A concrete temperature monitoring intelligent early warning method

Abstract

The present application relates to a kind of concrete temperature monitoring intelligent early warning method, including establishing the three-level data organization relationship of bid section, equipment and passage, configure monitoring point semantic label, measuring point physical relationship and early warning rule template for each passage;Receive the original data frame uploaded by temperature collection terminal, obtain effective data frame after processing;Generate the effective sampling point sequence of each passage, and match early warning rule, generate measuring point pair;Calculate temperature value risk parameter, measuring point pair temperature difference risk parameter, temperature change rate risk parameter and historical temperature difference risk parameter, and form risk parameter vector;Generate early warning event record;Store temperature time series data and early warning event record, and execute multi-channel push.Using a kind of concrete temperature monitoring intelligent early warning method designed by the present application, can solve the problem that multi-physical measuring point temperature difference and temperature change rate risk are difficult to be stably identified and prone to false alarm and miss under the condition of asynchronous, unequal interval and abnormal temperature data.

Description

Technical Field

[0001] This invention relates to the fields of engineering construction monitoring and Internet of Things data processing technology, specifically to an intelligent early warning method for concrete temperature monitoring. Background Technology

[0002] During the pouring and curing of large-volume concrete, the cement hydration reaction continuously releases heat, causing changes in the internal and surface temperatures of the concrete, as well as the ambient temperature, throughout the construction process. In engineering construction, temperature sensors are typically embedded and data acquisition terminals are configured to periodically collect temperature data from different locations. The results are then used to evaluate the changes in concrete temperature. With the increasing scale of large-scale bridge, water conservancy, and infrastructure projects, a single project often involves multiple construction areas, multiple data acquisition devices, and a large number of temperature monitoring points. Therefore, the temperature data is characterized by large volumes, dispersed sampling times, and varying physical locations and monitoring purposes of the monitoring points.

[0003] Existing temperature monitoring methods typically use single-point temperature thresholds or fixed-time interval temperature differences as the basis for judgment. Although some systems have achieved remote data acquisition and cloud display, the physical relationship between different monitoring points is not fully involved in the early warning calculation. In the presence of network latency, retransmission, uneven sampling time, missing data at measuring points, and short-term fluctuations, judgments based directly on single measurements or data at fixed time points are easily affected by the sampling status. At the same time, the heat conduction state reflected by points inside the concrete, surface points, and measuring points in the same layer are not the same. Using a single-point threshold uniformly makes it difficult to comprehensively evaluate the temperature difference between the inside and the surface, the rate of temperature change, and continuous abnormal conditions, thus affecting the stability of temperature risk assessment and the effectiveness of early warning information.

[0004] Therefore, existing technologies have shortcomings and need to be improved and developed. Summary of the Invention

[0005] The present invention provides an intelligent early warning method for concrete temperature monitoring, which solves the problem in the prior art that it is difficult to reliably identify the risk of temperature difference and temperature change rate at multiple physical measuring points under asynchronous, unequal interval and abnormal temperature data conditions, and that it is easy to generate false alarms and missed alarms.

[0006] This invention provides an intelligent early warning method for concrete temperature monitoring, executed by a cloud server, comprising:

[0007] Establish a three-level data organization relationship for sections, equipment, and channels, and configure semantic tags for monitoring points, physical relationships between measuring points, and early warning rule templates for each channel;

[0008] Receive raw data frames uploaded by the temperature acquisition terminal, and perform device authentication, integrity verification, duplicate data detection and data quality marking on the raw data frames to obtain valid data frames;

[0009] The effective data frames and corresponding historical data are asynchronously time-aligned according to the acquisition time to generate a sequence of effective sampling points for each channel. The monitoring point semantic tags and measurement point physical relationships are used to match early warning rules and generate measurement point pairs.

[0010] Based on the effective sampling point sequence and measurement point pairs, the temperature value risk parameter, the measurement point pair temperature difference risk parameter, the temperature change rate risk parameter, and the historical temperature difference risk parameter are calculated respectively, and the over-limit state corresponding to each risk parameter is determined to form a risk parameter vector.

[0011] The warning status is updated based on the out-of-limit status, consecutive out-of-limit count, duration, out-of-limit percentage within the window, and hysteresis conditions of the risk parameter vector, and a warning event record is generated.

[0012] The system stores temperature time-series data and the records of the warning events, and determines the corresponding warning handling method based on the updated warning status. When the warning status enters the alarm status or escalation status, it performs multi-channel push according to the corresponding warning level.

[0013] Furthermore, the establishment of a three-level data organization relationship among sections, equipment, and channels, and the configuration of semantic tags for monitoring points, physical relationships between measuring points, and early warning rule templates for each channel, includes:

[0014] Establish a correspondence between bidding sections and construction projects, associate at least one piece of equipment under each bidding section, and associate one to eight channels under each piece of equipment;

[0015] Configure semantic tags for monitoring points for each channel, including pouring zone, point type, burial depth, spatial coordinates, associated points, and structural parts;

[0016] The pouring zone is used to identify the pouring section or construction zone to which the measuring point belongs; the point type is used to identify whether the measuring point is an internal point, surface point, environmental point or rebar point; the embedment depth is the distance of the measuring point from the concrete surface; the spatial coordinates are the position of the measuring point in three-dimensional space; the associated point is the measuring point identifier that is physically associated with the current measuring point; and the structural part is used to identify the structural component where the measuring point is located.

[0017] According to the semantic tags of the monitoring points, the corresponding early warning rule templates are matched, wherein the temperature rise rate early warning rule is matched for internal points, the temperature drop rate early warning rule is matched for surface points, the inner and outer surface temperature difference early warning rule is matched for internal and surface points in the same pouring zone, and the same layer temperature difference early warning rule is matched for measuring points in the same pouring zone with a burial depth difference of no more than 0.1 meters.

[0018] For early warning rules that require two measuring points to participate in the calculation, at least one pair of measuring points is generated based on the associated points, pouring zones, point types, burial depths, and the matched early warning rule template. Each pair of measuring points includes a first measuring point identifier, a second measuring point identifier, and a physical relationship type.

[0019] An initial threshold of 1.5 degrees Celsius per hour is configured for the internal point temperature rise rate warning rule, an initial threshold of 3.0 degrees Celsius per hour is configured for the surface point temperature drop rate warning rule, an initial threshold of 15 degrees Celsius is configured for the inner-to-outer temperature difference warning rule, and an initial threshold of 10 degrees Celsius is configured for the same-layer temperature difference warning rule; the initial thresholds are used as the rate alarm thresholds for the corresponding point type warning rules when the burial depth adjustment is not performed.

[0020] Each warning rule includes a rule identifier, its segment, rule name, parameter type, comparison relationship, threshold, rule type, and activation status. The comparison relationship includes greater than, less than, not less than, not greater than, and range.

[0021] Furthermore, the process of receiving the raw data frame uploaded by the temperature acquisition terminal, and performing device authentication, integrity verification, duplicate data detection, and data quality marking on the raw data frame to obtain a valid data frame includes:

[0022] The system receives raw data frames containing device identifier, channel identifier, channel engineering name, acquisition time, temperature value, signal quality, data status, frame sequence number, and checksum, and generates a reception time upon receiving the raw data frames. The device identifier is a unique identifier for the temperature acquisition terminal, used for device authentication and segment attribution lookup. The channel identifier is the physical measurement point number of the temperature acquisition terminal. The channel engineering name field records the engineering display name of the channel in the construction monitoring scenario; this engineering display name is independent of the point type used to represent internal points, surface points, environmental points, or rebar points. The acquisition time is a timestamp generated when the temperature acquisition terminal acquires the temperature value. The reception time is a timestamp generated when the cloud server receives the raw data frames. The signal quality is the communication signal strength or signal-to-noise ratio. The data status distinguishes between normal acquisition data, retransmission data, and the first frame data after the temperature acquisition terminal restarts. The frame sequence number is an incrementing sequence number generated by the temperature acquisition terminal. The checksum is an integrity check value generated based on the frame content of the raw data frames.

[0023] The legitimacy of the corresponding temperature acquisition terminal is verified based on the device identifier, and the integrity of the original data frame is verified based on the check code. When the temperature acquisition terminal fails the legitimacy verification or the original data frame fails the integrity verification, the data quality field of the corresponding data is marked as invalid, and the original data frame is no longer used for subsequent timing alignment and early warning calculations.

[0024] The combination of the device identifier and the frame sequence number is used as a unique message identifier, and the unique message identifier is queried in the recent message identifier cache with a preset cache duration.

[0025] When the unique message identifier is found, the original data frame is marked as duplicated, and its use in subsequent early warning calculations is stopped. When the unique message identifier is not found, the unique message identifier is written to the most recent message identifier cache, the data quality field of the original data frame is marked as valid, and the data frame that has completed the unified format conversion is written to the message queue. The data quality field is used to indicate the data quality judgment result of the data frame, and the valid status indicates that the corresponding data frame has passed device authentication, integrity verification, and duplicate data detection.

[0026] Furthermore, the asynchronous timing alignment of the valid data frames with the corresponding historical data according to the acquisition time to generate a sequence of valid sampling points for each channel includes:

[0027] Construct a time window with the acquisition time of the current valid data frame as the right endpoint. In the formula, The acquisition time of the current valid data frame. The length of the sliding window is measured in minutes.

[0028] Obtain the set of sampling points arranged in ascending order of acquisition time within the time window: In the formula, The number of sampling points within the time window. For the first The sampling time for each sampling point For the first Temperature values ​​at each sampling point The sampling point number, and ;

[0029] Calculate the time interval between adjacent sampling points: In the formula, For the first The sampling point and the first The time interval between sampling points For the first The sampling time for each sampling point For the first The collection time for each sampling point;

[0030] When the time interval between any two adjacent sampling points is less than the preset minimum time interval, delete the sampling point whose temperature value is closer to the median temperature value in the current time window, and retain the sampling point whose temperature value deviates significantly from the median. Rearrange the remaining sampling points in ascending order of collection time and recalculate the time interval between adjacent sampling points. Repeat the filtering process until the time interval between any two adjacent valid sampling points is not less than the preset minimum time interval, and obtain a valid sampling point sequence.

[0031] When the time difference between the acquisition time of the first valid sampling point and the left endpoint of the time window exceeds twice the minimum time interval, the current time window is determined to have insufficient data coverage, and the calculation of the temperature rise rate or temperature fall rate is stopped using the current time window.

[0032] For each channel participating in the temperature difference calculation of the measuring point pair, select an effective sampling point within the corresponding time window whose acquisition time is no later than the right end of the window and whose time difference with the right end of the window is the smallest. Combine the temperature values ​​of the effective sampling points corresponding to different channels with the temperature value pairs corresponding to the measuring point pair.

[0033] The effective sampling point sequence is used for temperature change rate calculation only when there are at least two effective sampling points and the first and last effective sampling points have different acquisition times.

[0034] Further, the calculation of the rate of temperature rise or fall includes:

[0035] The temperature rise rate or temperature fall rate is calculated using the endpoint difference method or the least squares fitting slope method. The temperature change rate calculation is performed only on valid sampling point sequences with at least two valid sampling points and different sampling times. The sampling time of each valid sampling point is converted into a time value in minutes according to a unified time base before being included in the calculation.

[0036] When using the endpoint difference method, the following calculation is performed based on the first and last valid sampling points in the valid sampling point sequence: In the formula, The rate of temperature rise or fall is expressed in degrees Celsius per hour. The temperature value of the first valid sampling point. This is the temperature value of the last valid sampling point. The time of collection for the first valid sampling point. The time of the last valid sampling point. The time difference is in minutes and is not zero. The value 60 is used to convert the temperature change rate in minutes to the temperature change rate in hours.

[0037] When using the least squares fitting slope method, a univariate linear regression is performed on the effective sampling point sequence: In the formula, To fit the temperature value, The data collection time is measured in minutes. The slope of the fitted line, The intercept of the fitted line;

[0038] The slope of the fitted straight line is calculated according to the following formula: In the formula, The slope of the fitted line, The number of effective sampling points, The valid sampling point number, For the first The collection time for each valid sampling point For the first Temperature values ​​of each valid sampling point;

[0039] Calculate the rate of temperature rise or rate of temperature fall based on the slope of the fitted straight line: In the formula, The rate of temperature rise or fall is expressed in degrees Celsius per hour. The slope of the fitted straight line is expressed in degrees Celsius per minute. The value 60 is used to convert the slope of the fitted straight line in minutes to the rate of temperature rise or fall in hours.

[0040] in, This indicates a rise in temperature. This indicates a drop in temperature.

[0041] Furthermore, based on the effective sampling point sequence and measurement point pairs, the risk parameters for temperature values, temperature differences between measurement point pairs, rate of temperature change, and historical temperature differences are calculated respectively, and the out-of-limit states corresponding to each risk parameter are determined to form a risk parameter vector, including:

[0042] The current temperature value of each channel is compared with the corresponding upper or lower temperature threshold to obtain the temperature risk parameter and the corresponding over-limit status.

[0043] Based on the time-aligned temperature values ​​corresponding to the first and second measuring points of the measuring point pair, the absolute value of the temperature difference between the first and second measuring points is calculated, and the corresponding maximum allowable temperature difference is determined according to the physical relationship type of the measuring point pair. The absolute value of the temperature difference is compared with the maximum allowable temperature difference to obtain the temperature difference risk parameter of the measuring point pair and the corresponding over-limit state.

[0044] Calculate the absolute value of the temperature difference between the current acquisition time and the historical acquisition time corresponding to the preset historical time interval for the same channel. Compare the absolute value of the temperature difference with the maximum allowable historical temperature difference to obtain the historical temperature difference risk parameter and the corresponding over-limit status. When there is no valid temperature data for the historical acquisition time corresponding to the preset historical time interval, no historical temperature difference calculation value is generated, and the corresponding historical temperature difference risk parameter is marked as insufficient data. The insufficient data status is not treated as an over-limit status.

[0045] The temperature rise rate is compared with the temperature rise rate alarm threshold, and the absolute value of the temperature fall rate is compared with the temperature fall rate alarm threshold to obtain the temperature change rate risk parameter and the corresponding over-limit status.

[0046] When adjusting the rate threshold based on the burial depth, the adjusted rate threshold is calculated using the following formula: In the formula, The rate threshold is adjusted based on the burial depth. The base rate alarm threshold is determined based on the current measurement point type and corresponding early warning rules. The depth of the measuring point is measured in meters.

[0047] When the risk parameter of any channel is out of limit, the risk parameter corresponding to the associated point of the channel is obtained; when the associated point shows a same-direction anomaly, the risk of the current channel is marked as a verified risk; when the associated point does not show a same-direction anomaly, the risk of the current channel is marked as an isolated risk and the push priority of the corresponding warning event is reduced; wherein, the same-direction anomaly means that the current channel and the associated point are both out of limit under the same risk parameter type, and for risk parameters with temperature change direction, the temperature change direction corresponding to the current channel and the associated point is the same;

[0048] The calculated values ​​of each risk parameter, the out-of-limit status, and the risk label are combined to form the risk parameter vector.

[0049] Furthermore, the step of updating the warning status and generating a warning event record based on the risk parameter vector's out-of-limit status, consecutive out-of-limit count, duration, out-of-limit percentage within the window, and hysteresis conditions includes:

[0050] Upon receiving a new valid data frame, the system generates the next frame sequentially, frame by frame, with the acquisition time of the new valid data frame as the right endpoint of the window. A sliding window and calculate the first The rate of temperature rise or temperature fall of each sliding window, where, For the first A sliding window, The sliding window number is the overlap duration of adjacent sliding windows, which is the difference between the sliding window length and the sampling interval. The proportion of exceeding limits within the window is defined as the ratio of the number of valid calculation results in the current judgment window to the total number of valid calculation results within the judgment window for the same early warning rule.

[0051] In the scenario of temperature rise, the following formula is used to determine the first... Are the number of sliding windows exceeded? In the formula, For the first The rate of temperature rise of each sliding window, The alarm threshold for the rate of temperature rise. The sliding window number;

[0052] In a temperature drop scenario, the following formula is used to determine the first... Are the number of sliding windows exceeded? In the formula, For the first The rate of temperature drop of each sliding window, The alarm threshold for the rate of temperature drop. The sliding window number;

[0053] When the current sliding window exceeds the limit, increment the consecutive window exceeding the limit count; when the current sliding window does not exceed the limit, reset the consecutive window exceeding the limit count.

[0054] Only when the consecutive out-of-limit window count reaches the consecutive out-of-limit window count threshold. When the corresponding temperature rise rate warning or temperature fall rate warning is triggered, the following will occur: The minimum number of consecutive out-of-limit windows required to trigger a temperature rise rate warning or temperature fall rate warning; the default value is 3.

[0055] The warning status is updated according to the normal status, observation status, alarm status, escalation status and deactivation status;

[0056] When a single anomaly occurs, the normal state is updated to the observation state and no warning is issued. For temperature rise rate risk and temperature fall rate risk, when the continuous over-limit window count reaches the continuous over-limit window count threshold N, the observation state is updated to the alarm state. For temperature value risk, measuring point pair temperature difference risk, and historical temperature difference risk, when any of the following conditions are met: the number of consecutive over-limits reaches N, the duration of over-limits reaches 10 minutes, or the proportion of valid over-limit calculation results within the judgment window to the total number of valid calculation results reaches 70%, the observation state is updated to the alarm state. When the corresponding anomaly is not continuous and the conditions for entering the alarm state are not met, the observation state is updated to the normal state.

[0057] When the number of consecutive normal occurrences reaches 3 and the duration of normal occurrence reaches 15 minutes, the hysteresis condition is further determined. For warning rules that use exceeding the upper threshold as the trigger condition, the hysteresis condition is met when the risk parameter is lower than the difference between the corresponding upper threshold and the hysteresis. For warning rules that use being lower than the lower threshold as the trigger condition, the hysteresis condition is met when the risk parameter is higher than the sum of the corresponding lower threshold and the hysteresis. The hysteresis is five percent of the absolute value of the corresponding trigger threshold. When the number of consecutive normal occurrences, the duration of normal occurrence, and the hysteresis condition are all met simultaneously, the alarm status is updated to the deactivated status.

[0058] When any of the following conditions are met: the deviation of a temperature-related risk parameter (in degrees Celsius) from the trigger threshold is not less than 5 degrees Celsius; the duration of exceeding the limit is not less than 2 hours; or multiple channels or multiple warning rules are triggered simultaneously, the alarm status is updated to an escalation status. When the corresponding risk parameter falls back or a manual confirmation is received, the escalation status is updated to an alarm status. The duration of exceeding the limit is the time between the acquisition time corresponding to the first valid calculation result of exceeding the limit in the current continuous exceeding the limit process and the acquisition time corresponding to the current valid calculation result of exceeding the limit.

[0059] Generate a record of early warning events, including the relevant section, corresponding channel or measurement point pair, early warning level, triggering conditions, early warning time, and push action.

[0060] Furthermore, the process of updating the warning status and generating warning event records based on risk parameters also includes:

[0061] Upon receiving each valid data frame, the corresponding segment to which the device belongs is queried based on the device identifier in the valid data frame, and the warning rules that are in the enabled state under the segment are loaded.

[0062] The risk parameter calculation corresponding to the loaded early warning rule is performed in parallel. The historical time range required for this calculation is determined based on the acquisition time of the current valid data frame and the time window corresponding to the early warning rule to be executed. Only historical data that has not yet been obtained by the current calculation task within the historical time range is queried, and the incremental calculation of the corresponding risk parameter is performed based on the current valid data frame and the historical data.

[0063] When multiple early warning rules require historical data within the same time window, the historical data obtained from a single query will be provided to the multiple early warning rules for joint use.

[0064] When the receiving order of valid data frames is inconsistent with the acquisition time order, the affected sliding window is determined according to the acquisition time of the late valid data frame, and the risk parameters are recalculated for the affected sliding window.

[0065] The unique message identifier corresponding to the valid data frame is used as the idempotency verification basis for the early warning calculation task; when it is detected that the same early warning rule corresponding to the unique message identifier has been calculated, the calculation of the early warning rule is stopped; before writing the early warning event record, the existence of the corresponding early warning event record is verified according to the unique message identifier and the corresponding rule identifier, and the repeated writing is stopped if the record already exists.

[0066] Furthermore, the storage of temperature time-series data and the recording of early warning events, and the determination of the corresponding early warning handling method based on the updated early warning status, wherein when the early warning status enters an alarm state or escalation state, multi-channel push is executed according to the corresponding early warning level, including:

[0067] Write the temperature timing data corresponding to each channel of each device into the timing database;

[0068] The section information, equipment information, user information, early warning rule configuration, early warning push configuration, and early warning event records are written into a relational database, and the early warning event records are stored in an early warning information table;

[0069] Write the generated report files and operation logs to object storage;

[0070] When the warning status is updated to the alarm status, the notification object bound to the corresponding section is determined according to the warning push configuration. An SMS notification is generated through the SMS gateway, a mobile application notification is generated through the application message push service, and an in-platform message is generated in the platform warning information list.

[0071] A 30-minute silence period is set for repeated push notifications of the same early warning event, where the silence period is the time interval at which repeated push notifications of the same early warning event are stopped. The number of push notifications for the same early warning event is limited to no more than 10 times. For SMS notifications, the number of early warning SMS messages received by the same notification number within a natural day is limited to no more than 20. When the limit is reached, the sending of early warning SMS messages to the notification number is suspended. When the corresponding limit is reached, the push notification to the notification number is suspended. The same early warning event refers to an early warning event that persists for the same section, the same channel, the same measuring point pair, and the same early warning rule, while the previous alarm state has not yet entered the de-escalation state.

[0072] When the warning status is updated to an upgraded status, telephone voice notification and notification to superiors will be added, and the silent period will be adjusted to 10 minutes.

[0073] Furthermore, after implementing tiered, multi-channel push notifications, it also includes:

[0074] Based on the received configuration modification information, the acquisition interval, channel engineering name, and early warning threshold are sent to the designated temperature acquisition terminal via the downlink command channel;

[0075] Through a unified HTTPS data interface with a RESTful style, data query, device management, rule configuration, and report export services are provided to web and mobile application clients.

[0076] Based on temperature time-series data, equipment information, and early warning event records, the system generates equipment overview data, historical temperature query results, temperature curves, early warning information lists, and statistical analysis results.

[0077] The equipment overview data includes battery remaining capacity, ambient temperature, and real-time temperature of each channel. The temperature curve includes a temperature trend curve for the past five days. The historical temperature query results are queried in combination according to time conditions and equipment conditions.

[0078] Generate the corresponding report file according to the report export request, write the report file to object storage, and provide the report file to the web page or mobile application.

[0079] Beneficial effects:

[0080] As can be seen from the above technical solutions, the present invention provides an intelligent early warning method for concrete temperature monitoring, which has the following beneficial effects:

[0081] 1. Improving the Correspondence between Multi-Monitoring Point Temperature Risk Calculation and Engineering Physical Relationships: This application does not merely use the physical channel number of the acquisition equipment as the risk calculation object, but further establishes semantic information such as pouring zone, point type, embedment depth, spatial coordinates, associated points, and structural parts for each channel, and makes this information directly participate in the selection of early warning rules and the pairing of measuring points. Therefore, the same set of raw temperature data has already formed a correspondence with the actual monitoring location in the concrete structure before entering the risk calculation. The temperature difference obtained after pairing internal points and surface points according to the inside-outside relationship can reflect the difference between the internal hydration heat accumulation state and the surface heat dissipation state of the concrete; measuring points with similar embedment depths, after being paired according to the same layer relationship, form risk parameters for the temperature difference in the same layer space. Thus, the risk calculation object is transformed from abstract channel data into measuring point data with definite engineering meaning, so that the subsequent early warning conditions establish a more direct correspondence with the actual heat conduction state of the concrete.

[0082] 2. Reducing the impact of unequal sampling intervals, retransmission, and data gaps on temperature change rate calculation: When a device uploads data via a mobile communication network, the actual arrival time is affected by network latency, retransmission, and device clock differences. Adjacent data does not always arrive strictly according to a fixed sampling period. This application constructs a sliding window based on the actual acquisition time of the device rather than the server's receiving time, sorts the sampling points within the window according to the acquisition time, and then filters out overly dense sampling points based on the minimum effective sampling interval. Simultaneously, it checks the data coverage on the left side of the window. Only the effective sampling point sequence that meets the data coverage condition is included in the rate calculation, allowing the time interval to directly participate in the calculation process. Based on this, the endpoint difference method calculates the rate based on the temperature changes at the beginning and end of the effective window, and the least squares fitting slope method further utilizes multiple effective sampling points within the window to reflect the overall change trend. Therefore, compared to simply comparing data from two receiving times, the dependence on the data arrival order and sampling interval changes is reduced.

[0083] 3. Reducing Information Loss from Single-Threshold Judgment through Joint Evaluation of Multiple Risk Parameters: Concrete temperature risk is not only manifested in absolute temperature exceeding a certain value, but also in excessive surface-to-internal temperature differences, abnormal temperature differences within the same layer, excessively rapid temperature rise or fall per unit time, and excessive cumulative temperature changes over a long period. This application generates risk parameters such as temperature values, temperature differences between measuring points, temperature change rates, and historical temperature differences, and combines calculated values, exceeding limits, and verification results from associated points into a risk parameter vector. Therefore, the same monitoring data is subject to evaluation of single-point conditions, spatial relationships, and temporal changes. When an anomaly occurs in a certain channel, checking whether the same type and direction of anomalies appear at its physically associated points helps distinguish between engineering risks formed by multiple points of common change and isolated anomalies generated by a single sensor or channel. This ensures that the early warning information not only reflects "whether the limit is exceeded," but also the physical consistency between the exceeding-limit data.

[0084] 4. Suppressing repetitive state switching near the critical value through continuous over-limit, duration, and hysteresis mechanisms: When temperature sampling data is affected by measurement accuracy, environmental disturbances, and short-term temperature changes, a single sampling result may briefly exceed the alarm threshold. If a formal alarm is generated immediately for each over-limit occurrence, it is easy to form a large number of short-duration events. This application divides the early warning process into normal, observation, alarm, escalation, and de-escalation states. Multiple consecutive sliding window over-limit conditions are used for the temperature change rate, and other risks are confirmed by combining the number of consecutive over-limit occurrences, duration, and the proportion of over-limit occurrences within the window, ensuring that a single anomaly first enters the observation state. When the alarm is de-escalated, the number of consecutive normal occurrences, the duration of normal occurrence, and the hysteresis condition are simultaneously assessed, ensuring that the risk parameter must leave the critical region near the threshold before entering the de-escalation state. Therefore, small repetitive changes of the same risk near the threshold will not continuously cause frequent switching between alarms and de-escalation, which is beneficial to improving the continuity of the early warning event lifecycle.

[0085] 5. Improve the real-time processing efficiency of early warning calculation under continuous reporting from a large number of devices: Under continuous sampling conditions involving multiple sections and multiple devices, if all historical data of a certain channel is rescanned for each received frame of data, the computational load will continuously increase with the accumulation of historical data. This application, based on the acquisition time of the currently added data and the time window required by the corresponding rule, only queries the historical data involved in the current risk parameter calculation, and allows multiple rules requiring the same historical window to share the data obtained once; for late data, the affected sliding window is determined according to the acquisition time of the late data, and only the relevant window is recalculated. At the same time, the unique data identifier formed by the device identifier and frame sequence number is used to control duplicate processing, and idempotency verification is performed in the early warning calculation and event writing stages. Thus, normal newly added data, late data, and duplicate data enter the corresponding calculation paths respectively, avoiding full duplicate scanning and repeated generation of the same event, which is more suitable for online early warning processing under continuous high-frequency sampling conditions.

[0086] 6. Establishing a complete traceability relationship between temperature monitoring data, early warning status, and notification results: This application organizes storage separately for different data types. Continuously generated channel temperature data is written to a time-series database, while structured information such as sections, equipment, rule configurations, and early warning events is written to a relational database. Result data, such as report files, is stored independently. Early warning events record the section, channel or measuring point pair, triggering conditions, early warning level, early warning entry time, and push action, enabling subsequent queries to trace back to the corresponding measuring point and its temperature time-series data. In alarm status, SMS, mobile application messages, and in-platform messages are sent according to the notification configuration. In escalation status, telephone voice messages and notifications to superiors are added, while silent periods and frequency limits are used to control repeated pushes. Therefore, a continuous data relationship is formed between monitoring data, risk calculation, status changes, event records, and notification handling, which is beneficial for construction process verification and subsequent quality traceability.

[0087] It should be understood that all combinations of the foregoing concepts and the additional concepts described in more detail below can be considered part of the inventive subject matter of this disclosure, provided that such concepts do not contradict each other.

[0088] The foregoing and other aspects, embodiments, and features of the teachings of the present invention will be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the invention, such as features and / or beneficial effects of exemplary embodiments, will become apparent from the following description or may be learned through practice of specific embodiments according to the teachings of the present invention. Attached Figure Description

[0089] The accompanying drawings are not drawn to scale. In the drawings, each identical or nearly identical component shown in the various figures may be denoted by the same reference numeral. For clarity, not every component is labeled in each figure. Embodiments of various aspects of the invention will now be described by way of example and with reference to the accompanying drawings, wherein:

[0090] Figure 1 This is a flowchart illustrating an intelligent early warning method for monitoring concrete temperature in an embodiment of this application.

[0091] Figure 2 This is a flowchart of step S102 of a concrete temperature monitoring and intelligent early warning method in an embodiment of this application.

[0092] Figure 3 This is a flowchart of step S104 of a concrete temperature monitoring and intelligent early warning method in an embodiment of this application.

[0093] Figure 4 This is a flowchart of step S106 of a concrete temperature monitoring and intelligent early warning method in an embodiment of this application.

[0094] Figure 5 This is a flowchart of step S108 of a concrete temperature monitoring and intelligent early warning method in an embodiment of this application.

[0095] Figure 6 This is a flowchart of step S110 of a concrete temperature monitoring and intelligent early warning method in an embodiment of this application.

[0096] Figure 7 This is a flowchart of step S112 of a concrete temperature monitoring and intelligent early warning method in an embodiment of this application.

[0097] Figure 8 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0098] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art to which this invention pertains.

[0099] The terms "first," "second," and similar words used in the specification and claims of this patent application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular forms of "an," "a," or "the," etc., do not indicate a quantity limitation, but rather indicate the presence of at least one. Terms such as "comprising" or "including" mean that the element or object preceding "comprising" encompasses the features, integrals, steps, operations, elements, and / or components listed following "comprising" or "including," and do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0100] Existing temperature monitoring methods typically use single-point temperature thresholds or fixed-time interval temperature differences as the basis for judgment. Although some systems have achieved remote data acquisition and cloud display, the physical relationship between different monitoring points is not fully involved in the early warning calculation. In the presence of network latency, retransmission, uneven sampling time, missing data at measuring points, and short-term fluctuations, judgments based directly on single measurements or data at fixed time points are easily affected by the sampling status. At the same time, the heat conduction state reflected by points inside the concrete, surface points, and measuring points in the same layer are not the same. Using a single-point threshold uniformly makes it difficult to comprehensively evaluate the temperature difference between the inside and the surface, the rate of temperature change, and continuous abnormal conditions, thus affecting the stability of temperature risk assessment and the effectiveness of early warning information.

[0101] Therefore, embodiments of the present invention provide an intelligent early warning method for concrete temperature monitoring, executed by a cloud server, as described above. Figure 1 ,include:

[0102] Step S102: Establish a three-level data organization relationship for sections, equipment, and channels, and configure semantic tags for monitoring points, physical relationships between measuring points, and early warning rule templates for each channel.

[0103] Step S104: Receive the raw data frame uploaded by the temperature acquisition terminal, perform device authentication, integrity verification, duplicate data detection and data quality marking on the raw data frame to obtain a valid data frame.

[0104] Step S106: Asynchronously align the valid data frames with the corresponding historical data according to the acquisition time, generate the valid sampling point sequence for each channel, and match the early warning rules and generate measurement point pairs based on the semantic tags of the monitoring points and the physical relationship of the measurement points.

[0105] Step S108: Based on the effective sampling point sequence and measurement point pairs, calculate the temperature value risk parameter, the temperature difference risk parameter of the measurement point pair, the temperature change rate risk parameter, and the historical temperature difference risk parameter, and determine the over-limit state corresponding to each risk parameter to form a risk parameter vector.

[0106] Step S110: Update the warning status and generate a warning event record based on the risk parameter vector's out-of-limit status, number of consecutive out-of-limit times, duration, proportion of out-of-limits within the window, and hysteresis conditions.

[0107] Step S112: Store temperature time-series data and early warning event records, and determine the corresponding early warning handling method based on the updated early warning status. When the early warning status enters the alarm status or escalation status, perform multi-channel push according to the corresponding early warning level.

[0108] The cloud server first organizes temperature monitoring objects according to construction management relationships, establishing a hierarchical correspondence between construction sections, on-site temperature acquisition equipment, and the temperature acquisition channels connected to the equipment. It further maps these channels to actual pouring zones, monitoring point locations, and physical relationships. After the on-site terminals upload data, the cloud server checks the data source and integrity, eliminating illegal equipment data, damaged data, and duplicate data. Then, using the actual acquisition time of the equipment as the time base, it organizes the current data and corresponding historical data into a continuous and valid sampling sequence. Subsequently, based on the physical location relationships between the measuring points, it constructs a combination of measuring points to participate in temperature difference evaluation, and forms multi-dimensional risk parameters based on single-point temperature, temperature difference between measuring points, temperature change rate, and temperature difference over historical time spans. These risk parameters are not directly converted into external alarms; instead, they are combined with the duration of the anomaly and the current warning status to generate corresponding warning events, which are then handled according to the warning level.

[0109] The semantics of engineering objects, data quality, time series processing, and early warning status judgment are organized into a continuous processing chain. Risk judgment is not performed directly on abstract equipment channels, but rather after the engineering meaning of measuring points is identified and the data time series is processed. This ensures that the data entering the risk calculation has both engineering location and time series meaning. For on-site communication situations such as network retransmission, data retransmission, and asynchronous sampling across different channels, the time relationship and message duplication issues are handled separately by the acquisition time and unique data identifier. The processing results are then submitted to the risk calculation process, avoiding the direct equation of communication anomalies with concrete temperature anomalies.

[0110] Because the data used in risk calculation has undergone source verification, integrity checks, duplicate data identification, and time series processing, the temperature data entering the early warning stage has a relatively stable correspondence with the actual sampling status on site. Different measuring points are also associated based on the construction area and physical location, allowing changes in temperature at a single point to be evaluated together with changes at neighboring or physically related measuring points. By adding a status confirmation process before the official push, short-term single-point anomalies will not immediately turn into external alarms, while persistent anomalies will enter the corresponding early warning level according to their persistence. Therefore, this implementation method simultaneously addresses data quality, spatial relationships, temporal variations, and anomaly persistence issues, making the temperature risk assessment process more consistent with the actual characteristics of continuous construction monitoring data.

[0111] In some embodiments, a three-level data organization relationship is established among sections, devices, and channels. Semantic tags for monitoring points, physical relationships between measuring points, and early warning rule templates are configured for each channel, referring to... Figure 2 ,include:

[0112] Step S1021: Establish the correspondence between the bidding section and the construction project, associate at least one device under each bidding section, and associate one to eight channels under each device.

[0113] Step S1022: Configure semantic tags for monitoring points for each channel, including pouring zone, point type, burial depth, spatial coordinates, associated points and structural parts.

[0114] Among them, the pouring zone is used to identify the pouring section or construction zone to which the measuring point belongs, the point type is used to identify whether the measuring point is an internal point, surface point, environmental point or rebar point, the embedment depth is the distance of the measuring point from the concrete surface, the spatial coordinates are the position of the measuring point in three-dimensional space, the associated point is the identifier of other measuring points that are physically associated with the current measuring point, and the structural part is used to identify the structural component where the measuring point is located.

[0115] Step S1023: Match the corresponding early warning rule template according to the semantic tags of the monitoring points, wherein the temperature rise rate early warning rule is matched for internal points, the temperature drop rate early warning rule is matched for surface points, the inner and outer surface temperature difference early warning rule is matched for internal and surface points in the same pouring zone, and the same layer temperature difference early warning rule is matched for measuring points in the same pouring zone with a burial depth difference of no more than 0.1 meters.

[0116] Step S1024: For early warning rules that require two measuring points to participate in the calculation, at least one pair of measuring points is generated based on the associated points, pouring zones, point types, burial depths and the matched early warning rule template. Each pair of measuring points includes a first measuring point identifier, a second measuring point identifier and a physical relationship type.

[0117] Step S1025: Configure an initial threshold of 1.5 degrees Celsius per hour for the internal point temperature rise rate warning rule, configure an initial threshold of 3.0 degrees Celsius per hour for the surface point temperature drop rate warning rule, configure an initial threshold of 15 degrees Celsius for the inner-to-outer temperature difference warning rule, and configure an initial threshold of 10 degrees Celsius for the same-layer temperature difference warning rule; the initial threshold is used as the rate alarm threshold for the warning rule of the corresponding point type when the burial depth adjustment is not performed.

[0118] Each warning rule includes a rule identifier, the segment it belongs to, the rule name, the parameter type, the comparison relationship, the threshold, the rule type, and the activation status. The comparison relationship includes greater than, less than, not less than, not greater than, and range.

[0119] During the project configuration phase, an equipment set is established for each construction section, and a corresponding physical acquisition channel is created for each piece of equipment. Each channel, in addition to its equipment number and channel number, records the pouring zone, point type, embedment depth, spatial coordinates, associated points, and structural location. The point type distinguishes between internal points, surface points, environmental points, and reinforcement points; the pouring zone limits the monitoring points participating in the risk assessment of the same construction area; and the embedment depth and spatial location characterize the actual position of the monitoring point within the concrete structure. The rule matching process determines the corresponding risk assessment type based on these labels. For example, internal points correspond to temperature rise rate assessment, surface points correspond to temperature fall rate assessment, internal and surface points within the same pouring zone form an inner-to-outer temperature difference assessment relationship, and measuring points with similar embedment depths form a same-layer temperature difference assessment relationship.

[0120] In this embodiment, the engineering semantics are not merely channel display names in the interface, but rather input parameters for the risk calculation process. Whether a monitoring point participates in a certain type of early warning, which monitoring point it forms a measurement point combination with, and which type of temperature difference rule is used are all determined by structured tags and physical relationships. This approach separates the channel number at the equipment level from the monitoring location in the construction structure: even if the same model of equipment contains one to eight channels, each channel in different sections still acquires different engineering meanings according to its actual buried location. Based on this, the rule engine performs different risk assessments on measurement points with different physical roles, establishing a definite mapping relationship between monitoring data and the location of concrete structures from the data organization stage.

[0121] By using structured monitoring point semantics, rule configuration no longer relies entirely on manual selection for each channel. Internal points, surface points, and same-layer measuring points enter the corresponding risk assessment process according to predetermined label conditions, enabling data organization based on unified engineering attributes even when adding equipment or adjusting channels in large-scale projects. The calculation of temperature difference between the inside and surface is limited to measuring points with a clear internal-to-surface relationship, while the calculation of temperature difference within the same layer is limited to measuring points in the same construction area with similar burial depths. Therefore, the resulting temperature differences have clear engineering physical meanings. For construction sites involving multiple sections, multiple equipment, and a large number of measuring points, this approach also facilitates the unified loading and adjustment of rule parameters according to the section, reducing the impact of changes in equipment physical numbers on the upper-level risk calculation logic.

[0122] In some embodiments, raw data frames uploaded by a temperature acquisition terminal are received, and device authentication, integrity verification, duplicate data detection, and data quality marking are performed on the raw data frames to obtain valid data frames, as referred to... Figure 3 ,include:

[0123] Step S1041: Receive the original data frame containing the device identifier, channel identifier, channel engineering name, acquisition time, temperature value, signal quality, data status, frame sequence number, and checksum, and generate the reception time when receiving the original data frame; wherein, the device identifier is the unique identifier of the temperature acquisition terminal, used for device authentication and segment attribution query; the channel identifier is the physical measurement point number of the temperature acquisition terminal; the channel engineering name field is used to record the engineering display name of the channel in the construction monitoring scenario, and the engineering display name is independent of the point type used to represent internal points, surface points, environmental points, or rebar points; the acquisition time is the timestamp generated when the temperature acquisition terminal acquires the temperature value; the reception time is the timestamp generated when the cloud server receives the original data frame; the signal quality is the communication signal strength or signal-to-noise ratio; the data status is used to distinguish normal acquisition data, supplementary transmission data, and the first frame data after the temperature acquisition terminal restarts; the frame sequence number is the incremental sequence number generated by the temperature acquisition terminal; and the checksum is the integrity check value generated based on the frame content of the original data frame.

[0124] Step S1042: Verify the legality of the corresponding temperature acquisition terminal according to the device identifier, and verify the integrity of the original data frame according to the check code; when the temperature acquisition terminal fails the legality verification or the original data frame fails the integrity verification, mark the data quality field of the corresponding data as invalid, and stop using the original data frame for subsequent timing alignment and early warning calculation.

[0125] Step S1043: Use the combination of device identifier and frame sequence number as a unique message identifier, and query the unique message identifier in the recent message identifier cache with a preset cache duration.

[0126] Step S1044: When a unique message identifier is found, the original data frame is marked as duplicate and its use in subsequent warning calculations is stopped; when no unique message identifier is found, the unique message identifier is written to the most recent message identifier cache, the data quality field of the original data frame is marked as valid, and the data frame that has completed the unified format conversion is written to the message queue; wherein, the data quality field is used to indicate the data quality judgment result of the data frame, and the valid status indicates that the corresponding data frame has passed device authentication, integrity verification and duplicate data detection.

[0127] The on-site temperature acquisition terminal generates a data frame containing device identifier, channel identifier, engineering name, acquisition time, temperature value, signal quality, data status, frame sequence number, and checksum, and uploads it to the cloud. The cloud generates a separate reception time upon receiving data; the acquisition time reflects the moment the temperature was formed, and the reception time reflects the moment the data entered the cloud, both stored separately. The cloud checks the legitimacy of the uploading device based on the device identifier and the completeness of the data frame content based on the checksum. The device identifier and frame sequence number are combined to form a unique data identifier. This unique data identifier is then searched in a recent message cache with a valid retention period. Existing messages are identified as duplicates; non-existent messages proceed to further processing, and their unique identifiers are written to the cache.

[0128] To address the common data retransmission problem in field wireless communication, a distinction is made between "the same temperature value reappearing" and "the same data frame being repeatedly transmitted." The identification of duplicate data is not simply based on temperature values, but rather on device identification and the device's incrementing frame sequence number to determine the message's identity. Therefore, if the same temperature value is continuously collected by the same channel during a stable temperature period, it is still considered different sampled data, while identical frames generated by network retransmissions are identified as duplicate data. Furthermore, the acquisition time and reception time are stored in different fields, ensuring that subsequent timing calculations use the acquisition time and do not directly alter the time interval in the temperature change rate due to network latency.

[0129] Device legitimacy and integrity verification precede risk calculation, preventing data from abnormal sources or corrupted content from entering the temperature analysis process. Duplicate detection based on unique message identifiers prevents retransmitted data from repeatedly participating in risk calculations, avoiding the same temperature anomaly being counted multiple times for consecutive exceedances due to duplicate messages. Since the original sampling time is not replaced by the server reception time, even if some data is delayed or retransmitted, it is still processed according to the actual temperature collection order. Therefore, the data access phase establishes basic quality control in four aspects: data source, data content, message identity, and time attributes, providing input data with clear temporal meaning for subsequent sliding window calculations.

[0130] In some embodiments, the valid data frames and corresponding historical data are asynchronously time-aligned according to the acquisition time to generate a sequence of valid sampling points for each channel, referring to... Figure 4 ,include:

[0131] Step S1061: Construct a time window with the acquisition time of the current valid data frame as the right endpoint of the window. In the formula, The acquisition time of the current valid data frame. The length of the sliding window is measured in minutes.

[0132] Step S1062: Obtain the set of sampling points arranged in ascending order of acquisition time from the time window: In the formula, The number of sampling points within the time window. For the first The sampling time for each sampling point For the first Temperature values ​​at each sampling point The sampling point number, and .

[0133] Step S1063: Calculate the time interval between adjacent sampling points: In the formula, For the first The sampling point and the first The time interval between sampling points For the first The sampling time for each sampling point For the first The collection time for each sampling point.

[0134] Step S1064: When the time interval between any two adjacent sampling points is less than the preset minimum time interval, delete the sampling point whose temperature value is closer to the median temperature value in the current time window, and retain the sampling point whose temperature value deviates significantly from the median; re-sort the remaining sampling points in ascending order of acquisition time and recalculate the time interval between adjacent sampling points, repeat the filtering process until the time interval between any two adjacent valid sampling points is not less than the preset minimum time interval, and obtain the valid sampling point sequence.

[0135] Step S1065: When the time difference between the acquisition time of the first valid sampling point and the left end of the time window exceeds twice the minimum time interval, the current time window is determined to have insufficient data coverage, and the calculation of the temperature rise rate or temperature fall rate using the current time window is stopped.

[0136] Step S1066: For each channel participating in the temperature difference calculation of the measuring point pair, select a valid sampling point within the corresponding time window whose acquisition time is no later than the right end of the window and whose time difference with the right end of the window is the smallest, and combine the temperature values ​​of the valid sampling points corresponding to different channels into a temperature value pair corresponding to the measuring point pair.

[0137] Step S1067: The effective sampling point sequence is used for temperature change rate calculation only when the number of effective sampling points is not less than two and the first effective sampling point and the last effective sampling point have different acquisition times.

[0138] When processing newly added temperature data, the current data acquisition time is used as the right endpoint of a sliding window. A time range of length L is selected in the historical direction, and all valid data from the corresponding channels within this time range are obtained. After the data is sorted in ascending order of acquisition time, the actual time interval between every two adjacent sampling points is calculated. When the interval between adjacent sampling points is less than the preset minimum time interval, overly dense data is filtered, and the remaining sampling points are reordered and checked until all adjacent valid sampling points meet the minimum time interval requirement. After filtering, the data coverage in the left area of ​​the window is further checked. When the first valid data is too far from the left edge of the window, the window is marked as having insufficient data coverage and is not used for temperature change rate evaluation.

[0139] Describing window effectiveness using actual time intervals rather than the number of sampling points is suitable for situations where the sampling period varies due to network conditions and retransmission processes. A fixed number of sampling points represents different actual time spans at different sampling intervals, while using window length and actual timestamps ensures that each rate calculation corresponds to a defined observation time range. Simultaneously, minimum sampling interval filtering controls the excessive weighting of densely transmitted data within a short period on the fitting results, and window data coverage checks prevent the calculation of the entire time window's rate of change when only a small amount of data is available at the end of the window. For different channels participating in temperature difference evaluation, a unified right-hand window time is used as a common time reference, giving the temperature differences between measurement points a corresponding temporal meaning.

[0140] By establishing a sliding window using the actual acquisition time, changes in data reporting frequency do not directly alter the physical time length represented by the window. Overly dense sampled data, after filtering, will not occupy an excessive proportion in temperature trend calculations due to a large number of data points within the same short time period; rate calculations are stopped when too much data is missing at the left end of the window, avoiding the use of data with significantly insufficient actual span to extrapolate long-term temperature changes. In situations such as concentrated retransmission after network outages at construction sites, this process restores the data to its actual acquisition time position, helping to maintain the correspondence between the temperature change process and the actual construction time.

[0141] In some embodiments, calculating the rate of temperature rise or the rate of temperature fall includes:

[0142] The temperature rise rate or temperature fall rate is calculated using the endpoint difference method or the least squares fitting slope method. The temperature change rate is calculated only for valid sampling point sequences with at least two valid sampling points and different sampling times. The sampling time of each valid sampling point is converted into a time value in minutes according to a unified time base before being included in the calculation.

[0143] When using the endpoint difference method, the calculation is based on the first and last valid sampling points in the valid sampling point sequence: In the formula, The rate of temperature rise or fall is expressed in degrees Celsius per hour. The temperature value of the first valid sampling point. This is the temperature value of the last valid sampling point. The time of collection for the first valid sampling point. The time of the last valid sampling point. The time difference is in minutes and is not zero. The value 60 is used to convert the temperature change rate in minutes to the temperature change rate in hours.

[0144] When using the least squares fitting slope method, perform univariate linear regression on the effective sampling point sequence: In the formula, To fit the temperature value, The data collection time is measured in minutes. The slope of the fitted line, This is the intercept of the fitted line.

[0145] The slope of the fitted line is calculated using the following formula: In the formula, The slope of the fitted line, The number of effective sampling points, The valid sampling point number, For the first The collection time for each valid sampling point For the first Temperature values ​​of each valid sampling point.

[0146] Calculate the rate of temperature rise or rate of temperature fall based on the slope of the fitted straight line: In the formula, The rate of temperature rise or fall is expressed in degrees Celsius per hour. The slope of the fitted straight line is expressed in degrees Celsius per minute. The value 60 is used to convert the slope of the fitted straight line in minutes to the rate of temperature rise or fall in hours. This indicates a rise in temperature. This indicates a drop in temperature.

[0147] For a sampling point sequence that meets the requirements for effective data quantity and coverage, the temperature change rate is calculated using either the endpoint difference method or the least squares fitting slope method, depending on the data quality and computational configuration. The endpoint difference method uses the first and last sampling points within the effective window to establish a proportional relationship between the temperature difference between the two points and the actual time difference, converting this into a temperature change rate in hours. The least squares fitting slope method utilizes all effective sampling points within the window, converting the acquisition time into a uniform minute value, obtaining the fitting slope of temperature change over time through univariate linear fitting, and then converting the slope into a temperature change rate in hours. The positive or negative sign of the obtained rate represents an overall temperature increase or decrease, respectively.

[0148] The endpoint difference method and the least squares fitting slope method correspond to different data utilization approaches. The former only requires the first and last two valid sampling points of the window, making the calculation process more direct; the latter considers the time position and temperature value of multiple sampling points within the window simultaneously, focusing more on extracting the overall trend of change throughout the window. Neither method assumes that adjacent sampling points have exactly the same time interval, but directly uses the actual acquisition time in the calculation. Therefore, when network latency, retransmission, or device clock factors cause the sampling interval to be uneven, as long as the valid sampling points meet the time coverage requirements, the rate of temperature change is still determined by the relationship between the amount of temperature change and the actual time.

[0149] The rate of temperature change reflects not whether the temperature is too high at a particular moment, but rather the speed of temperature change over a certain period of time. The construction risks associated with a rapid increase in internal concrete temperature differ from those with a rapid decrease in surface temperature. Therefore, by storing the direction and absolute value of the rate separately, the rule engine can perform different judgments for internal and surface points. The endpoint difference method reduces computational load and is suitable for monitoring windows with relatively stable data quality; the least squares fitting method utilizes more effective sampling information and exhibits statistical characteristics different from the two-point difference for small perturbations at a single sampling point. Both calculation paths share the same effective sampling sequence, ensuring consistency between the data preprocessing logic and the rate evaluation logic.

[0150] In some embodiments, based on the effective sampling point sequence and measurement point pairs, temperature value risk parameters, measurement point pair temperature difference risk parameters, temperature change rate risk parameters, and historical temperature difference risk parameters are calculated respectively, and the out-of-limit states corresponding to each risk parameter are determined to form a risk parameter vector, which is then referenced. Figure 5 ,include:

[0151] Step S1081: Compare the current temperature value of each channel with the corresponding upper or lower temperature threshold to obtain the temperature risk parameter and the corresponding over-limit status.

[0152] Step S1082: Based on the time-aligned temperature values ​​corresponding to the first and second measuring points of the measuring point pair, calculate the absolute value of the temperature difference between the first and second measuring points, determine the corresponding maximum allowable temperature difference based on the physical relationship type of the measuring point pair, compare the absolute value of the temperature difference with the maximum allowable temperature difference, and obtain the temperature difference risk parameters and corresponding over-limit states of the measuring point pair.

[0153] Step S1083: Calculate the absolute value of the temperature difference between the current acquisition time and the historical acquisition time corresponding to the preset historical time interval for the same channel. Compare the absolute value of the temperature difference with the maximum allowable historical temperature difference to obtain the historical temperature difference risk parameter and the corresponding over-limit status. Wherein, when there is no valid temperature data for the historical acquisition time corresponding to the preset historical time interval, the current historical temperature difference calculation value is not generated, and the corresponding historical temperature difference risk parameter is marked as insufficient data status. The insufficient data status is not treated as an over-limit status.

[0154] Step S1084: Compare the temperature rise rate with the temperature rise rate alarm threshold, and compare the absolute value of the temperature drop rate with the temperature drop rate alarm threshold to obtain the temperature change rate risk parameter and the corresponding over-limit status.

[0155] Step S1085: When adjusting the rate threshold according to the burial depth, calculate the adjusted rate threshold according to the following formula: In the formula, The rate threshold is adjusted based on the burial depth. The base rate alarm threshold is determined based on the current measurement point type and corresponding early warning rules. The depth of the measuring point is measured in meters.

[0156] Step S1086: When the risk parameter of any channel is in an out-of-limit state, obtain the risk parameter corresponding to the associated point of the channel; when the associated point shows a same-direction anomaly, mark the risk of the current channel as a verified risk; when the associated point does not show a same-direction anomaly, mark the risk of the current channel as an isolated risk and reduce the push priority of the corresponding warning event; where, a same-direction anomaly means that the current channel and the associated point are both in an out-of-limit state under the same risk parameter type, and for risk parameters with a temperature change direction, the temperature change direction corresponding to the current channel and the associated point is the same.

[0157] Step S1087: Combine the calculated values ​​of each risk parameter, the over-limit status, and the risk label into a risk parameter vector.

[0158] Risk parameter calculation includes multiple independent but shared time-series data evaluation dimensions. Single-point temperature evaluation compares the current effective temperature value with the configured upper or lower temperature limit; the temperature difference evaluation between measurement points calculates the absolute value of the temperature difference based on the temperature values ​​of two measurement points with a defined physical relationship at the same time reference, and compares it with the maximum allowable temperature difference for the corresponding physical relationship; the historical temperature difference evaluation compares the change in temperature between the current temperature of the same channel and the temperature corresponding to a predetermined historical time interval; and the temperature change rate evaluation compares the heating or cooling rate calculated using an effective sliding window with the corresponding alarm threshold. Each evaluation result includes the calculated risk parameter value and its exceeding status, and together they enter the early warning status judgment process.

[0159] Different risk parameters correspond to different dimensions of concrete temperature state. Absolute temperature describes the current state of a single point; temperature difference between measuring points describes the temperature difference between two locations with a physical relationship in space; historical temperature difference describes the cumulative change at the same location over a longer time span; and the rate of temperature change describes the severity of the change process within a continuous time window. Therefore, historical temperature difference and rate of temperature change are not different expressions of the same parameter: the former depends on the temperature difference between two fixed time locations, while the latter is calculated based on the time change relationship within a sliding window. For physically related measuring points, after a single point exceeds the limit, the risk state of the corresponding related points is further checked, and verified risks or isolated risks are marked according to whether the associated anomalies are consistent.

[0160] Spatial temperature difference, temporal temperature difference, and rate of temperature change are calculated separately to avoid covering different physical problems with only one type of threshold. For example, even if the current temperature has not exceeded the absolute temperature threshold, rapid surface cooling may still pose a risk of temperature change rate fluctuations; even if the short-term rate of temperature change is normal at the same measuring point, the cumulative temperature difference over a longer period may still reach historical temperature difference conditions. Verification of associated monitoring points further utilizes the common changes between multiple monitoring locations: when multiple related measuring points simultaneously experience anomalies in the same direction, the risk exhibits stronger data consistency; when only a single point anomaly exists, this risk is distinguished from multiple-point common anomalies, providing a basis for subsequent event prioritization and sensor status checks.

[0161] In some embodiments, the warning status is updated based on the out-of-limit status, consecutive out-of-limit count, duration, out-of-limit percentage within the window, and hysteresis conditions of the risk parameter vector, and a warning event record is generated, referring to... Figure 6 ,include:

[0162] Step S1101: Upon receiving a new valid data frame, generate the first frame frame by frame, using the acquisition time of the new valid data frame as the right endpoint of the window. A sliding window And calculate the first The rate of temperature rise or temperature fall of each sliding window, where, For the first A sliding window, The sliding window number is the overlap duration of adjacent sliding windows, which is the difference between the sliding window length and the sampling interval. The percentage of out-of-limit states within a window is defined as the ratio of the number of valid calculation results in the current judgment window to the total number of valid calculation results within that judgment window for the same early warning rule.

[0163] In the scenario of temperature rise, the following formula is used to determine the first... Are the number of sliding windows exceeded? In the formula, For the first The rate of temperature rise of each sliding window, The alarm threshold for the rate of temperature rise. This is the sliding window number.

[0164] In a temperature drop scenario, the following formula is used to determine the first... Are the number of sliding windows exceeded? In the formula, For the first The rate of temperature drop of each sliding window, The alarm threshold for the rate of temperature drop. This is the sliding window number.

[0165] Step S1102: When the current sliding window exceeds the limit, increment the consecutive window exceeding the limit count; when the current sliding window does not exceed the limit, reset the consecutive window exceeding the limit count.

[0166] Step S1103: Only when the consecutive out-of-limit window count reaches the consecutive out-of-limit window count threshold. When the corresponding temperature rise rate warning or temperature fall rate warning is triggered, the following will occur: The minimum number of consecutive over-limit windows required to trigger a temperature rise rate warning or temperature fall rate warning; the default value is 3.

[0167] Step S1104: Update the warning status according to the normal status, observation status, alarm status, escalation status and deactivation status.

[0168] When a single anomaly occurs, the normal state is updated to the observation state and no warning is issued. For temperature rise rate risk and temperature fall rate risk, when the continuous over-limit window count reaches the continuous over-limit window count threshold N, the observation state is updated to the alarm state. For temperature value risk, measuring point pair temperature difference risk, and historical temperature difference risk, when any of the following conditions are met: the number of consecutive over-limits reaches N, the duration of over-limits reaches 10 minutes, or the proportion of valid over-limit calculation results within the judgment window to the total number of valid calculation results reaches 70%, the observation state is updated to the alarm state. When the corresponding anomaly is not continuous and the conditions for entering the alarm state are not met, the observation state is updated to the normal state.

[0169] When the number of consecutive normal occurrences reaches 3 and the duration of normal occurrence reaches 15 minutes, the hysteresis condition is further determined. For warning rules that use exceeding the upper threshold as the trigger condition, the hysteresis condition is met when the risk parameter is lower than the difference between the corresponding upper threshold and the hysteresis. For warning rules that use being lower than the lower threshold as the trigger condition, the hysteresis condition is met when the risk parameter is higher than the sum of the corresponding lower threshold and the hysteresis. The hysteresis is 5% of the absolute value of the corresponding trigger threshold. When the number of consecutive normal occurrences, the duration of normal occurrence, and the hysteresis condition are all met simultaneously, the alarm status is updated to the cleared status.

[0170] When any of the following conditions are met: the deviation of a temperature-related risk parameter (in degrees Celsius) from the trigger threshold is not less than 5 degrees Celsius; the duration of continuous exceedance is not less than 2 hours; or multiple channels or multiple warning rules are triggered simultaneously, the alarm status will be updated to an escalation status. When the corresponding risk parameter falls back or a manual confirmation is received, the escalation status will be updated to an alarm status. The duration of continuous exceedance is the time between the acquisition time corresponding to the first valid calculation result of the current continuous exceedance process and the acquisition time corresponding to the current valid calculation result of the exceedance.

[0171] Step S1105: Generate a warning event record that includes the section to which it belongs, the corresponding channel or measurement point pair, the warning level, the triggering condition, the time of entering the warning and the push action.

[0172] For each new frame of valid data received, a new sliding window is formed based on its acquisition time, and the corresponding risk parameters are calculated. For the rate of temperature change, it is continuously recorded whether each sliding window exceeds the heating rate threshold or cooling rate threshold, and a continuous out-of-limit window count is maintained. The warning status transitions through normal, observation, alarm, escalation, and de-escalation: a single anomaly initially enters the observation state; upon reaching the corresponding persistence judgment condition, it enters the alarm state; when the alarm persistence reaches the escalation condition, it enters the escalation state; after risk recovery, it is not immediately de-escalated, but instead, the number of consecutive normal occurrences, the duration of normal occurrence, and the hysteresis condition are checked simultaneously, and it only enters the de-escalation state after all conditions are met. This status object continuously records the occurrence, persistence, escalation, and recovery process of the anomaly.

[0173] Instead of equating each rule calculation result directly to an independent alarm, the system maps continuous data judgment results to the same state lifecycle. A continuous multi-window over-limit mechanism identifies whether the temperature rise and fall rates are persistently abnormal; the observation state absorbs occasional single over-limit occurrences; and the hysteresis condition handles small, recurring changes near the threshold. Duration, number of consecutive occurrences, and the proportion of over-limits within a window reflect different forms of abnormal persistence. The escalation condition further combines the degree of deviation, duration, and the combined triggering of multiple channels and rules to determine a higher level of response. The state object thus becomes an intermediate judgment layer between the risk parameter calculation results and the warning event.

[0174] A single sampling anomaly may be caused by transient interference, measurement error, or short-term environmental changes. If a formal alarm is immediately triggered, it will increase the number of short-lived warning events. By adopting an observation state, a formal alarm is only triggered when the anomaly persists for a set period, thus verifying the individual anomaly in terms of temporal continuity. The deactivation phase requires continuous data recovery and exceeding the hysteresis range to prevent frequent deactivation and re-triggering of the alarm when the temperature value remains near the threshold for an extended period. For events with longer durations, larger deviations, or simultaneous anomalies at multiple monitoring locations, a higher-level notification method is used by escalating the status, ensuring a correlation between the intensity of the warning response and the duration of the risk.

[0175] In some embodiments, the process of updating the warning status and generating a warning event record based on risk parameters further includes:

[0176] Upon receiving each valid data frame, the corresponding segment to which the device belongs is queried based on the device identifier in the valid data frame, and the warning rules that are enabled under the segment are loaded.

[0177] The calculation of risk parameters corresponding to the loaded early warning rules is performed in parallel. The historical time range required for this calculation is determined based on the acquisition time of the current valid data frame and the time window corresponding to the early warning rule to be executed. Only historical data that has not yet been obtained by the current calculation task within the historical time range is queried, and the incremental calculation of the corresponding risk parameters is performed based on the current valid data frame and historical data.

[0178] When multiple alert rules require historical data within the same time window, the historical data obtained from a single query will be provided to multiple alert rules for joint use.

[0179] When the receiving order of valid data frames is inconsistent with the acquisition time order, the affected sliding window is determined according to the acquisition time of the late valid data frame, and the risk parameters are recalculated for the affected sliding window.

[0180] The unique message identifier corresponding to the valid data frame is used as the idempotency verification basis for the early warning calculation task; when it is detected that the same early warning rule corresponding to the unique message identifier has been calculated, the calculation of the early warning rule is stopped; before writing the early warning event record, the existence of the corresponding early warning event record is verified according to the unique message identifier and the corresponding rule identifier, and the repeated writing is stopped if it already exists.

[0181] The early warning calculation adopts an incremental processing approach for newly added data. After each frame of valid data enters the processing flow, the corresponding construction section is obtained based on the device identifier, and the currently enabled early warning rules for that section are loaded. For each rule, the required historical data is determined based on the current data collection time and the observation time range required by the rule, acquiring only the data within the time window involved in the current calculation; when multiple rules use the same historical time range, the already acquired historical data is shared. For data received later than its actual collection time, the potentially affected sliding window is located based on the data's collection time, and the calculation is recalculated. Duplicate data processing uses a unique data identifier to perform idempotency checks, preventing identical messages from generating duplicate risk calculation results and event records.

[0182] The difference between incremental computation and traditional full scan lies in the fact that the historical data query scope is determined by "the currently added data and specific rules," rather than rereading all records of a certain channel each time. If different rules require the same time range, the time series database is reduced by sharing query results. Late data is handled differently from normal sequential data: normal data is only advanced into the current window, while late data is backtracked to the window affected by its actual collection time for partial recalculation. The unique message identifier participates in both the duplicate detection in the data access phase and the idempotency judgment in the early warning calculation phase, ensuring that message-level duplicates do not become business-level duplicate events.

[0183] As the construction period extends, the historical temperature data for the same channel will continue to increase. If every new data frame scans the entire historical record, the computational load will accumulate with the data volume. By limiting the historical query window, each calculation is mainly related to the time span required by the current rule, rather than directly depending on the total amount of historical data. Sharing historical data among multiple rules further reduces the number of times the same window is queried multiple times. For late data caused by the wireless network, only the windows that are actually affected are recalculated, avoiding a full recalculation. Combined with idempotency verification, normal data, late data, and duplicate data all have definite processing paths, which is beneficial for maintaining the consistency of the early warning calculation process under continuous reporting from multiple devices.

[0184] In some embodiments, temperature time-series data and early warning event records are stored, and the corresponding early warning handling method is determined based on the updated early warning status. When the early warning status enters the alarm status or escalation status, multi-channel push is executed according to the corresponding early warning level, referring to... Figure 7 ,include:

[0185] Step S1121: Write the temperature timing data corresponding to each channel of each device into the timing database.

[0186] Step S1122: Write the section information, equipment information, user information, early warning rule configuration, early warning push configuration, and early warning event records into the relational database, and store the early warning event records in the early warning information table.

[0187] Step S1123: Write the generated report file and operation log to object storage.

[0188] Step S1124: When the warning status is updated to the alarm status, determine the notification object bound to the corresponding segment according to the warning push configuration, generate an SMS notification through the SMS gateway, generate a mobile application notification through the application message push service, and generate an in-platform message in the platform warning information list.

[0189] Step S1125: Set a 30-minute silence period for repeated push notifications of the same early warning event, where the silence period is the time interval at which repeated push notifications of the same early warning event are stopped; limit the number of push notifications of the same early warning event to no more than 10 times; for SMS notifications, limit the number of early warning SMS messages received by the same notification number within a natural day to no more than 20; when the limit is reached, stop sending more early warning SMS messages to the notification number; when the corresponding limit is reached, stop pushing notifications to the notification number; where, the same early warning event refers to an early warning event that continues to exist during the period when the previous alarm state has not yet entered the de-escalation state, targeting the same section, the same channel, or the same measurement point pair and the same early warning rule.

[0190] Step S1126: When the warning status is updated to the upgraded status, add telephone voice notification and notification to superiors, and adjust the silence period to 10 minutes.

[0191] Monitoring data is categorized and stored according to data type. Continuous temperature sampling data is written to a time-series database, while structured information such as segments, equipment, users, warning rules, notification configurations, and warning events are written to a relational database. Report files and related file data are stored independently. When a risk state enters an alarm state, the corresponding notification recipient is obtained based on the event's segment and notification configuration, and an alarm notification is executed via SMS, mobile application messages, and in-platform messages. The same alarm event is considered to have the same event lifecycle before entering the de-escalation state, and a silence time and frequency limit are set for repeated notifications. After an event escalates, telephone voice calls and notifications to superiors are added, while the silence time is shortened.

[0192] The system distinguishes between "early warning events" and "single over-limit calculation results." A persistent risk constitutes the same event from the time it enters the alarm state until it is finally lifted. Therefore, the silence period and the number of push notifications are counted around the event lifecycle, rather than recreating the notification count for each frame of over-limit data received. Temperature time-series data and early warning events are stored using methods suitable for their respective data characteristics. The former is oriented towards continuous queries by time range, while the latter is oriented towards structured queries such as segments, rules, event states, and notification records. The push channel changes according to the alarm level, so that changes in the early warning status directly drive changes in the notification strategy.

[0193] Event-level push control reduces the repeated sending of numerous identical notifications during periods of sustained anomalies, while retaining space for periodic re-reminders. Regular alarms utilize multiple notification paths—SMS, mobile application messages, and platform information—with additional notifications via telephone and superiors when the event persists or reaches escalation criteria, allowing for different notification intensities for different risk levels. Long-term storage of temperature data and event records enables tracing historical temperature changes of corresponding channels based on event occurrence time, and allows for retrospective analysis of concurrent event status using specific temperature curves, providing a continuous data foundation for construction quality analysis, anomaly handling verification, and subsequent project data compilation.

[0194] In some embodiments, after performing tiered multi-channel push notifications, the following steps are also included:

[0195] Based on the received configuration modification information, the acquisition interval, channel engineering name, and early warning threshold are sent to the designated temperature acquisition terminal via the downlink command channel.

[0196] Through a unified HTTPS data interface with a RESTful style, data query, device management, rule configuration, and report export services are provided to web and mobile application clients.

[0197] Based on temperature time-series data, equipment information, and early warning event records, the system generates equipment overview data, historical temperature query results, temperature curves, early warning information lists, and statistical analysis results.

[0198] The equipment overview data includes battery remaining capacity, ambient temperature, and real-time temperature of each channel. The temperature curves include temperature trend curves for the past five days, and historical temperature query results can be combined and queried according to time and equipment conditions.

[0199] Generate the corresponding report file based on the report export request, write the report file to object storage, and provide the report file to the web page or mobile application.

[0200] In addition to performing temperature risk calculations, the cloud platform also synchronizes data collection intervals, channel engineering names, and related configuration parameters to designated field terminals via downlink communication channels, ensuring that the data collection configuration of field devices corresponds to the cloud project configuration. For web and mobile application clients, a unified HTTPS data interface provides temperature data query, device management, rule configuration, and report export services. Based on time-series temperature data, device information, and early warning events, the platform generates an overall device operation overview, historical temperature data, temperature trend curves, early warning information, and statistical results. Upon receiving a report export request, the platform generates and saves the corresponding report file based on the query conditions and then provides the file to the requesting terminal.

[0201] The system connects on-site terminal configuration, cloud-based risk management, and user-end data access within the same data framework. Equipment and channel identifiers determine the origin of raw data, while engineering names and project configurations map physical channels to construction monitoring objects. A unified interface provides query results based on the same contract section, equipment, channel, and time conditions. Configuration modifications are not merely stored in the webpage display layer but are synchronized to designated terminals via downlink channels, establishing a data link between on-site data collection and cloud configuration. Report content is directly derived from pre-saved temperature time-series data, equipment data, and event data, avoiding inconsistencies between statistical results and platform query results caused by using independent data sources.

[0202] Through a unified data access interface, the web and mobile applications use the same backend data and business rules, reducing the possibility of different terminals interpreting the same temperature data differently. Engineering managers can view the current equipment and channel temperatures, query historical data based on time and equipment conditions, and analyze the construction temperature control process by combining temperature trends with concurrent early warning events. Configuration ensures a correspondence between on-site data collection parameters and cloud management parameters, and report export converts existing monitoring results into file data suitable for archiving. Therefore, a continuous data processing path is formed from on-site data collection, cloud analysis, early warning response, historical querying, and report output.

[0203] Another embodiment of the present invention also provides a smart early warning device for monitoring concrete temperature, comprising:

[0204] Create a module to establish a three-level data organization relationship between sections, equipment, and channels, and configure semantic tags for monitoring points, physical relationships between measuring points, and early warning rule templates for each channel.

[0205] The receiving module is used to receive raw data frames uploaded by the temperature acquisition terminal, and to perform device authentication, integrity verification, duplicate data detection and data quality marking on the raw data frames to obtain valid data frames.

[0206] The processing module is used to asynchronously align the valid data frames with the corresponding historical data according to the acquisition time, generate the valid sampling point sequence of each channel, and match the early warning rules and generate measurement point pairs based on the semantic tags of the monitoring points and the physical relationship of the measurement points.

[0207] The calculation module is used to calculate the temperature value risk parameter, the temperature difference risk parameter of the measurement point pair, the temperature change rate risk parameter, and the historical temperature difference risk parameter based on the effective sampling point sequence and measurement point pair, and to determine the over-limit state corresponding to each risk parameter, forming a risk parameter vector.

[0208] The update module is used to update the warning status and generate warning event records based on the risk parameter vector's over-limit status, number of consecutive over-limits, duration, proportion of over-limits within the window, and hysteresis conditions.

[0209] The storage module is used to store temperature time-series data and early warning event records, and determines the corresponding early warning handling method based on the updated early warning status. When the early warning status enters the alarm status or escalation status, it performs multi-channel push according to the corresponding early warning level.

[0210] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0211] Based on the same inventive concept as the above method embodiments, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it enables the electronic device to implement the control method described in the above embodiments.

[0212] In one embodiment, the electronic device may be a server, and in this embodiment, the structure of the electronic device may be as follows: Figure 8 As shown, it includes a memory, a communication module, and one or more processors.

[0213] Memory is used to store computer programs executed by the processor. Memory can be mainly divided into a program storage area and a data storage area. The program storage area can store the operating system and programs required to run instant messaging functions, etc.; the data storage area can store various instant messaging information and operation instruction sets, etc.

[0214] Memory can be volatile memory, such as random access memory (RAM); memory can also be non-volatile memory, such as read-only memory, flash memory, hard disk drive (HDD), or solid-state drive (SSD); or memory can be any other medium capable of carrying or storing a desired computer program having the form of instructions or data structures and accessible by a computer, but is not limited thereto. Memory can be a combination of the above-mentioned types of memory.

[0215] A processor may include one or more central processing units (CPUs) or digital processing units, etc. A processor is used to implement the aforementioned data processing methods when a computer program stored in memory is invoked.

[0216] The communication module is used to communicate with terminal devices and other servers.

[0217] This application embodiment does not limit the specific connection medium between the above-described memory, communication module, and processor. This application embodiment... Figure 8 The memory and processor are connected via a bus, and the bus is in... Figure 8 The connections between other components are illustrated with arrows and are for illustrative purposes only, not as limiting information. Buses can be categorized as address buses, data buses, control buses, etc. For ease of description, Figure 8 The text uses only one arrow to describe it, but does not indicate that there is only one bus or one type of bus.

[0218] Based on the same inventive concept as the above-described method embodiments, embodiments of the present invention also provide a computer-readable storage medium for storing a computer program. When the computer program is run on a computer, it enables an electronic device to implement the control methods described in the above embodiments. The computer-readable storage medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0219] Based on the same inventive concept as the above-described method embodiments, embodiments of the present invention also provide a computer program product. The computer program product includes a computer program that, when run on an electronic device, causes the electronic device to perform the steps of the control methods described above according to various exemplary embodiments of this application. The program product may take the form of any combination of one or more readable media. These computer program commands can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the commands executed by the processor of the computer or other programmable data processing device generate a process for implementing... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0220] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A method for intelligent early warning of concrete temperature monitoring, executed by a cloud server, characterized in that, include: Establish a three-level data organization relationship for sections, equipment, and channels, and configure semantic tags for monitoring points, physical relationships between measuring points, and early warning rule templates for each channel; Receive raw data frames uploaded by the temperature acquisition terminal, and perform device authentication, integrity verification, duplicate data detection and data quality marking on the raw data frames to obtain valid data frames; The effective data frames and corresponding historical data are asynchronously time-aligned according to the acquisition time to generate a sequence of effective sampling points for each channel. The monitoring point semantic tags and measurement point physical relationships are used to match early warning rules and generate measurement point pairs. Based on the effective sampling point sequence and measurement point pairs, the temperature value risk parameter, the measurement point pair temperature difference risk parameter, the temperature change rate risk parameter, and the historical temperature difference risk parameter are calculated respectively, and the over-limit state corresponding to each risk parameter is determined to form a risk parameter vector. The warning status is updated based on the out-of-limit status, consecutive out-of-limit count, duration, out-of-limit percentage within the window, and hysteresis conditions of the risk parameter vector, and a warning event record is generated. The system stores temperature time-series data and the records of the warning events, and determines the corresponding warning handling method based on the updated warning status. When the warning status enters the alarm status or escalation status, it performs multi-channel push according to the corresponding warning level.

2. The intelligent early warning method for concrete temperature monitoring according to claim 1, characterized in that, The establishment of a three-level data organization relationship among sections, equipment, and channels, and the configuration of semantic tags for monitoring points, physical relationships between measuring points, and early warning rule templates for each channel, includes: Establish a correspondence between bidding sections and construction projects, associate at least one piece of equipment under each bidding section, and associate one to eight channels under each piece of equipment; Configure semantic tags for monitoring points for each channel, including pouring zone, point type, burial depth, spatial coordinates, associated points, and structural parts; The pouring zone is used to identify the pouring section or construction zone to which the measuring point belongs; the point type is used to identify whether the measuring point is an internal point, surface point, environmental point or rebar point; the embedment depth is the distance of the measuring point from the concrete surface; the spatial coordinates are the position of the measuring point in three-dimensional space; the associated point is the measuring point identifier that is physically associated with the current measuring point; and the structural part is used to identify the structural component where the measuring point is located. According to the semantic tags of the monitoring points, the corresponding early warning rule templates are matched, wherein the temperature rise rate early warning rule is matched for internal points, the temperature drop rate early warning rule is matched for surface points, the inner and outer surface temperature difference early warning rule is matched for internal and surface points in the same pouring zone, and the same layer temperature difference early warning rule is matched for measuring points in the same pouring zone with a burial depth difference of no more than 0.1 meters. For early warning rules that require two measuring points to participate in the calculation, at least one pair of measuring points is generated based on the associated points, pouring zones, point types, burial depths, and the matched early warning rule template. Each pair of measuring points includes a first measuring point identifier, a second measuring point identifier, and a physical relationship type. An initial threshold of 1.5 degrees Celsius per hour is configured for the internal point temperature rise rate warning rule, an initial threshold of 3.0 degrees Celsius per hour is configured for the surface point temperature drop rate warning rule, an initial threshold of 15 degrees Celsius is configured for the inner-to-outer temperature difference warning rule, and an initial threshold of 10 degrees Celsius is configured for the same-layer temperature difference warning rule; the initial thresholds are used as the rate alarm thresholds for the corresponding point type warning rules when the burial depth adjustment is not performed. Each warning rule includes a rule identifier, its segment, rule name, parameter type, comparison relationship, threshold, rule type, and activation status. The comparison relationship includes greater than, less than, not less than, not greater than, and range.

3. The intelligent early warning method for concrete temperature monitoring according to claim 1, characterized in that, The process of receiving raw data frames uploaded by the temperature acquisition terminal, performing device authentication, integrity verification, duplicate data detection, and data quality marking on the raw data frames to obtain valid data frames includes: The system receives raw data frames containing device identifier, channel identifier, channel engineering name, acquisition time, temperature value, signal quality, data status, frame sequence number, and checksum, and generates a reception time upon receiving the raw data frames. The device identifier is a unique identifier for the temperature acquisition terminal, used for device authentication and segment attribution lookup. The channel identifier is the physical measurement point number of the temperature acquisition terminal. The channel engineering name field records the engineering display name of the channel in the construction monitoring scenario; this engineering display name is independent of the point type used to represent internal points, surface points, environmental points, or rebar points. The acquisition time is a timestamp generated when the temperature acquisition terminal acquires the temperature value. The reception time is a timestamp generated when the cloud server receives the raw data frames. The signal quality is the communication signal strength or signal-to-noise ratio. The data status distinguishes between normal acquisition data, retransmission data, and the first frame data after the temperature acquisition terminal restarts. The frame sequence number is an incrementing sequence number generated by the temperature acquisition terminal. The checksum is an integrity check value generated based on the frame content of the raw data frames. The legitimacy of the corresponding temperature acquisition terminal is verified based on the device identifier, and the integrity of the original data frame is verified based on the check code. When the temperature acquisition terminal fails the legitimacy verification or the original data frame fails the integrity verification, the data quality field of the corresponding data is marked as invalid, and the original data frame is no longer used for subsequent timing alignment and early warning calculations. The combination of the device identifier and the frame sequence number is used as a unique message identifier, and the unique message identifier is queried in the recent message identifier cache with a preset cache duration. When the unique message identifier is found, the original data frame is marked as duplicated, and its use in subsequent early warning calculations is stopped. When the unique message identifier is not found, the unique message identifier is written to the most recent message identifier cache, the data quality field of the original data frame is marked as valid, and the data frame that has completed the unified format conversion is written to the message queue. The data quality field is used to indicate the data quality judgment result of the data frame, and the valid status indicates that the corresponding data frame has passed device authentication, integrity verification, and duplicate data detection.

4. The intelligent early warning method for concrete temperature monitoring according to claim 1, characterized in that, The step of asynchronously aligning the valid data frames with the corresponding historical data according to the acquisition time to generate a sequence of valid sampling points for each channel includes: Construct a time window with the acquisition time of the current valid data frame as the right endpoint. In the formula, The acquisition time of the current valid data frame. The length of the sliding window is measured in minutes. Obtain the set of sampling points arranged in ascending order of acquisition time within the time window: In the formula, The number of sampling points within the time window. For the first The sampling time for each sampling point For the first Temperature values ​​at each sampling point The sampling point number, and ; Calculate the time interval between adjacent sampling points: In the formula, For the first The sampling point and the first The time interval between sampling points For the first The sampling time for each sampling point For the first The collection time for each sampling point; When the time interval between any two adjacent sampling points is less than the preset minimum time interval, delete the sampling point whose temperature value is closer to the median temperature value in the current time window, and retain the sampling point whose temperature value deviates significantly from the median. Rearrange the remaining sampling points in ascending order of collection time and recalculate the time interval between adjacent sampling points. Repeat the filtering process until the time interval between any two adjacent valid sampling points is not less than the preset minimum time interval, and obtain a valid sampling point sequence. When the time difference between the acquisition time of the first valid sampling point and the left endpoint of the time window exceeds twice the minimum time interval, the current time window is determined to have insufficient data coverage, and the calculation of the temperature rise rate or temperature fall rate is stopped using the current time window. For each channel participating in the temperature difference calculation of the measuring point pair, select an effective sampling point within the corresponding time window whose acquisition time is no later than the right end of the window and whose time difference with the right end of the window is the smallest. Combine the temperature values ​​of the effective sampling points corresponding to different channels with the temperature value pairs corresponding to the measuring point pair. The effective sampling point sequence is used for temperature change rate calculation only when there are at least two effective sampling points and the first and last effective sampling points have different acquisition times.

5. The intelligent early warning method for concrete temperature monitoring according to claim 4, characterized in that, Calculating the rate of temperature rise or fall includes: The temperature rise rate or temperature fall rate is calculated using the endpoint difference method or the least squares fitting slope method. The temperature change rate calculation is performed only on valid sampling point sequences with at least two valid sampling points and different sampling times. The sampling time of each valid sampling point is converted into a time value in minutes according to a unified time base before being included in the calculation. When using the endpoint difference method, the following calculation is performed based on the first and last valid sampling points in the valid sampling point sequence: In the formula, The rate of temperature rise or fall is expressed in degrees Celsius per hour. The temperature value of the first valid sampling point. This is the temperature value of the last valid sampling point. The time of collection for the first valid sampling point. The time of the last valid sampling point. The time difference is in minutes and is not zero. The value 60 is used to convert the temperature change rate in minutes to the temperature change rate in hours. When using the least squares fitting slope method, a univariate linear regression is performed on the effective sampling point sequence: In the formula, To fit the temperature value, The data collection time is measured in minutes. The slope of the fitted line, The intercept of the fitted line; The slope of the fitted straight line is calculated according to the following formula: In the formula, The slope of the fitted line, The number of effective sampling points, The valid sampling point number, For the first The collection time for each valid sampling point For the first Temperature values ​​of each valid sampling point; Calculate the rate of temperature rise or rate of temperature fall based on the slope of the fitted straight line: In the formula, The rate of temperature rise or fall is expressed in degrees Celsius per hour. The slope of the fitted straight line is expressed in degrees Celsius per minute. The value 60 is used to convert the slope of the fitted straight line in minutes to the rate of temperature rise or fall in hours. in, This indicates a rise in temperature. This indicates a drop in temperature.

6. The intelligent early warning method for concrete temperature monitoring according to claim 2, characterized in that, Based on the effective sampling point sequence and measurement point pairs, the risk parameters for temperature values, temperature differences between measurement point pairs, rate of temperature change, and historical temperature differences are calculated respectively, and the out-of-limit states corresponding to each risk parameter are determined to form a risk parameter vector, including: The current temperature value of each channel is compared with the corresponding upper or lower temperature threshold to obtain the temperature risk parameter and the corresponding over-limit status. Based on the time-aligned temperature values ​​corresponding to the first and second measuring points of the measuring point pair, the absolute value of the temperature difference between the first and second measuring points is calculated, and the corresponding maximum allowable temperature difference is determined according to the physical relationship type of the measuring point pair. The absolute value of the temperature difference is compared with the maximum allowable temperature difference to obtain the temperature difference risk parameter of the measuring point pair and the corresponding over-limit state. Calculate the absolute value of the temperature difference between the current acquisition time and the historical acquisition time corresponding to the preset historical time interval for the same channel. Compare the absolute value of the temperature difference with the maximum allowable historical temperature difference to obtain the historical temperature difference risk parameter and the corresponding over-limit status. When there is no valid temperature data for the historical acquisition time corresponding to the preset historical time interval, no historical temperature difference calculation value is generated, and the corresponding historical temperature difference risk parameter is marked as insufficient data. The insufficient data status is not treated as an over-limit status. The temperature rise rate is compared with the temperature rise rate alarm threshold, and the absolute value of the temperature fall rate is compared with the temperature fall rate alarm threshold to obtain the temperature change rate risk parameter and the corresponding over-limit status. When adjusting the rate threshold based on the burial depth, the adjusted rate threshold is calculated using the following formula: In the formula, The rate threshold is adjusted based on the burial depth. The base rate alarm threshold is determined based on the current measurement point type and corresponding early warning rules. The depth of the measuring point is measured in meters. When the risk parameter of any channel is out of limit, the risk parameter corresponding to the associated point of the channel is obtained; when the associated point shows a same-direction anomaly, the risk of the current channel is marked as a verified risk; when the associated point does not show a same-direction anomaly, the risk of the current channel is marked as an isolated risk and the push priority of the corresponding warning event is reduced; wherein, the same-direction anomaly means that the current channel and the associated point are both out of limit under the same risk parameter type, and for risk parameters with temperature change direction, the temperature change direction corresponding to the current channel and the associated point is the same; The calculated values ​​of each risk parameter, the over-limit status, and the risk label are combined to form the risk parameter vector.

7. The intelligent early warning method for concrete temperature monitoring according to claim 1, characterized in that, The step of updating the warning status based on the risk parameter vector's out-of-limit status, consecutive out-of-limit count, duration, out-of-limit percentage within the window, and retracement conditions, and generating a warning event record, includes: Upon receiving a new valid data frame, the system generates the next frame sequentially, frame by frame, with the acquisition time of the new valid data frame as the right endpoint of the window. A sliding window and calculate the first The rate of temperature rise or temperature fall of each sliding window, where, For the first A sliding window, The sliding window number is the overlap duration of adjacent sliding windows, which is the difference between the sliding window length and the sampling interval. The proportion of exceeding limits within the window is defined as the ratio of the number of valid calculation results in the current judgment window to the total number of valid calculation results within the judgment window for the same early warning rule. In the scenario of temperature rise, the following formula is used to determine the first... Are the number of sliding windows exceeded? In the formula, For the first The rate of temperature rise of each sliding window, The alarm threshold for the rate of temperature rise. The sliding window number; In a temperature drop scenario, the following formula is used to determine the first... Are the number of sliding windows exceeded? In the formula, For the first The rate of temperature drop of each sliding window, The alarm threshold for the rate of temperature drop. The sliding window number; When the current sliding window exceeds the limit, increment the consecutive window exceeding the limit count; when the current sliding window does not exceed the limit, reset the consecutive window exceeding the limit count. Only when the consecutive out-of-limit window count reaches the consecutive out-of-limit window count threshold. When the corresponding temperature rise rate warning or temperature fall rate warning is triggered, the following will occur: The minimum number of consecutive out-of-limit windows required to trigger a temperature rise rate warning or temperature fall rate warning; the default value is 3. The warning status is updated according to the normal status, observation status, alarm status, escalation status and deactivation status; When a single anomaly occurs, the normal state is updated to the observation state and no warning is issued. For temperature rise rate risk and temperature fall rate risk, when the continuous over-limit window count reaches the continuous over-limit window count threshold N, the observation state is updated to the alarm state. For temperature value risk, measuring point pair temperature difference risk, and historical temperature difference risk, when any of the following conditions are met: the number of consecutive over-limits reaches N, the duration of over-limits reaches 10 minutes, or the proportion of valid over-limit calculation results within the judgment window to the total number of valid calculation results reaches 70%, the observation state is updated to the alarm state. When the corresponding anomaly is not continuous and the conditions for entering the alarm state are not met, the observation state is updated to the normal state. When the number of consecutive normal occurrences reaches 3 and the duration of normal occurrence reaches 15 minutes, the hysteresis condition is further determined. For warning rules that use exceeding the upper threshold as the trigger condition, the hysteresis condition is met when the risk parameter is lower than the difference between the corresponding upper threshold and the hysteresis. For warning rules that use being lower than the lower threshold as the trigger condition, the hysteresis condition is met when the risk parameter is higher than the sum of the corresponding lower threshold and the hysteresis. The hysteresis is five percent of the absolute value of the corresponding trigger threshold. When the number of consecutive normal occurrences, the duration of normal occurrence, and the hysteresis condition are all met simultaneously, the alarm status is updated to the deactivated status. When any of the following conditions are met: the deviation of a temperature-related risk parameter (in degrees Celsius) from the trigger threshold is not less than 5 degrees Celsius; the duration of exceeding the limit is not less than 2 hours; or multiple channels or multiple warning rules are triggered simultaneously, the alarm status is updated to an escalation status. When the corresponding risk parameter falls back or a manual confirmation is received, the escalation status is updated to an alarm status. The duration of exceeding the limit is the time between the acquisition time corresponding to the first valid calculation result of exceeding the limit in the current continuous exceeding the limit process and the acquisition time corresponding to the current valid calculation result of exceeding the limit. Generate a record of early warning events, including the relevant section, corresponding channel or measurement point pair, early warning level, triggering conditions, early warning time, and push action.

8. The intelligent early warning method for concrete temperature monitoring according to claim 7, characterized in that, The process of updating the warning status and generating warning event records based on risk parameters also includes: Upon receiving each valid data frame, the corresponding segment to which the device belongs is queried based on the device identifier in the valid data frame, and the warning rules that are in the enabled state under the segment are loaded. The risk parameter calculation corresponding to the loaded early warning rule is performed in parallel. The historical time range required for this calculation is determined based on the acquisition time of the current valid data frame and the time window corresponding to the early warning rule to be executed. Only historical data that has not yet been obtained by the current calculation task within the historical time range is queried, and the incremental calculation of the corresponding risk parameter is performed based on the current valid data frame and the historical data. When multiple early warning rules require historical data within the same time window, the historical data obtained from a single query will be provided to the multiple early warning rules for joint use. When the receiving order of valid data frames is inconsistent with the acquisition time order, the affected sliding window is determined according to the acquisition time of the late valid data frame, and the risk parameters are recalculated for the affected sliding window. The unique message identifier corresponding to the valid data frame is used as the idempotency verification basis for the early warning calculation task; when it is detected that the same early warning rule corresponding to the unique message identifier has been calculated, the calculation of the early warning rule is stopped; before writing the early warning event record, the existence of the corresponding early warning event record is verified according to the unique message identifier and the corresponding rule identifier, and the repeated writing is stopped if the record already exists.

9. The intelligent early warning method for concrete temperature monitoring according to claim 1, characterized in that, The system stores the time-series temperature data and the early warning event records, and determines the corresponding early warning handling method based on the updated early warning status. When the early warning status enters the alarm state or escalation state, it performs multi-channel push according to the corresponding early warning level, including: Write the temperature timing data corresponding to each channel of each device into the timing database; The section information, equipment information, user information, early warning rule configuration, early warning push configuration, and early warning event records are written into a relational database, and the early warning event records are stored in an early warning information table; Write the generated report files and operation logs to object storage; When the warning status is updated to the alarm status, the notification object bound to the corresponding section is determined according to the warning push configuration. An SMS notification is generated through the SMS gateway, a mobile application notification is generated through the application message push service, and an in-platform message is generated in the platform warning information list. A 30-minute silence period is set for repeated push notifications of the same early warning event, where the silence period is the time interval at which repeated push notifications of the same early warning event are stopped. The number of push notifications for the same early warning event is limited to no more than 10 times. For SMS notifications, the number of early warning SMS messages received by the same notification number within a natural day is limited to no more than 20. When the limit is reached, the sending of early warning SMS messages to the notification number is suspended. When the corresponding limit is reached, the push notification to the notification number is suspended. The same early warning event refers to an early warning event that persists for the same section, the same channel, the same measuring point pair, and the same early warning rule, while the previous alarm state has not yet entered the de-escalation state. When the warning status is updated to an upgraded status, telephone voice notification and notification to superiors will be added, and the silent period will be adjusted to 10 minutes.

10. The intelligent early warning method for concrete temperature monitoring according to claim 1, characterized in that, After implementing tiered, multi-channel push notifications, it also includes: Based on the received configuration modification information, the acquisition interval, channel engineering name, and early warning threshold are sent to the designated temperature acquisition terminal via the downlink command channel; Through a unified HTTPS data interface with a RESTful style, data query, device management, rule configuration, and report export services are provided to web and mobile application clients. Based on temperature time-series data, equipment information, and early warning event records, the system generates equipment overview data, historical temperature query results, temperature curves, early warning information lists, and statistical analysis results. The equipment overview data includes battery remaining capacity, ambient temperature, and real-time temperature of each channel. The temperature curve includes a temperature trend curve for the past five days. The historical temperature query results are queried in combination according to time conditions and equipment conditions. Generate the corresponding report file according to the report export request, write the report file to object storage, and provide the report file to the web page or mobile application.

Images