and other low-power wide-area network technologies to realize stable uploading of monitoring data to the back-end platform, adapting to the scene requirements of scattered deployment of urban manhole covers. The back-end management platform has basic data storage and threshold comparison functions, which can determine whether the manhole cover is abnormal through a simple preset threshold. Some systems also support marking the manhole cover position and "normal / abnormal" basic state on an electronic map and pushing simple operation and maintenance orders to operation and maintenance personnel, which preliminarily realizes the transformation of manhole cover monitoring from "passive repair" to "active early warning" and improves the operation and maintenance response speed to some extent.

[0003] However, the existing manhole cover monitoring system based on the Internet of Things still has significant technical defects in addressing the three core issues of manhole cover vibration noise, road flatness and manhole cover subsidence, as well as the requirement for accurate identification of device state, making it difficult to meet the requirements of refined operation and maintenance. The existing system lacks pertinence in setting warning thresholds and fails to integrate the differences in manhole cover types and installation scenarios. The system mostly uses a unified fixed threshold without considering the characteristic differences of manhole covers of different materials and different installation environments. SUMMARY

[0004] The purpose of the present application is to provide a real-time monitoring and intelligent early warning system for the state of a manhole cover based on the Internet of Things, which solves the following technical problems:

[0005] The existing manhole cover monitoring system based on the Internet of Things still has significant technical defects in addressing the three core issues of manhole cover vibration noise, road flatness and manhole cover subsidence, as well as the requirement for accurate identification of device state, making it difficult to meet the requirements of refined operation and maintenance.

[0006] The purpose of the present application can be achieved by the following technical solutions:

[0007] The real-time monitoring and intelligent early warning system for the state of a manhole cover based on the Internet of Things comprises:

[0008] The data acquisition module is used for synchronously collecting device running state data, including vibration frequency, vibration noise and inclination angle, by integrating various sensors in the well lid monitoring terminal.

[0009] The early warning device marking module is used for marking the well lid on the local map according to the device running state, and determining the well lid with normal device state.

[0010] The well lid fault pre-diagnosis module is used for pre-diagnosing and marking the well lid with abnormal precursor by analyzing the time sequence change trend of the long-term running state of the well lid and mining the progressive abnormal characteristics before the fault occurs.

[0011] The early warning threshold determination module is used for classifying the well lid with normal device state, and determining the inclination angle threshold, vibration frequency threshold and vibration noise threshold of the well lid vibration noise risk, well lid subsidence risk or road surface flatness risk according to different well lid types.

[0012] The hierarchical early warning signal triggering module is used for determining whether the well lid with abnormal precursor appears the well lid vibration noise risk, well lid subsidence risk or road surface flatness risk, and triggering different level early warning signals according to the determination result.

[0013] The well lid alarm module is used for pushing early warning information and operation and maintenance work orders according to the different level early warning signals, and notifying the staff through the corresponding alarm mode.

[0014] As a further scheme of the application, in the early warning device marking module, the process of marking the well lid on the local map according to the device running state comprises the following steps:

[0015] The early warning device marking module receives the device running state data output by the data acquisition module, and the sensor heartbeat data and real-time uploading state of the collected data.

[0016] If the device running state data meets the normal range of the defined device running state data, and the sensor heartbeat data and collected data are uploaded in real time without interruption, the corresponding well lid is marked as a blue mark on the map.

[0017] If the sensor heartbeat data or collected data appears real-time uploading interruption, the module triggers the manual cover opening inspection process.

[0018] When the manual cover opening operation is performed, if the device synchronously uploads the collected data, the corresponding well lid is marked as an orange mark on the map.

[0019] If the manual cover opening operation is completed, and the device still does not upload the heartbeat data and collected data, it is determined that the device state is abnormal, and the corresponding well lid is marked as a red mark on the map.

[0020] As a further scheme of the present application: in the well lid failure pre-diagnosis module, the process of real-time pre-diagnosis and marking of abnormal well lids with failure precursors is:

[0021] Collecting equipment operation history data of the target well lid monitoring terminal for consecutive n days, defining a failure precursor period by associating contemporaneous failure records, and associating the operation history data with the precursor label and marking positive and negative samples, wherein n is a preset value;

[0022] Calculating the interdiurnal change rate of the equipment operation state data and the set sliding window trend difference, and fusing into a time series progressive feature set;

[0023] Using a time series convolution network as the architecture, training with the time series progressive feature set and the precursor label, and obtaining a single well lid exclusive failure precursor recognition model through back propagation optimization;

[0024] Real-time collection of equipment operation state data output by the well lid monitoring terminal to generate time series progressive features, input into the failure precursor recognition model to obtain a probability value, and if the probability value meets the standard, the abnormal well lid with a failure precursor is marked.

[0025] As a further scheme of the present application: in the early warning threshold determination module, the process of determining the inclination angle threshold, the vibration frequency threshold and the vibration noise threshold of the well lid vibration noise risk, the well lid subsidence risk or the road surface flatness risk is:

[0026] Extracting installation environment attributes and physical specification attributes of well lids with normal equipment state to form a classification original data set;

[0027] Based on the classification original data set, using a hierarchical clustering algorithm combined with domain rules, the well lids are divided into heavy traffic type, regular road type and slow walking area type;

[0028] According to the classification results, the vibration frequency, inclination angle and vibration noise historical monitoring data of the equipment operation state normal period under the corresponding category are screened to construct a classification exclusive database;

[0029] From the classification exclusive database, normal period monitoring data of different types of well lids in the last x months are taken; according to the well lid installation environment attributes and physical specification attributes, exclusive safety redundancy coefficients, finite element model mechanical parameters and flatness analysis windows are set; in order of well lid vibration noise risk, well lid subsidence risk and road surface flatness risk, the unified derivation logic is reused with exclusive parameters to obtain the inclination angle, vibration frequency and vibration noise threshold values of each risk, wherein x is a preset value.

[0030] As a further scheme of the present application: based on the classification original data set, using a hierarchical clustering algorithm combined with domain rules, the well lids are divided into heavy traffic type, regular road type and slow walking area type:

[0031] Docking with the city geographic information system, traffic flow monitoring database and vibration source mapping system, the road type, daily average traffic level and surrounding vibration source intensity value of the well lid are extracted;

[0032] The well lid production archive database is called to extract the bearing level, well lid material and well lid diameter;

[0033] The installation environment attribute and the physical specification attribute are associated, the missing value sample is removed, the one-hot encoding is performed on the non-numeric attribute, the standardization processing is performed on the numeric attribute, and the classification original data set is generated;

[0034] The classification original data set is taken as input, the Ward clustering criterion and the Euclidean distance measurement are adopted, the maximum distance threshold value in the class is set as the clustering termination condition, and the initial classification cluster is obtained;

[0035] Based on the city well lid operation and maintenance specification, the initial cluster meeting the trunk road + heavy load bearing + high daily average traffic, branch + standard bearing + medium daily average traffic and sidewalk + light bearing + low daily average traffic is divided into heavy traffic type, conventional road type and slow walking area type respectively; the remaining samples are re-distributed according to the weighted similarity of the installation environment attribute and the physical specification attribute, and the final classification is determined.

[0036] As a further scheme of the application, the process of screening the vibration frequency, inclination angle and vibration noise historical monitoring data of the equipment running normal period under the corresponding category according to the classification result and constructing the classification exclusive database is:

[0037] The well lid classification result is associated with the historical monitoring data stored by the data acquisition module through the unique identification of the well lid, including the vibration frequency, inclination angle, vibration noise and acquisition time stamp, and the associated data set is generated;

[0038] The associated data set is verified according to the parameter normal range of the early warning threshold determination module, the period in which the three parameters of the same well lid unique identification are continuously collected for 12 cycles and are in the normal range and have no data loss and interruption is marked as the equipment running normal period;

[0039] The associated data set is grouped according to the well lid classification result, the monitoring data in the equipment running normal period is retained in each group according to the acquisition time stamp, and the classification screening data set of different types of well lids is obtained;

[0040] The classification exclusive database containing the specified field is created for the classification screening data set of different types of well lids, the corresponding data is imported, and the timing synchronization mechanism is set to receive new normal data.

[0041] As a further scheme of the application, the process of obtaining the inclination angle, vibration frequency and vibration noise threshold of each risk is:

[0042] Obtain the monitoring data of the heavy traffic type, the regular road type and the slow-moving area type well lid in the normal period of the last x months from the classified exclusive database to generate the target data set of each type;

[0043] According to the installation environment attribute and the physical specification attribute of the well lid of different types, the exclusive safety redundancy coefficient, the finite element model mechanical parameter and the road surface flatness analysis time window are set;

[0044] According to the vibration noise risk, the well lid subsidence risk and the road surface flatness risk in order, the data preprocessing, the finite element simulation and the space-time feature analysis logic are reused, and the inclination angle, the vibration frequency and the vibration noise threshold value of each type of risk are obtained by substituting the exclusive parameters.

[0045] The mapping table of the well lid category-risk type-three threshold values is summarized, the attribute identification is associated, and the threshold value database of the early warning threshold value determination module is stored for calling by the graded early warning signal triggering module.

[0046] As a further scheme of the present application, in the graded early warning signal triggering module, the process of determining whether the well lid with a failure precursor anomaly appears the well lid vibration noise risk, the well lid subsidence risk or the road surface flatness risk is as follows:

[0047] The type of the target well lid is obtained, and the inclination angle threshold value, the vibration frequency threshold value and the vibration noise threshold value of the vibration noise risk of the target well lid in the type are obtained; the inclination angle threshold value, the vibration frequency threshold value and the vibration noise threshold value of the well lid subsidence risk; and the inclination angle threshold value, the vibration frequency threshold value and the vibration noise threshold value of the road surface flatness risk.

[0048] The vibration frequency, the vibration noise and the inclination angle of the target well lid are obtained in real time, and if any one of the vibration frequency, the vibration noise and the inclination angle exceeds the corresponding threshold value, it is determined that the well lid appears the corresponding risk warning.

[0049] As a further scheme of the present application, the specific way of triggering different level warning signals according to the determination result of the intelligent well lid is as follows:

[0050] If one of the vibration frequency, the vibration noise and the inclination angle exceeds the corresponding threshold value, it is determined that the well lid is a primary risk warning.

[0051] If two of the vibration frequency, the vibration noise and the inclination angle exceed the corresponding threshold value, it is determined that the well lid is a middle risk warning.

[0052] If all of the vibration frequency, the vibration noise and the inclination angle exceed the corresponding threshold value, it is determined that the well lid is a serious risk warning.

[0053] The present application has the following beneficial effects:

[0054] The present application can accurately solve the defects of the existing system that the monitoring of the three core problems of manhole cover vibration noise, road flatness and manhole cover subsidence is not accurate, and meets the fine operation and maintenance requirements. The data acquisition module synchronously acquires three types of key data of vibration frequency, vibration noise and inclination angle by integrating various sensors in the manhole cover monitoring terminal, and can comprehensively capture the characteristic signals of the three types of problems - the vibration noise data is directly related to the noise disturbance problem, the inclination angle data reflects the subsidence trend of the manhole cover, and the cooperative data of the vibration frequency and the inclination angle can accurately judge the road flatness condition, avoiding the one-sidedness of the single parameter monitoring of the existing system; at the same time, the early warning threshold determination module classifies the manhole covers with normal equipment state first, and then customizes exclusive thresholds for different types of manhole covers, completely solves the problem that the existing system unified threshold does not match the manhole cover material and installation environment, greatly reduces the false alarm and risk of missing report without noise, reduces invalid operation and maintenance actions, reduces operation and maintenance cost, and ensures that the three core problems can be identified in time and accurately.

[0055] The present application can break through the limitations of the existing system that the equipment state recognition is fuzzy and the early warning response is inefficient, and improve the accuracy and timeliness of operation and maintenance response. The early warning device marking module determines the manhole covers with normal equipment state through map marking, which not only provides accurate basic data support for subsequent classification and threshold determination, but also can be clearly distinguished from abnormal state, avoiding the defects of the existing system that the "normal / abnormal" is divided generally; the hierarchical early warning signal triggering module realizes primary, intermediate and serious three-level early warning according to the number of parameter over-standard items, so that the staff can prioritize high-risk problems according to the early warning level, avoiding resource mismatch; the manhole cover alarm module pushes early warning information and operation and maintenance work order based on different levels of early warning signals, which can ensure that the three types of risks and equipment abnormal conditions are quickly transmitted to the staff, shorten the fault response time - avoid over-response to minor risks, and prevent serious risk processing lag, ultimately reduce safety accidents and public opinion complaints caused by manhole cover problems, and ensure road traffic safety and residents' living environment. BRIEF DESCRIPTION OF DRAWINGS

[0056] The present application will be further described below in conjunction with the drawings.

[0057] Figure 1 It is a structure diagram of the manhole cover state real-time monitoring and intelligent early warning system based on the Internet of Things. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor belong to the scope of protection of the present application.

[0059] Please refer toFigure 1 The application is a real-time monitoring and intelligent early warning system for manhole covers based on the Internet of Things, comprising:

[0060] A data acquisition module is configured to integrate multiple sensors in the manhole cover monitoring terminal and synchronously acquire device operation state data, including vibration frequency, vibration noise and tilt angle.

[0061] An early warning device marking module is configured to mark the location of the manhole cover on a map according to the device operation state and determine the manhole cover with a normal device state.

[0062] A manhole cover fault pre-diagnosis module is configured to analyze the time series trend of the long-term operation state of the manhole cover, mine the progressive abnormal features before the fault occurs, and pre-diagnose and mark the manhole cover with a precursor abnormality in real time.

[0063] An early warning threshold determination module is configured to classify the manhole cover with a normal device state and determine the tilt angle threshold, vibration frequency threshold and vibration noise threshold of the manhole cover vibration noise risk, manhole cover subsidence risk or road flatness risk according to different manhole cover types.

[0064] A hierarchical early warning signal triggering module is configured to determine whether the manhole cover with a precursor abnormality has the manhole cover vibration noise risk, manhole cover subsidence risk or road flatness risk, perform hierarchical early warning on the intelligent manhole cover according to the determination result, trigger different levels of early warning signals accordingly, and monitor the abnormal displacement and water immersion state of the manhole cover.

[0065] A manhole cover alarm module is configured to push early warning information and operation and maintenance work orders and notify the staff through the corresponding alarm mode according to the triggering of different levels of early warning signals.

[0066] In the data acquisition module, multiple sensors are integrated in the manhole cover monitoring terminal to synchronously acquire device operation state data, including vibration frequency, vibration noise and tilt angle.

[0067] In the monitoring of the manhole cover, the three parameters of vibration frequency, vibration noise and tilt angle are collected through the built-in high-precision three-axis acceleration sensor of the intelligent manhole cover safety monitoring terminal. The vibration frequency is indirectly calculated by combining the original acceleration data collected and the signal processing algorithm; the tilt angle is indirectly calculated by using the perceived gravity acceleration component through the algorithm; and the vibration noise is collected by monitoring the vibration frequency: the greater the vibration frequency, the more intense the vibration, and the greater the vibration noise. Specifically, after the vibration frequency is obtained, the frequency spectrum of the vibration and the acceleration amplitude corresponding to each frequency are obtained by the three-axis acceleration sensor, and then the vibration energy is calculated by the frequency and amplitude together. Subsequently, the vibration energy is mapped to the sound pressure level parameter in the air according to the acoustic radiation model, so as to obtain the characteristic parameter of the vibration noise.

[0068] In the early warning device marking module, the manhole cover is marked on the map according to the device running state, and the process of determining the manhole cover with a normal device state is as follows:

[0069] The early warning device marking module receives the device running state data (including vibration frequency, inclination angle and other running parameters) output by the data collection module in real time, as well as the sensor heartbeat data and the real-time uploading state of the collected data.

[0070] If the device running state data meets the normal range of the defined device running state data (the vibration frequency, inclination angle and other parameters are within the threshold), and the heartbeat data and the collected data are uploaded in real time without interruption, the corresponding intelligent manhole cover is marked as blue on the map (corresponding to a normal device state).

[0071] If the heartbeat data or the collected data appears real-time uploading interruption (signal interruption), the module triggers the manual cover opening inspection process:

[0072] When the manual cover opening operation is performed, if the device synchronously uploads the collected data (data reporting when the cover is opened), the corresponding intelligent manhole cover is marked as orange on the map (corresponding to device signal loss).

[0073] If the device still does not upload the heartbeat data and the collected data after the manual cover opening operation is completed, it is determined that the device state is abnormal, and the corresponding intelligent manhole cover is marked as red on the map (corresponding to an abnormal device state).

[0074] In the manhole cover fault pre-diagnosis module, the process of real-time pre-diagnosis and marking of the manhole cover with a fault precursor is as follows:

[0075] Collect the device running state historical data of the target manhole cover monitoring terminal for 60 consecutive days, and the data dimensions include daily average vibration frequency, daily average vibration noise and daily average inclination angle, to ensure uniform data sampling time interval (collecting once a day at a fixed time period);

[0076] Correlate the historical operation and maintenance records of the manhole cover during the same period, filter out the confirmed fault records (such as maintenance records and vibration noise exceeding standard treatment records), and define the 14 days before the fault as the fault precursor time;

[0077] Correlate the historical running state data with the fault precursor period label to construct a single manhole cover time series running data set containing running parameters-time stamp-fault precursor label, wherein the data in the fault precursor period is labeled as positive samples, and the normal running data in the non-fault precursor period is labeled as negative samples.

[0078] Based on the single well cover timing operation data set, the inter-daily change rate of three operation parameters is calculated respectively, and the calculation formula is: (the parameter value of the day-the parameter value of the previous day) / the parameter value of the previous day*100%, which is used to reflect the relative change amplitude of each daily parameter;

[0079] The trend difference of the three operation parameters is calculated respectively within a 7-day sliding window, which is specifically: taking the current day as the end of the window, calculating the difference between the average value of the previous 7 days and the average value of the previous 7 days, which is used to capture the gradual trend change of the parameters within a two-week cycle;

[0080] The inter-daily change rate of the vibration frequency, the inter-daily change rate of the vibration noise, the inter-daily change rate of the tilt angle, the 7-day sliding window trend difference of the vibration frequency, the 7-day sliding window trend difference of the vibration noise, and the 7-day sliding window trend difference of the tilt angle are fused to form a time series progressive feature set.

[0081] A time series convolution network is used as the basic model architecture, which has the ability to capture long-term time series dependencies and can effectively identify the progressive law of parameter changes;

[0082] The generated time series progressive feature set is used as the input layer data of the model, and the fault precursor label (positive sample / negative sample) in the single well cover timing operation data set is used as the output layer label of the model. 80% of the data set is divided into a training set, and 20% of the data set is divided into a validation set;

[0083] The model parameters are iteratively optimized through the back propagation algorithm to minimize the misjudgment rate of the model on the fault precursor period until the precursor identification accuracy of the model on the validation set is stable, and the single well cover exclusive fault precursor identification model is obtained.

[0084] Real-time acquisition of device operation state data output by the well cover monitoring terminal (acquisition once every 2 hours), and synchronous generation of the inter-daily change rate and 7-day sliding window trend difference corresponding to the real-time data according to the same calculation method, to form real-time time series progressive features;

[0085] The real-time time series progressive features are input into the trained single well cover exclusive fault precursor identification model, and the model outputs the probability value of the current time belonging to the fault precursor period;

[0086] If the probability value output by the model reaches the preset judgment standard, the well cover is marked as a fault precursor abnormal well cover, and the marking time and the corresponding real-time operation parameters are recorded synchronously.

[0087] In the early warning threshold determination module, the well covers with normal equipment states are classified, and the tilt angle threshold, vibration frequency threshold and vibration noise threshold of the well cover vibration noise risk, well cover subsidence risk or road flatness risk are determined according to different well cover types.

[0088] Extract the installation environment attribute and physical specification attribute of the manhole cover with normal equipment state to form a classification raw data set;

[0089] Based on the classification raw data set, adopt hierarchical clustering algorithm combined with domain rules to divide the manhole cover into heavy traffic type, regular road type and slow walking area type;

[0090] According to the classification result, filter the vibration frequency, inclination angle and vibration noise historical monitoring data of the equipment during the normal period, and construct a classification exclusive database;

[0091] Take the monitoring data of the three types of manhole covers in the normal period for x months from the classification exclusive database; according to the installation environment attribute and physical specification attribute, set exclusive safety redundancy coefficient, finite element model mechanical parameter and flatness analysis window; according to the order of manhole cover vibration noise risk, manhole cover subsidence risk and road surface flatness risk, reuse the unified derivation logic to substitute the exclusive parameters to obtain the inclination angle, vibration frequency and vibration noise threshold value of each risk, wherein x is a preset value.

[0092] In the process of extracting the installation environment attribute and physical specification attribute of the manhole cover with normal equipment state to form a classification raw data set, and based on the classification raw data set, adopting hierarchical clustering algorithm combined with domain rules to divide the manhole cover into heavy traffic type, regular road type and slow walking area type, the following contents are included:

[0093] Connect the city geographic information system, traffic flow monitoring database and vibration source mapping system to extract the road type, daily vehicle flow level and surrounding vibration source intensity value of the manhole cover, wherein the city geographic information system systematically records the official classification of each road (such as trunk road, branch road and pedestrian road) according to the city road planning and construction archives, so the type of the road where the manhole cover is located can be accurately obtained through the system; the traffic flow monitoring database continuously counts the vehicle traffic volume in a unit time through the monitoring equipment arranged on the road section, and forms the vehicle flow level according to the preset period (such as daily average), which can directly reflect the traffic load frequency and intensity that the manhole cover bears daily; the vibration source mapping system determines the specific position of the vibration source such as rail transit line and construction area through field survey, and then converts the distance into intensity value according to the vibration propagation law, which reflects the degree of external vibration interference faced by the manhole cover; the three types of information together constitute the installation environment attribute of the manhole cover, which provides external condition basis for subsequent classification.

[0094] Call the manhole cover production archive database to extract the load-bearing grade, manhole cover material, and manhole cover diameter. Because the production enterprises will test the performance and record the parameters of each batch of products according to the national standards before the manhole cover leaves the factory, the load-bearing grade determines the maximum load limit that the manhole cover can withstand, the manhole cover material determines its mechanical properties (such as compressive and bending strength), and the manhole cover diameter affects the stress distribution state when the force is applied. These physical specification attributes directly reflect the anti-risk ability of the manhole cover itself and are indispensable internal condition basis for classification.

[0095] Correlate the installation environment attributes and physical specification attributes, remove missing value samples, perform one-hot encoding on non-numeric attributes and standardization processing on numeric attributes to generate a classification raw data set. The correlation operation is achieved through the unique identification of the manhole cover (such as the factory number) to ensure that the external environment information and the physical information of the same manhole cover correspond one-to-one and avoid data misplacement. Removing missing value samples is because the absence of any attribute (such as missing load-bearing grade) will result in incomplete classification basis, which may cause clustering bias. Non-numeric attributes (such as road type and manhole cover material) cannot be directly involved in algorithm calculation, and one-hot encoding can convert them into discrete numerical form while preserving the attribute category difference. The dimensions and numerical ranges of numeric attributes (such as the intensity value of the surrounding vibration source and the diameter of the manhole cover) are different (such as the intensity value is a coefficient and the diameter is a length), and standardization processing can eliminate the dimension influence, making each attribute have a fair weight in the clustering process. The data after these treatments can meet the input requirements of hierarchical clustering algorithm and form a classification raw data set.

[0096] With the classification raw data set as input, Ward's clustering criterion and Euclidean distance measurement are used, and the maximum distance threshold within the class is set as the clustering termination condition to obtain the initial classification cluster. Ward's clustering criterion is selected because it merges clusters by minimizing the within-cluster sum of squares, ensuring that the internal samples of the merged cluster have high attribute similarity, which meets the core requirement of similar attributes in manhole cover classification. Euclidean distance measurement is used to calculate the difference between two manhole cover samples in multiple attribute dimensions, which is a common and effective method for measuring sample similarity and is suitable for standardized numerical data. Setting the maximum distance threshold within the class as the clustering termination condition is because when the maximum attribute distance of the samples within the cluster exceeds the threshold, further merging will result in too large differences within the cluster, losing the classification significance. The initial classification cluster obtained after stopping merging can ensure the attribute consistency of the samples within the cluster and provide a basis for subsequent rule modification.

[0097] Based on the city manhole cover operation and maintenance specification, the initial cluster meeting the main road + heavy load bearing + high daily average traffic volume, branch + standard load bearing + medium daily average traffic volume, sidewalk + light load bearing + low daily average traffic volume is divided into heavy traffic type, conventional road type and slow-moving area type respectively. The remaining samples are re-distributed according to the weighted similarity of 60% installation environment attribute and 40% physical specification attribute to determine the final classification. The city manhole cover operation and maintenance specification is based on the long-term municipal operation and maintenance practice summary, which clearly defines the manhole cover bearing grade and corresponding traffic volume range that need to be matched for different road types (such as heavy load manhole cover for high traffic volume on main road). Therefore, the initial cluster meeting the above combination conditions completely fits the three typical application scenarios in actual operation and maintenance, and can be directly divided into the corresponding types. The remaining samples are special cases that do not meet the typical combination in the initial clustering (such as light load manhole cover on main road). The principle of distributing the remaining samples according to the weighted similarity is that the installation environment attribute (road type, traffic volume, etc.) determines the external load intensity that the manhole cover bears, and has a more critical impact on classification. Therefore, it is given a weight of 60%. The physical specification attribute is the basis for the manhole cover to adapt to the external load, and is given a weight of 40%. By calculating the weighted similarity of the remaining samples and the three typical types, the remaining samples are distributed to the type with the most matched attributes, finally ensuring that all manhole cover samples can be classified into the category that meets the actual application scenario, and the classification is completed.

[0098] In the process of screening the vibration frequency, inclination angle and vibration noise historical monitoring data of the equipment in a normal state in the corresponding category according to the classification result, a classification exclusive database is constructed, which specifically includes the following contents:

[0099] By associating the classification result with the historical monitoring data (including vibration frequency, inclination angle, vibration noise and collection timestamp) stored by the data collection module, an associated data set is generated. The unique identification of the manhole cover is an exclusive code set for each manhole cover when it is manufactured, which can ensure that the classification result (heavy traffic type, conventional road type and slow-moving area type) accurately corresponds to the historical monitoring data of the manhole cover, avoiding confusion of data of different manhole covers. The data collection module will continuously collect the vibration frequency, inclination angle and vibration noise of the manhole cover during operation at a fixed period, and record the collection timestamp. These data can completely reflect the operation state of the manhole cover at different time periods. By associating the classification result with these monitoring data through the unique identification, each monitoring data can be clearly attributed to the corresponding category of manhole cover, providing clear category labels for subsequent screening, thereby generating an associated data set.

[0100] The parameter normal range of the early warning threshold determination module is determined based on a large amount of manhole cover monitoring data in a normal state, can accurately define the reasonable fluctuation interval of the three parameters in the normal operation of the manhole cover, and can be used as a standard for judging whether the manhole cover is normal. The continuous 12 collection cycles are selected as the verification duration because the single collection data may be affected by accidental factors (such as a temporarily passing heavy vehicle) and appear a short-term fluctuation, and the stable data of the continuous multiple cycles can better reflect the real normal operation state of the manhole cover. The period of data loss and interruption is excluded because such data cannot fully reflect the operation of the manhole cover and may lead to a false state. Only the period in which the three parameters meet the normal range and the data are continuous and complete can be confirmed as the normal operation period of the device.

[0101] The associated data set is grouped according to the classification result, the monitoring data with the collection timestamp in the normal operation period of the device is retained in each group, and three types of classification filtering data sets are obtained. The associated data set is grouped according to the classification result, that is, the manhole cover data belonging to the heavy traffic type, the regular road type and the slow area type are respectively divided into three data groups, so that each group of data only corresponds to a single type of manhole cover, avoiding the mixing of different types of data affecting subsequent processing. The collection timestamp can accurately locate the generation period of the monitoring data. By comparing the timestamp with the marked normal operation period of the device, the data in the normal period is retained, and the monitoring data in the abnormal period (such as parameter out-of-range and data interruption) is excluded. Finally, three types of classification filtering data sets containing only the normal operation state monitoring data of each type of manhole cover are obtained, providing a pure data basis for constructing a dedicated database.

[0102] A classification dedicated database containing specified information is created for the three types of classification filtering data sets, the corresponding data is imported, and a timing synchronization mechanism is set to receive new normal data. An independent classification dedicated database is created for each type of classification filtering data set. The normal operation parameter characteristics of different types of manhole covers are different, and independent storage can avoid data cross interference and facilitate subsequent operations such as threshold derivation for a single type of manhole cover. The three types of classification filtering data sets are respectively imported into the corresponding database, which can ensure that the data storage and the category are one-to-one, and realize the ordered management of data. The timing synchronization mechanism is set, the data collection module will continuously generate new manhole cover monitoring data, the timing synchronization can automatically filter the part of the new data that meets the normal operation period of the device, and import the classification dedicated database, so that the data in the database is always the latest normal state data, providing timeliness support for subsequent threshold updating and other work.

[0103] Taking three types of well covers from the classification-specific database for x months of normal period monitoring data; according to its installation environment attributes and physical specification attributes, setting specific safety redundancy coefficient, finite element model mechanical parameters and flatness analysis window; according to the order of well cover vibration noise risk, well cover subsidence risk and road surface flatness risk, reuse unified derivation logic into specific parameters, get the inclination angle, vibration frequency and vibration noise threshold of each type of risk, including the following contents:

[0104] Taking heavy traffic type, conventional road type and slow area type well covers from the classification-specific database for x months of normal period monitoring data to generate various types of target data sets, wherein the classification-specific database is a database that stores corresponding normal period monitoring data independently according to well cover categories, which can directly locate and extract specific data of each type of well cover, avoiding mixing of different category data; selecting data for nearly x months is because the monitoring data in recent period can truly reflect the current traffic environment of the well cover (such as recent traffic flow changes, surrounding vibration source dynamics) and the current mechanical performance state of the well cover itself, avoiding the data from being out of touch with the actual working condition due to too long time, ensuring that the data used for threshold derivation has timeliness and authenticity; generating various types of target data sets is to collect each type of well cover data separately, providing accurate and pure data basis for subsequent differentiated threshold derivation for different types of well covers, ensuring that each type of threshold derivation is based only on the normal operation data of this type of well cover, avoiding interference from cross-category data.

[0105] According to the installation environment attributes and physical specification attributes of the three types of manhole covers, the exclusive safety redundancy coefficient, the finite element model mechanical parameters and the road surface flatness analysis time window are set. The installation environment attribute of the heavy load traffic type manhole cover is "main road + high daily vehicle flow", the physical specification attribute is "heavy load bearing + high strength material (such as ductile cast iron)", and it bears high frequency heavy load vehicle impact and strong vibration interference every day. Therefore, a higher safety redundancy coefficient is set to reserve enough risk buffer space, the finite element model mechanical parameters need to match the compression and bending resistance of high strength material to simulate the real stress state, and the road surface flatness analysis time window is set to a longer time to cover the parameter fluctuation of the dense traffic period. The installation environment attribute of the conventional road type manhole cover is "branch + daily vehicle flow", the physical specification attribute is "standard bearing + conventional material (such as reinforced concrete)", and the load and vibration it bears are between heavy load and slow moving area. Therefore, the safety redundancy coefficient, the finite element model mechanical parameters and the flatness analysis time window are set to a medium level. The installation environment attribute of the slow moving area type manhole cover is "sidewalk + low daily vehicle flow", the physical specification attribute is "light load bearing + light weight material (such as composite material)", and the load and vibration it bears are the smallest. Therefore, the safety redundancy coefficient is the lowest, the finite element model mechanical parameters match the performance of light weight material, and the flatness analysis time window is the shortest to adapt to the low frequency parameter change of pedestrian traffic. The exclusive parameters are set because there are essential differences in the external load intensity and the self anti-risk ability of different types of manhole covers. The exclusive parameters can ensure that the subsequent threshold derivation fits the actual working conditions of each type of manhole cover, and avoid that the unified parameters lead to too high or too low threshold.

[0106] The three types of thresholds are mapped according to the order of vibration noise risk, manhole cover subsidence risk, and road flatness risk, and the logic of data preprocessing, finite element simulation, and spatiotemporal feature analysis is reused to obtain the inclination angle, vibration frequency, and vibration noise threshold of each type of risk by substituting the exclusive parameters. The unified logic is used to ensure that the threshold derivation methods of the three types of manhole covers in the same risk dimension are consistent, to ensure the comparability and reliability of the thresholds of different types of manhole covers, and to avoid threshold deviation caused by method differences. The data preprocessing logic is used for vibration noise risk because there may be short-term random fluctuations in normal period monitoring data (such as transient vibration caused by temporary passing small vehicles). By data preprocessing, such fluctuations can be eliminated, highlighting the true correlation between vibration noise and vibration frequency and inclination angle. After substituting the exclusive safety redundancy coefficient, a reasonable noise risk threshold can be determined based on the normal data range. The finite element simulation logic is used for manhole cover subsidence risk because manhole cover subsidence is essentially a structural stress deformation problem. Finite element simulation can simulate stress distribution under different inclination angles by constructing a manhole cover mechanical model. When the stress reaches the material limit, the corresponding parameters are the risk threshold. After substituting the exclusive finite element model mechanical parameters, the subsidence risk threshold that conforms to the actual bearing capacity can be derived by accurately matching the material properties of each type of manhole cover. The spatiotemporal feature analysis logic is used for road flatness risk because road flatness needs to be reflected through parameter fluctuations over a period of time (such as changes in angle and frequency caused by continuous pedestrian traffic or small vehicles passing through). By dividing the data according to the exclusive road flatness analysis time window, the parameter fluctuation characteristics within the window can be calculated. After correlating with the municipal flatness standard, the risk threshold can be determined. After substituting the exclusive time window, the threshold that fits the actual flatness requirements can be derived by adapting to the use scenarios of each type of manhole cover (such as heavy traffic periods or sparse pedestrian periods).

[0107] The mapping table of manhole cover category-risk type-three thresholds is formed by summarizing the inclination angle, vibration frequency, and vibration noise threshold of each type of manhole cover corresponding to the three types of risks according to the "category-risk" dimension, and is stored as a threshold database of the early warning threshold determination module after associating with attribute identifiers for the subsequent tracing of the corresponding manhole cover installation environment attributes and physical specification attributes. When the attributes of the manhole cover change (such as material replacement), the corresponding threshold can be quickly located and updated. The threshold database is stored in the early warning threshold determination module because the hierarchical early warning signal triggering module needs to retrieve the threshold of the corresponding category and corresponding risk for comparison when determining the risk of the manhole cover. Database storage can ensure the timeliness and accuracy of threshold retrieval and provide stable threshold support for subsequent hierarchical early warning.

[0108] In the hierarchical early warning signal triggering module, the process of determining whether the manhole cover appears a manhole cover vibration noise risk warning, a manhole cover subsidence risk warning, or a road flatness risk warning is as follows:

[0109] The type of the target manhole cover is obtained, and the inclination angle threshold, vibration frequency threshold and vibration noise threshold of the target manhole cover in the vibration noise risk warning, manhole cover subsidence risk warning or road flatness risk warning of the type are obtained. Specifically, the unique identification (such as the factory number or installation number) of the target manhole cover is first associated with the manhole cover category record stored in the previous classification stage (which has clearly defined each manhole cover belonging to heavy traffic type, regular road type or slow area type), because after the classification of the manhole cover, the unique identification of each manhole cover is permanently bound with the corresponding category information, and through the unique identification, the specific type of the target manhole cover can be accurately located and extracted, avoiding classification matching errors. And obtaining the three types of risk thresholds of the corresponding type is because the threshold database of the early warning threshold determination module has stored data according to the fixed correspondence of "manhole cover category-risk type-threshold" - that is, each type of manhole cover has exclusive vibration noise risk, manhole cover subsidence risk and road flatness risk thresholds, and after the type of the target manhole cover is determined, the inclination angle threshold, vibration frequency threshold and vibration noise threshold corresponding to the three types of risks under the type can be obtained from the database, because the threshold derivation process in the early stage has formulated exclusive thresholds that fit the actual working conditions of manhole covers of different categories according to their installation environment and physical characteristics, and the type and the threshold have a unique matching relationship, so the accurate corresponding threshold can be directly obtained through the type.

[0110] Real-time acquisition of the vibration frequency, vibration noise and inclination angle of the target manhole cover, and if any of the vibration frequency, vibration noise and inclination angle exceeds the corresponding threshold value, it is determined that the manhole cover has the corresponding risk warning, wherein the real-time acquisition of the parameters is realized by the special sensors installed on the target manhole cover monitoring terminal - the vibration sensor continuously captures the vibration frequency generated by the manhole cover under external force (such as vehicle rolling, pedestrian stepping), the noise sensor real-time acquires the vibration noise generated by the friction or vibration of the manhole cover and well seat, and the inclination sensor monitors the inclination angle of the manhole cover and the horizontal road surface. These sensors will transmit the collected parameters to the hierarchical warning signal triggering module at a preset period, because the sensors can directly sense the physical state change of the manhole cover and are the core components for acquiring real-time operation data, which can ensure the timeliness and authenticity of parameter acquisition; and the logic of determining the risk warning is to compare the real-time parameters with the corresponding threshold value one by one: if the vibration frequency exceeds the vibration frequency threshold value corresponding to the vibration noise risk or the manhole cover subsidence risk or the road surface flatness risk, it means that the current vibration intensity of the manhole cover has exceeded the safety range under this risk; if the vibration noise exceeds the vibration noise threshold value corresponding to the vibration noise risk, it means that the noise generated by the manhole cover has reached the interference degree that needs to be warned; if the inclination angle exceeds the inclination angle threshold value corresponding to the manhole cover subsidence risk or the road surface flatness risk, it means that the manhole cover may have a subsidence trend or the road surface flatness has not met the standard. Since each parameter directly corresponds to the core characteristics of a certain type of risk, any parameter exceeding the threshold value means that the risk condition of the parameter has been met, and if it is not timely warned, it may lead to the expansion of the risk (such as the inclination angle exceeding the threshold value may cause the manhole cover to fall off), therefore, as long as any parameter exceeds the corresponding threshold value, it is determined that the manhole cover has the warning of the risk type to which the parameter belongs, ensuring that the risk can be identified in time and there is no omission.

[0111] The manhole cover abnormal displacement monitoring realizes real-time acquisition of displacement and attitude signals by integrating a three-axis acceleration sensor in the manhole cover monitoring terminal; the edge node compares the real-time signals with the device baseline model, identifies abnormal displacement based on time sequence change and triggers hierarchical warning. The water immersion monitoring function uses conductivity sensors and relative humidity sensors to detect the liquid level and humidity mutation in the manhole cover cavity; the edge node realizes immediate marking and operation and maintenance work order generation for water immersion events. Abnormal displacement and water immersion events realize linkage disposal, synchronously update map markers and push operation and maintenance notifications.

[0112] In the hierarchical warning signal triggering module, the intelligent manhole cover is given hierarchical warning according to the determination result, and the specific way of triggering different level warning signals is as follows:

[0113] If one of the vibration frequency, vibration noise and inclination angle exceeds the corresponding threshold value, it is determined that the manhole cover is a primary risk warning;

[0114] If two of the vibration frequency, vibration noise and inclination angle exceed the corresponding threshold value, it is determined that the manhole cover is a middle risk warning;

[0115] If the vibration frequency, the vibration noise and the inclination angle all exceed the corresponding threshold values, it is determined that the well lid is in a serious risk warning.

[0116] The core logic of the whole grading warning process is that the number of parameter items exceeding the threshold value is positively correlated with the risk severity. Because the vibration frequency, the vibration noise and the inclination angle of the well lid correspond to different key dimensions of its running state, single dimension anomaly only represents a local minor problem, and multi-dimension anomaly reflects the overall state deterioration. By counting the number of parameters exceeding the threshold value to divide the warning level, the warning level can be accurately matched with the actual risk severity, which can not only avoid resource waste caused by excessive response to minor risks, but also prevent safety accidents caused by delayed response to serious risks.

[0117] The above describes one embodiment of the present application in detail, but the content described is only a preferred embodiment of the present application and cannot be considered as limiting the scope of the present application. Any equivalent changes and improvements made within the scope of the present application should still belong to the patent coverage of the present application.

Claims

1. The well lid state real-time monitoring and intelligent early warning system based on Internet of Things, characterized in that, The method comprises the following steps: A data acquisition module is used to integrate multiple sensors in the manhole cover monitoring terminal, and to synchronously collect device operation state data, including vibration frequency, vibration noise, and inclination angle; An early warning device marking module is used to mark the location of the manhole cover on a map according to the device operation state, and to determine the manhole cover with a normal device state; A manhole cover fault pre-diagnosis module is used to analyze the time series change trend of the long-term operation state of the manhole cover, to mine the progressive abnormal features before the fault occurs, to perform real-time pre-diagnosis, and to mark the manhole cover with a precursor abnormality; An early warning threshold determination module is used to classify the manhole cover with a normal device state, and to determine the inclination angle threshold, vibration frequency threshold, and vibration noise threshold of the manhole cover vibration noise risk, manhole cover subsidence risk, and road surface flatness risk according to different manhole cover types; the process comprises: Extracting the installation environment attributes and physical specification attributes of the manhole cover with a normal device state to form a classification original data set; Based on the classification original data set, the manhole cover is divided into heavy traffic type, regular road type, and slow area type by using a hierarchical clustering algorithm combined with domain rules; A graded early warning signal triggering module is used to determine whether the manhole cover with a precursor abnormality appears the manhole cover vibration noise risk, manhole cover subsidence risk, or road surface flatness risk; according to the determination result, the intelligent manhole cover is graded for early warning, and different levels of early warning signals are triggered accordingly; meanwhile, the manhole cover abnormal displacement and water immersion conditions are monitored; A manhole cover alarm module is used to push early warning information and operation and maintenance work orders according to the different levels of early warning signals triggered, and to notify the staff through the corresponding alarm mode.

2. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status as described in claim 1, characterized in that, In the early warning device marking module, the process of marking the location of the manhole cover on the map according to the device operation state comprises: The early warning device marking module receives the device operation state data output by the data acquisition module, as well as the sensor heartbeat data and real-time upload state of the collected data in real time; If the device operation state data is within the normal range of the defined device operation state data, and the sensor heartbeat data and collected data are uploaded in real time without interruption, the corresponding manhole cover is marked as a blue mark on the map; If the sensor heartbeat data or collected data appears real-time upload interruption, the module triggers the manual cover opening inspection process: When the manual cover opening operation is performed, if the device synchronously uploads the collected data, the corresponding manhole cover is marked as an orange mark on the map; If the manual cover opening operation is completed, and the device still does not upload the heartbeat data and collected data, it is determined that the device state is abnormal, and the corresponding manhole cover is marked as a red mark on the map.

3. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status according to claim 1, characterized in that, In the manhole cover fault pre-diagnosis module, the process of real-time pre-diagnosis and marking the manhole cover with a precursor abnormality comprises: Collecting the device operation history data of the target manhole cover monitoring terminal for n consecutive days, defining the fault precursor period by associating the same period fault record, associating and labeling the positive and negative samples of the operation history data and the precursor label, wherein n is a preset value; Calculating the inter-daily change rate of the device operation state data and the set sliding window trend difference, and fusing them into a time series progressive feature set; Taking the time series convolution network as the architecture, training the time series progressive feature set and the precursor label, and obtaining the single manhole cover exclusive fault precursor recognition model through back propagation optimization; Real-time acquisition of device running state data output by the manhole cover monitoring terminal generates time sequence progressive features, inputs a failure precursor identification model to obtain a probability value, and marks a manhole cover as a failure precursor anomaly when the probability value meets a standard.

4. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status according to claim 1, characterized in that, In the early warning threshold determination module, the process of determining the inclination angle threshold, the vibration frequency threshold, and the vibration noise threshold of the manhole cover vibration noise risk, the manhole cover subsidence risk, and the road flatness risk further includes: According to the classification result, the vibration frequency, the inclination angle, and the vibration noise historical monitoring data of the equipment running state normal period under the corresponding category are screened to construct a classification exclusive database. From the classification exclusive database, the normal period monitoring data of different types of manhole covers in the last x months is obtained; according to the manhole cover installation environment attributes and physical specification attributes, an exclusive safety redundancy coefficient, a finite element model mechanical parameter, and a flatness analysis window are set; according to the order of the manhole cover vibration noise risk, the manhole cover subsidence risk, and the road flatness risk, the unified derivation logic is reused and the exclusive parameters are substituted to obtain the inclination angle, the vibration frequency, and the vibration noise threshold of each type of risk, wherein x is a preset value.

5. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status according to claim 4, characterized in that, The process of dividing the manhole cover into heavy traffic type, regular road type, and slow area type based on the classification original data set, using hierarchical clustering algorithm combined with field rules, is as follows: Interface with the city geographic information system, traffic flow monitoring database, and vibration source mapping system to extract the road type, daily average traffic level, and surrounding vibration source intensity value of the manhole cover; Call the manhole cover production archive database to extract the bearing grade, manhole cover material, and manhole cover diameter; Associate the installation environment attributes and physical specification attributes, remove missing value samples, perform one-hot encoding on non-numeric attributes and standardization processing on numeric attributes to generate a classification original data set; Using the classification original data set as input, using Ward clustering criteria and Euclidean distance measurement, setting the clustering termination condition as the maximum distance threshold within the class, obtaining the initial classification cluster; Based on the city manhole cover operation and maintenance specification, the initial cluster that meets the trunk road + heavy load bearing + high daily average traffic, branch + standard bearing + medium daily average traffic, and sidewalk + light bearing + low daily average traffic is divided into heavy traffic type, regular road type, and slow area type respectively; for the remaining samples, the weighted similarity of the installation environment attributes and physical specification attributes is redistributed to determine the final classification.

6. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status according to claim 4, characterized in that, The process of filtering the vibration frequency, inclination angle, and vibration noise historical monitoring data of the equipment running state normal period under the corresponding category according to the classification result to construct a classification exclusive database is as follows: Associate the manhole cover classification result and the historical monitoring data stored by the data acquisition module through the manhole cover unique identifier, including vibration frequency, inclination angle, vibration noise, and collection timestamp, to generate an associated data set; According to the parameter normal range of the early warning threshold determination module, verify the associated data set, filter the period when the three parameters of the same manhole cover unique identifier are within the normal range for 12 consecutive collection periods without data loss or interruption, and mark it as a device running normal period; Group the associated data set according to the manhole cover classification result, and retain the monitoring data with collection timestamps in the device running normal period in each group to obtain different types of manhole cover classification filtered data sets; Create a classification-specific database containing specified fields for the classified and screened dataset of different types of manhole covers, import the corresponding data, and set a timing synchronization mechanism to receive newly added normal data.

7. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status according to claim 4, characterized in that, The process of obtaining the inclination angle, vibration frequency, and vibration noise threshold of each type of risk is as follows: Obtain the monitoring data of the manhole covers of the overload traffic type, the regular road type, and the slow-moving area type from the classification-specific database for the past x months, and generate the target dataset for each type; According to the installation environment attributes and physical specification attributes of different types of manhole covers, set the exclusive safety redundancy coefficient, the finite element model mechanical parameters, and the road surface flatness analysis time window; According to the order of vibration noise risk, manhole cover subsidence risk, and road surface flatness risk, reuse the data preprocessing, finite element simulation, and spatiotemporal feature analysis logic, and substitute the exclusive parameters to obtain the inclination angle, vibration frequency, and vibration noise threshold of each type of risk; Summarize and form a mapping table of manhole cover category-risk type-three threshold values, associate the attribute identifiers, and store them in the threshold value database of the early warning threshold value determination module for the hierarchical early warning signal triggering module to call.

8. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status according to claim 1, characterized in that, In the hierarchical early warning signal triggering module, the process of determining whether the manhole cover with a failure precursor anomaly is at risk of manhole cover vibration noise, manhole cover subsidence, or road surface flatness is as follows: Obtain the type of the target manhole cover, and obtain the inclination angle threshold, vibration frequency threshold, and vibration noise threshold of the target manhole cover in the vibration noise risk of that type; The inclination angle threshold, vibration frequency threshold, and vibration noise threshold of the manhole cover subsidence risk; And the inclination angle threshold, vibration frequency threshold, and vibration noise threshold of the road surface flatness risk; Real-time acquisition of the vibration frequency, vibration noise, and inclination angle of the target manhole cover, if any of the vibration frequency, vibration noise, and inclination angle exceeds the corresponding threshold, then determine that the manhole cover is at risk of corresponding early warning.

9. The IoT-based real-time monitoring and intelligent early warning system for manhole cover status according to claim 8, characterized in that, The specific way of hierarchical early warning of the intelligent manhole cover according to the determination result and the corresponding triggering of different levels of early warning signals is as follows: If one of the vibration frequency, vibration noise, and inclination angle exceeds the corresponding threshold, then determine that the manhole cover is at the primary risk early warning; If two of the vibration frequency, vibration noise, and inclination angle exceed the corresponding threshold, then determine that the manhole cover is at the intermediate risk early warning; If all of the vibration frequency, vibration noise, and inclination angle exceed the corresponding threshold, then determine that the manhole cover is at the severe risk early warning.

Images