A heat supply long-distance pipeline network monitoring system and method

Abstract

The present application relates to pipe network engineering management method, specifically relates to heat supply long-distance pipe network monitoring system, including cloud server and the monitoring terminal that sends data to cloud server through low power wide area network, the monitoring terminal obtains the compression sequence of real-time water pressure data in the first time window with on-chip time as the standard, the first sequence of fixed length is intercepted from back to front in the compression sequence, the cloud server receives the first sequence and looks for the first sequence of two monitoring terminals that match. Through looking for the first sequence that matches and the difference sequence after alignment, the demonstration of one or more physical quantities such as the water pressure of corresponding monitoring node, the water pressure difference of monitoring node, the water pressure drop of pipeline between monitoring nodes is obtained.

Description

Technical Field

[0001] This disclosure belongs to the field of pipeline system technology, and relates to pipeline network engineering management methods, particularly to a monitoring system and method for long-distance heating pipeline networks. Background Technology

[0002] In the intelligent transformation of existing heating networks, especially primary networks, heating companies face technical challenges when deploying IoT terminals. These challenges stem from a mismatch between existing infrastructure and modern monitoring requirements. The existing heating networks involved in the transformation include a large number of primary networks and a small number of secondary networks, mostly directly buried or embedded in walls, without pre-reserved sensor installation compartments. To control transformation costs and minimize disruption to residents' lives, feasible deployment locations for new IoT monitoring terminals primarily include existing valve wells, inspection wells, and heating chambers. These locations are mainly used for valve operation, pipeline maintenance, and venting. Modern IoT monitoring systems require high accuracy in real-time sensing of network operating status, such as pressure. Therefore, suitable monitoring node locations for installing IoT monitoring terminals need to be carefully selected.

[0003] The primary heating network is mainly a long-distance pipeline network, responsible for the transmission of heat energy from the heat source to the heating station. Some renovation plans choose to use monitoring terminals connected to low-power wide area networks (LPWAN) to deploy the sensing layer network to adapt to the deployment conditions of the old network renovation. However, on the one hand, due to the requirements of refined heating management, both primary and secondary networks involve frequent adjustments of booster pumps or control valves, which can easily cause random fluctuations in network pressure. On the other hand, due to the low-power design, the on-chip clock of such monitoring terminals is unstable, and they are in standby mode most of the time. The communication cycle is long, the data transmission time is inaccurate, and the data reception time is not synchronized, making it difficult to provide pressure data to meet the needs of subsequent real-time fault analysis and prediction.

[0004] US Patent 12091842B2 discloses a method for monitoring pipeline anomalies based on transient features and artificial neural network analysis. It tends to monitor pipeline anomalies by acquiring transient pressure data at various points over a long period, but it does not disclose its specific data acquisition method, and its sample data volume is clearly unsuitable for currently available LPWAN technologies. Chinese Patent CN113465019A discloses a system and method for monitoring abnormal water loss in heating pipe networks. It involves comparing the maximum and minimum data points of the pressure characteristic signal sequence at the monitoring front end with a benchmark value as the basis for anomaly alarms. However, its monitoring method still focuses on transient analysis at specific nodes, making it impractical for low-power terminal communication sensing conditions. Summary of the Invention

[0005] This disclosure aims to provide a monitoring system and method for long-distance heating pipelines deployed on low-cost LPWANs, which can significantly improve the accuracy of pipe pressure sensing in primary heating pipelines to meet the data needs of subsequent pipeline fault analysis and prediction.

[0006] In existing technologies for low-power, low-cost IoT terminal devices, when it is necessary to calculate the difference between physical quantities from different sensors, to avoid errors caused by the processing of the difference physical quantities, the sampling time synchronization is generally ensured by sharing an on-chip clock and merging physical communication links and packaging them together (such as the CoAP message protocol of NB-IoT). This method is affected by the protocol, has low data volume, and cannot meet the requirements of long-distance sampling. Even when deployed in the same location (such as the supply and return water pressure difference of the same inspection well), the on-chip resource constraints prevent the use of flexible filtering algorithms or complex conversions. Alternatively, a high-precision communication time synchronization protocol can be used to synchronize the on-chip clock before sampling and transmission, but remote time synchronization requires frequent calls to the on-chip communication module, which seriously shortens the terminal standby time. Due to the above technical limitations, actual engineering sensing and monitoring still mainly relies on point-by-point sampling by each terminal. Because the data of each monitoring node lacks synchronization, it is difficult to provide effective basic data to analysis systems such as large-scale fault analysis models when the accuracy requirements of fault analysis increase. The first aspect of this disclosure provides a technical solution through multiple system embodiments: a long-distance heating pipeline monitoring system, belonging to the Internet of Things (IoT) system deployed based on Low Power Wide Area Network (LPWAN). These systems all include at least a cloud server located at the platform layer or application layer, and a monitoring terminal that sends field sensing data to the system. The monitoring terminal and the cloud server are data coupled through the system's communication network, and the coupling link includes the LPWAN; that is, the monitoring terminal directly or indirectly sends field sensing data to these cloud servers. Wherein:

[0007] Some of these monitoring terminals deployed in the sensing layer are deployed in the water supply and return pipelines of various monitoring nodes in the long-distance heating pipeline network to sense relevant water pressure data of the monitoring nodes. They are configured to: continuously collect multiple real-time water pressure data at equal intervals within a first time window; obtain a compressed sequence of the real-time water pressure data within the first time window; extract a first sequence of fixed length from the compressed sequence from the end to the beginning; and send the first sequence to the cloud server at a configured sending time.

[0008] Some cloud servers deployed at the platform layer or application layer of the system, coupled with the data from the aforementioned monitoring terminals, are configured to: receive the first sequence, search for matching sequences of two monitoring terminals, where the matching relationship satisfies that the reception times of the first sequences are both within a second time window, and the monitoring terminals are located on the same monitoring node's water supply and return pipelines, or adjacent upstream and downstream monitoring nodes on the same pipeline. The search by the cloud servers can be either cyclically triggered or passively responded to based on requests.

[0009] In this system, the end time of the first time window and the transmission time have a relatively fixed on-chip clock interval. These on-chip clock intervals are used to control the difficulty of aligning the data sequences of subsequent matching monitoring terminals and to prepare for the on-chip processing unit to complete the transmission of compressed data. On some monitoring terminals using real-time streaming compression, since compression is already completed upon reception, no additional compression processing is needed, and this time can be very short, controlled to be greater than a transmission success probability. In other possible solutions, it is necessary to separately compress the first time window portion or all of the collected data in the buffer, such as in some dynamically adaptive compression algorithms. This time should be the minimum time to ensure compression completion. Simultaneously, the length of the second time window is less than the interval between two adjacent transmission times configured by the monitoring terminal to avoid the same monitoring terminal receiving invalid aligned data when the accumulated time error is too large.

[0010] In other embodiments of the first aspect, the monitoring terminal deployed by the system is also used to send the water temperature of its deployed pipeline corresponding to a first time window to the cloud server. Since water temperature data generally has a slow rate of change, these water temperatures can be obtained directly by the monitoring terminal at a single point in the first time window from the sensor of the sampling channel, or they can be obtained through multiple acquisitions within the first time window to assess the measurement error of the sample water temperature. In some specific examples, this water temperature data can be used by the cloud server to calculate the relative static head, or static pressure difference, caused by the water temperature difference between the two ends of the pipe section, unaffected by flow velocity.

[0011] In other embodiments of the first aspect, when two matched monitoring terminals are located at adjacent monitoring nodes on the same pipeline, the cloud server is further configured to locate the static pressure difference of the pipeline segment between the adjacent monitoring nodes, the static pressure difference including the static pressure difference caused by the height difference between adjacent monitoring nodes. In some specific implementations where the monitoring terminal is also configured to send the water temperature of its deployed pipeline corresponding to a first time window to the cloud server, the static pressure difference obviously also includes the static pressure difference caused by the water temperature difference at both ends of the pipeline segment.

[0012] In other embodiments of the first aspect, the cloud server is further configured to decompress the first sequence of the two matched monitoring terminals and perform time series alignment to obtain a first difference sequence of the aligned portion.

[0013] In other embodiments of the first aspect, the cloud server is further configured to filter the first sequence received from the monitoring terminal to generate a second sequence.

[0014] In some possible improved embodiments, in conjunction with the preferred embodiment of the first aspect described above, the second difference sequence of the aligned portion is obtained by decompressing the second sequences of the two matched monitoring terminals and aligning them according to time series.

[0015] In some possible improved embodiments, in conjunction with the preferred embodiment of the first aspect described above, the difference sequence is the dynamic pressure difference after deducting the static pressure head of the corresponding pipeline.

[0016] In conjunction with the preferred embodiment of the first aspect described above, some possible improved embodiments of the compression method for obtaining the compressed sequence of real-time water pressure data within the first time window is a linear fitting lossy compression algorithm with a maximum error condition.

[0017] The second aspect of this disclosure provides a method for monitoring long-distance heating pipelines by combining technical solutions provided in multiple system embodiments, for application in these long-distance heating pipeline monitoring systems. Some specific implementations of the method include the step of: when the first sequence of two matching monitoring terminals returns empty or the duration of the aligned portion obtained through the two returned first sequences is less than a threshold, adjusting the on-chip clocks of the two matching monitoring terminals so that their respective first time windows are substantially the same standard time. The system embodiments of this disclosure focus on disclosing a new long-distance heating pipeline monitoring system that enables the acquisition of the true difference value with minimal time synchronization. Although the system tolerance is extremely large, it cannot prevent the significant drift of the on-chip time of monitoring terminals operating in long-term dormant mode. The second aspect of these method examples provides technical solutions for addressing the timing and target of this significant drift.

[0018] This disclosure, in its third aspect, provides a method for monitoring long-distance heating pipelines by combining technical solutions from multiple system embodiments, for application to these long-distance heating pipeline monitoring systems. Clearly, these application methods based on the new long-distance heating pipeline monitoring systems can bring direct technical benefits. Some specific implementations include the step of: calculating and / or updating the water pressure readings and / or water pressure difference readings of the two monitoring terminals when a first sequence of two matching monitoring terminals returns non-empty. In some direct implementations, a cloud server provides users with readings of one or more physical quantities, including the water pressure at the monitoring nodes, the supply and return water pressure difference at the monitoring nodes, and the water pressure difference in the pipelines between the monitoring nodes, through a human-computer interaction terminal. It is easy to understand that the processing of these readings is not limited to cloud-based data servers or application servers. The calculation of these readings can be completed by local applications. For example, in some applications, the human-machine interface terminal directly obtains the first or second sequence of two matching monitoring terminals on the corresponding pipe section from the cloud and stores it locally. In some application interfaces, the pressure or differential pressure reading at a certain point in time can be calculated locally without being connected to the network. The system itself may not contain these devices and applications that are coupled through API data, but these application methods should be included in the protection scope of this disclosure.

[0019] While those skilled in the art can directly understand the significant technical effects of the various aspects of the technical solutions disclosed herein from the description of multiple aspects, it is still necessary to explain that the effects obtained by this disclosure in solving the following specific technical problems in the art are more prominent. For long-distance pipeline monitoring systems, especially the pressure monitoring part, in response to the difficulty in unifying the pressure data of multiple devices in time and space, such as different upload frequencies and timestamps of different devices, and the fact that the data are distributed in different geographical locations, making direct comparison and joint analysis of the data impossible, some examples of this disclosure adopt spatiotemporal matching and secondary processing technology for multi-point pressure data. In response to the problem of inconsistent data upload times of different devices, a time-window-based time-series alignment and resampling interpolation processing method is introduced to perform unified time-scale matching of multi-source pressure data. At the same time, combined with the geographical topology of the pipeline network, discrete pressure monitoring points are mapped into a "node-pipe segment" model to realize the spatial organization and association of multi-point pressure data. Building upon this foundation, the traditional single-point pressure display has been upgraded to a unified presentation of the following combined calculation results: supply water pressure, return water pressure, and supply-return water pressure difference at the same location; supply water pressure drop and return water pressure drop between different locations (e.g., from point A to point B, point A to point C, point B to point D); and comparative analysis of pressure change characteristics across multiple paths and pipe sections. Addressing the issue of insufficient accuracy in calculating network pressure difference and pressure drop, such as the failure to effectively eliminate the static pressure influence caused by terrain differences due to varying equipment installation heights, resulting in pressure difference values ​​that do not accurately reflect the hydraulic state of the network, this disclosure introduces static pressure correction methods for calculating pressure difference and pressure drop. This solves the static pressure interference caused by differences in equipment installation heights. By introducing a static pressure correction model into the calculation process and combining it with the geographical elevation information of the equipment, the static pressure component caused by terrain undulations is calculated and eliminated, thereby obtaining more effective pressure difference and pressure drop results that better reflect actual hydraulic conditions. This effectively avoids systematic errors caused by geographical differences, making the calculation results more reliable and comparable in engineering applications. To address the reliability issues of single-point pressure alarm modes, considering the susceptibility of single-device pressure to user behavior and instantaneous disturbances, the difficulty in setting reasonable alarm thresholds, and the resulting high false alarm rates and delayed anomaly detection, this disclosure presents several examples of intelligent alarm methods based on inter-location pressure drop characteristics. For instance, in terms of alarm strategy, unlike traditional alarm modes based on single-point pressure values, these methods can detect anomalies based on inter-location pressure drop characteristics. By monitoring pipe sections or routes, the changes in supply and return water pressure drops are analyzed in real time. Since pressure drop exhibits strong stability under normal operating conditions, these methods can significantly narrow the range of alarm upper and lower thresholds. When pipeline leaks, partial blockages, or hydraulic anomalies occur, the pressure drop index will show abnormal fluctuations first, facilitating rapid identification and accurate location of pipeline network problems. Attached Figure Description

[0020] For those skilled in the art, the clear description of various embodiments in this disclosure is sufficient for them to understand the scope of the technical solutions claimed in this disclosure. The accompanying drawings described below are merely some exemplary specific embodiments and technical aspects of this disclosure. Other drawings can be obtained based on these drawings without any creative effort. The following is a brief introduction to the drawings used in the description of the specific embodiments. Obviously, since each drawing only describes one technical aspect, when its description is used in conjunction with other drawings or technical aspects to explain multiple aspects of the implementation in different specific embodiments, the content shown in the drawings has a distinguishing scope of reference when understood in context.

[0021] Figure 1 This is a schematic diagram of the structure of a long-distance heating pipeline monitoring system according to one embodiment of the present disclosure;

[0022] Figure 2 This is a schematic diagram showing the deployment location of the monitoring terminal at the monitoring node in one embodiment of this disclosure;

[0023] Figure 3 A timing diagram showing the transmission of the corresponding first sequence by the matching monitoring terminal in one embodiment of this disclosure;

[0024] Figure 4 This is a schematic diagram showing the deployment location of the monitoring terminal at the monitoring node in another embodiment of this disclosure;

[0025] Figure 5 This is a schematic diagram illustrating the acquisition of the first sequence alignment portion using a time alignment algorithm in one embodiment of this disclosure;

[0026] Figure 6 This is a schematic diagram illustrating the acquisition of the first sequence alignment portion through a time alignment algorithm in another embodiment of this disclosure;

[0027] Figure 7 This is a schematic diagram of the hardware platform structure of the monitoring terminal in one embodiment of this disclosure. Detailed Implementation

[0028] First, it should be noted that phrases such as "in one embodiment" or "in an embodiment" in this specification do not necessarily refer to the same embodiment, but rather provide specific technical aspects for combining particular embodiments, wherein specific features, structures, or characteristics can be combined in any suitable manner consistent with this disclosure. The terms "comprising" and "including" are open-ended, as used in the claims, and do not exclude additional structures or steps. Consider the following cited claim: "A system comprising one or more cloud servers..." Such claims do not exclude the inclusion of additional devices (e.g., gateway devices, communication load balancing devices, etc.). Various terminals, servers, or other devices may be described or stated as being "configured" to perform one or more tasks or task steps. In such a context, "configured" implies a structure (e.g., circuitry) that indicates the terminal / server / computer device includes a structure (e.g., circuitry) that performs the one or more tasks and task steps by its processing unit during operation. Thus, the terminal / server / computer device is allegedly configured to perform the task or specific steps within the task even when the specified terminal / server / computer device is currently inoperable or not running (e.g., not connected). Terminal / server / computer devices used with the language “configured as” include hardware—such as circuits, memory storing executable program instructions to perform operations, etc. Furthermore, “configured as” can include general-purpose structures (e.g., general-purpose circuits) manipulated by software or firmware (e.g., FPGAs or general-purpose processors executing software) to operate in a manner capable of performing one or more tasks to be solved. “Configured as” can also include adjusting manufacturing processes (e.g., semiconductor fabrication facilities) to manufacture devices (e.g., integrated circuits) suitable for implementing or performing one or more tasks. As used herein, indicative terms such as “first” and “second” act as labels for the nouns preceding them and do not imply any type of ordering (e.g., spatial, temporal, logical, etc.). For example, a terminal / server or the task and task-specific characteristics of a terminal / server configured to perform may be described herein as the execution of a “first” task / step / algorithm and a “second” task / step / algorithm. The terms “first” and “second” do not necessarily imply that the second algorithm must be executed before the first algorithm. As used herein, the terms “based on,” “according to,” or “depending on” are used to describe one or more factors influencing a determination, and these terms do not exclude additional factors that may influence the determination. That is, the determination may be based solely on these factors or at least in part on these factors. Consider the phrase “A is determined based on B,” in which case B is a factor influencing the determination of A. Such phrases do not exclude that the determination of A may also be based on C, and in other instances, A may be determined solely on B. When used in the claims, the term “or” is used as an inclusive or, not an exclusive, or.For example, the phrase "at least one of x, y, or z" means any one of x, y, and z, and any combination thereof.

[0029] In one aspect, the specific implementations provided in this disclosure are mainly applied to long-distance heating pipeline monitoring systems that use low-power wide-area networks to sense field data. Figure 1 This paper presents a typical four-layer IoT architecture for such systems, which mainly includes the sensing layer, network layer, platform layer, and application layer. The data link of the network layer is built based on low-power wide-area network, and its construction technology includes, but is not limited to, NB-IoT, LoRa, etc.

[0030] according to Figure 1 , 2 As shown in Figure 3, in a specific demonstration using an NB-IoT network, the sensing layer mainly consists of monitoring terminals E1, E2, E3, E4, etc., deployed at monitoring nodes along the pipeline network. These terminals generally integrate core sensing units such as pressure sensors and temperature sensors, as well as NB-IoT communication modules and microcontrollers (MCUs). They also adopt on-site power supply methods. For example, in some specific implementations in the field or underground environments without mains power, the terminals use solar panels + batteries or single batteries for power supply.

[0031] This implementation uses a low-power wide-area network (LPWAN) at the network layer based on the NB-IoT cellular network provided by the telecommunications operator. Data collected by the monitoring terminal is directly transmitted to a nearby network layer base station via the NB-IoT module, and then transmitted to the cloud platform via core network routing. In this implementation, the monitoring terminal is installed in valve wells along urban long-distance pipelines, without a self-built dedicated communication network, resulting in lower deployment costs and complexity. In other implementations involving special enclosed environments, private base stations or relays (wireless gateways or DTUs, etc.) are deployed. When considering signal coverage, other solutions such as LoRa are used to construct the communication network.

[0032] This implementation utilizes a cloud-based IoT platform to receive, store, and manage data from numerous monitoring terminals. It provides fundamental data services for device management, status monitoring, and data visualization. Due to distributed storage design and communication load balancing requirements, it comprises multiple physical or virtual cloud servers, such as PS1 and PS2, to form the platform layer system. In some specific implementations, the cloud-based data server in the platform layer system only provides data services to upper-layer business logic processing applications through API interfaces, ensuring that the business logic processing applications and monitoring terminals are only coupled through data.

[0033] This implementation's application layer is a monitoring system for operations and maintenance personnel, including PC and mobile human-machine interface terminals T1, T2, etc., primarily used for configuring business logic or retrieving monitoring data, and providing online services deployed in the cloud via application servers AS1, AS2, etc. The core functions of the monitoring system generally include: data overview and GIS display, i.e., displaying the location, real-time data (pressure, temperature), and alarm status of all monitoring terminals on an electronic map; multi-dimensional query and statistics, supporting filtering historical data by device, time, parameter thresholds, and other conditions, and displaying statistical results in the form of graphs; intelligent early warning function, supporting the platform to set normal threshold ranges for parameters such as pressure, differential pressure, and temperature, and automatically generating alarms and pushing them to the human-machine interface terminals when monitoring data is abnormal (such as a sudden drop in differential pressure in a pipeline section, or an abnormal increase in differential pressure), indicating potential risks such as leakage or pressure loss; and on-site auxiliary functions, where the monitoring system provides equipment navigation, on-site photo uploading, and simple equipment debugging (such as parameter reading via Bluetooth connection) through the human-machine interface terminals to assist on-site inspections and maintenance. It is easy to understand that cloud servers for processing relevant data can be deployed at both the platform layer and the application layer. The main difference lies in whether a single communication protocol is used to directly or indirectly connect with the monitoring terminal, such as a network connection at the network layer in public technologies like LPWAN.

[0034] like Figure 2 As shown, monitoring nodes P1, P2...P11 are located near the junction of the water supply and return pipelines, in upstream pipelines, at water supply inlets / outlets, and in the middle of long straight pipes—locations where installation is feasible. The monitoring nodes are connected to both the supply and return pipelines. Hollow arrows in the diagram indicate the heating direction, while solid arrows indicate the actual water supply and return directions in the pipelines. Monitoring terminals are deployed at each monitoring node, with one terminal deployed on both the supply and return pipelines of each node. Specifically, monitoring terminals near the junction of the pipelines are deployed on the upstream main pipeline in the heating direction of that monitoring node. Specifically, monitoring terminals E1 and E2 are deployed on the water supply and return pipelines of monitoring node P2, respectively; monitoring terminals E3 and E4 are deployed on the water supply and return pipelines of the upstream main pipeline in the heating direction of monitoring node P3, respectively; monitoring terminals E5 and E6 are deployed on the water supply and return pipelines of monitoring node P4, respectively; and monitoring terminals E7 and E8 are deployed on the water supply and return pipelines of monitoring node P6, respectively. Clearly, monitoring terminals E1, E3, E5, and E7 are deployed on the water supply pipelines, and E1 and E3, E3 and E5, and E3 and E7 are monitoring terminals of adjacent upstream and downstream monitoring nodes located on the same pipeline. Monitoring terminals E1 and E2, E3 and E4, and E5 and E6 are monitoring terminals located on the same monitoring node.

[0035] In some specific implementations, the monitoring terminal's data acquisition port adopts an integrated temperature and pressure acquisition design. The acquisition port faces the side of the pipe to avoid sedimentation and contact with air bubbles inside the pipe. The slope of the connecting pressure guide pipe is greater than 1:10 to prevent liquid accumulation, thereby obtaining more stable and direct pressure samples. The monitoring terminal's communication module is led to an external antenna of the monitoring node via a feeder to increase the communication success rate. The distance between adjacent monitoring nodes is determined by the upper and lower limits of the pressure wave velocity measured on-site to meet the constraints of monitoring requirements. In a demonstration where water hammer waves caused by filter pipe adjustments are required, the constraints on the engineering estimation of the distance between upstream and downstream monitoring nodes are as follows:

[0036]

[0037] in, The distance between two matched monitoring terminals upstream and downstream of the same pipeline. , These represent the maximum and minimum measured pressure wave velocities during the heating season, respectively. The measured maximum dominant frequency of water hammer wave in the pipeline. A fixed sampling interval is configured for the monitoring terminal. The sampling duration configured for the monitoring terminal, The total sampling time difference for two matched monitoring terminals sampling within the same first time window includes sampling time difference, processing time difference, and communication time difference. For example, in a measured heating pipeline, due to differences in gas content and temperature fluctuations, the measured pressure wave velocity is 800 to 1000 meters per second. With a fixed sampling duration of 60 seconds and a sampling interval of 0.1 seconds, the maximum predicted total time difference is no greater than 20 seconds, and the maximum water hammer frequency component is approximately 5 Hz. Therefore, the pipeline distance between two adjacent matched monitoring terminals upstream and downstream of this pipeline segment should be less than 2400 meters and greater than 100 meters. When primarily measuring the pressure drop of the pipeline segment, the maximum value within this range should be taken as much as possible. Obviously, in this implementation, the configuration parameters of the monitoring terminals in different deployment locations in the system can be appropriately modified to meet the needs of their installation locations. Simultaneously, similar to monitoring the pressure difference between the supply and return water pipelines at monitoring nodes, the sampling interval and sampling duration can be configured to meet the following engineering estimation constraints:

[0038]

[0039] in, A fixed sampling interval is configured for the monitoring terminal. The sampling duration configured for the monitoring terminal, The highest frequency of the pressure signal that needs to be measured on-site. The safety factor related to the average random noise at the site is generally taken as 5 or higher. The maximum water hammer wave period at the site to be covered.

[0040] In the above configuration examples, the sampling duration configured for the monitoring terminal in each sampling run is the first time window described in this disclosure. If both terminals are configured with a baseline sampling duration of 3 minutes, sampling and transmitting three times per hour, starting at 02:00, 22:00, and 42:00 after the hour, then the second sampling after 13:00 UTC each day, which is generally between 13:42 and 13:45, satisfies the same first time window. The sampling duration is within one error range of the baseline sampling duration of 3 minutes, and the sampling completion time is within one error range of the preset time of 13:45 UTC. In the demonstration, each monitoring terminal operates within the first time window, continuously collecting multiple real-time water pressure data at equal intervals. The requirement is to obtain a time series with default order information suitable for compression and decompression without needing to specify the specific sampling time in the communication. The specific sampling interval needs to be pre-configured in the compression and decompression process, but does not need to be transmitted during communication.

[0041] Combination Figure 3 In one example, monitoring terminals E1 and E2 are located on the same monitoring node's water supply and return pipelines, or on adjacent upstream and downstream monitoring nodes on the same pipeline, within a specific first time window during the terminal sampling phase. Each process completes one sampling cycle during its runtime, followed by processing and transmission. For example, it might be configured to sample at equal intervals from 2026-07-15T13:41:30Z (UTC) to 2026-07-15T13:44:30Z (UTC), with real-time compression during sampling. After interception and communication packaging, transmission begins at 2026-07-15T13:45:00Z (UTC) using a communication interrupt. In actual operation... and Each of them represents its actual first sampling window. and These are the actual end time of their first time window and their respective pre-configured sending time. and The relatively fixed on-chip clock interval between these intervals can be called timing advance (TA), which is used to ensure that the monitoring terminal completes one sampling, compression, and transmission preparation. It's easy to understand that, in this example, the actual sampling window duration during operation is first affected by crystal oscillator inconsistency. and All are approximately Not entirely with They are equal, although each has its own pre-configured sending time. and All are configured to the preset settings. The time is as follows: 2026-07-15T13:45:30Z (UTC). However, in the demonstration, due to cumulative errors such as the on-chip clock reference clock and clock drift, the actual sampling start time of E1 is slightly later than the set time. The actual sampling start time of E2 is slightly later than the set starting time. It is worth noting that, to minimize energy consumption caused by time synchronization, these accumulated errors have a relatively large tolerance in this embodiment of the disclosure, such as not exceeding 1 / 5 to 1 / 3 of the length of the first time window, based on the alignment index. As an example, due to the uncertainty brought about by the large network latency fluctuations in low-power wide-area network communication, the actual E1 received on the cloud server side is... The first sequence of timestamps obtained On the contrary, it may be earlier than, or in some instances, close to or significantly later than, the received E2. The first sequence of timestamps obtained .

[0042] Since E1 and E2 already satisfy the spatial matching relationship of being either supply and return water pipelines located at the same monitoring node, or adjacent upstream and downstream pipelines located on the same pipeline, at the same time and The reception time of the first sequence, which is considered to be acquired, compressed, and truncated within the same first time window, and Time difference If the sequence is less than a configured second time window, the first sequences received by the two cloud servers are considered to satisfy the temporal matching relationship disclosed herein. Clearly, if E2 and E3 also satisfy the spatial matching relationship, their first sequences obtained in different first time windows can exist in different second time windows. For example, if E1 and E2 are each configured to have an on-chip transmission time of 2026-07-15T13:25:00Z (UTC), and E2 and E3 are configured to have another on-chip transmission time of 2026-07-15T13:45:00Z (UTC), then a second time window can be set to query whether the first sequence from E1 and E2 has a reception time between 2026-07-15T13:20:00Z (UTC) and 2026-07-15T13:30:00Z (UTC) to determine whether the two sequences meet the time matching condition. At the same time, a second time window can be set to query whether the first sequence from E2 and E3 has a reception time between 2026-07-15T13:44:00Z (UTC) and 13:50:00 UTC to determine whether the two sequences meet the time matching condition. In practice, the second time window used on the cloud server to determine whether two sets of first sequences from spatial matching monitoring terminals satisfy a temporal matching relationship can be dynamically adjusted. This adjustment should aim to maintain the alignment quality of the two first sequences at a high level, for example, by keeping the window as small as possible, while simultaneously maximizing the temporal matching success rate, for example, by making the window as large as possible. In some system embodiments using a static second time window, a time calibration strategy can be deployed on both the monitoring terminal side and the cloud side. This time calibration strategy is executed when the failure rate of finding matching first sequences from two monitoring terminals increases to a certain threshold, thereby minimizing the execution of the time calibration strategy on the monitoring terminal side to avoid communication power consumption.

[0043] It is readily understood that when different monitoring characteristics are required in the field engineering system in other instances, those skilled in the art can determine the deployment location of the monitoring or estimation terminal and configure the acquisition parameters of the monitoring terminal related to this disclosure based on common technical knowledge in the field. The first time window in some embodiments of this disclosure is determined by the number of machine cycles configured on-chip. For example, the sampling duration of each monitoring terminal is configured to 60k, or it can be configured by an external low-power timing circuit, or by an on-chip low-power timer (RTC) module. The configuration goal is that the duration of two consecutive samples from the same device should be consistent, and the sampling duration for different devices to calculate the difference should be basically consistent. Simultaneously, when each device continuously samples the same AD channel, the time interval between two samples should be consistent so that a more standard time series can be obtained when the monitoring terminal compresses the data. This eliminates the need for encoding and decoding operations on the sampling time of each sample. Since the crystal oscillator error of the monitoring terminal has a standard value, the overall sampling error of this sequence on the time axis is controllable.

[0044] In some exemplary implementations, the monitoring terminal is installed on a straight pipe section at the monitoring node. In some implementations, when a pre-designated monitoring node is unsuitable for deployment due to civil structure or other physical conditions (electromagnetic interference, network signal strength, or water hammer sources), the monitoring terminal can be deployed upstream or downstream, depending on the permissible deployment distance. For example, if the monitoring node has a valve, in general projects, the upstream distance from the valve should be at least 10 times the pipe diameter, and the downstream distance should be at least 5 times the pipe diameter, with the communication feeder still originating from that monitoring node. In other implementations, such as... Figure 4 As shown, if the preset monitoring node P3 cannot be deployed, suitable nodes P31 and P32 should be found in the two downstream branches near P3 during construction to deploy monitoring terminals. In other implementations, when P3 cannot be deployed, if the distance between P2 and P3 is close enough and the distance between P1 and P2 meets the measurement requirements, P31 and P32 may not be deployed downstream of P3, but the distance constraints for the P2-P4 and P2-P6 pipe sections must be met. In some implementations, the cloud server stores the associated geographic elevation of each monitoring node and maps it to specific pipe sections based on the GIS pipe network topology. Each pipe section is configured with pipe diameter, length, roughness, and other pipe section parameters. The purpose of each monitoring node in the system is to measure real-time supply water pressure, real-time return water pressure, real-time supply water temperature, and real-time return water temperature. Unless otherwise specified, all collected and calculated values ​​are real-time values. Align the data acquisition windows of the monitoring terminals in the same pipe section and calculate the supply water pressure drop and return water pressure drop within the same data acquisition window of each pipe section after correction of the static pressure head, which is basically independent of the flow velocity.

[0045] It is easy to understand that in the technical field of this disclosure, the existing monitoring terminal deployment methods can basically meet the requirement of obtaining water pressure data from any specific monitoring terminal in the system to provide a monitoring basis. However, based on the above-mentioned deployment and configuration of multiple monitoring terminals, it is easy to see that, due to the influence of deployment location, pipe pressure fluctuation characteristics, on-chip clock accuracy, time difference between devices, communication time difference, etc., the existing technology cannot actually obtain the true difference data from the data of two single monitoring terminals, and cannot achieve effective difference monitoring and other related measurement data, such as pipe section pressure drop, node supply and return water pressure difference, etc. In addition to the influence of conventional measurement, it is mainly affected by static pressure difference caused by factors such as node height difference, temperature difference, and differences in pipe structure, which is unrelated to the flow velocity of the pipe medium. By way of example, considering the height and temperature difference of the deployment location of the monitoring terminals at both ends of the pipe section, the theoretical static pressure difference in engineering can be regarded as:

[0046]

[0047] in, It is a reference density, such as the density at the design temperature, the equivalent density of the gas-liquid mixture measured during system operation, and other constants with controllable errors. The estimated overall static pressure difference for the project. The height difference between the deployment locations of the two monitoring terminals and The water temperature is collected by monitoring terminals at the inlet and outlet ends of the medium flow direction, respectively. is the gravitational constant.

[0048] It is easy to understand that when the acquisition error is controllable, the static pressure difference calculated according to this method has an assessable and quantifiable error range. The above implementation is only an example of how to obtain the static pressure difference in this disclosure. Because the static pressure difference of a pipe segment at a specified time and space in engineering transient analysis can be regarded as a fixed value, those skilled in the art can use any means to improve the static pressure difference calculation method in this disclosure to obtain higher accuracy and a calculable error range, so as to obtain the pressure difference value between matching monitoring terminals after deducting the static pressure difference. These differences include the pressure drop of a pipe segment or the supply and return water pressure difference at a branch location.

[0049] Although methods for obtaining the static pressure difference between two monitoring nodes are readily available in the art, due to the technical difficulties mentioned in this disclosure, it remains difficult to directly obtain the relative dynamic pressure difference between the two monitoring nodes after deducting the static pressure difference. Through the configuration of the sampling phase in several system embodiments of this disclosure, in some system embodiments, a cloud server coupled to the monitoring terminal data, including a data server, application server, or any distributed computing device deployed in the cloud, decompresses the first sequence of the matched two monitoring terminals and performs time series alignment to obtain the first difference sequence of the aligned portion.

[0050] In several embodiments of one aspect of this disclosure, the matched monitoring terminal is configured to obtain a compressed sequence of real-time water pressure data within its first time window. The specific compression method is a lossy linear fitting compression algorithm with a maximum error condition, such as Piecewise Linear Approximation (PLA) or Spinning Door Transformation. These algorithms can be used within a maximum error limit (…). This method performs real-time compression with error control under a metric, consuming fewer computational resources overall. While ensuring the compression meets the maximum error condition, the specific compression method is not limited to wavelet transform or even existing compression methods such as Chimp and Elf / Elf+ used locally to enhance local accuracy. Furthermore, due to the truncation strategy of this disclosure, when truncating the compressed sequence of real-time water pressure data from back to front, some linear fitting lossy compression algorithms that use continuous or semi-continuous fitting have smaller truncation errors at the truncation position. For discontinuous fitting methods such as rotating gates, the tolerance can be controlled during compression, or the degenerate endpoint can be fixed to ensure that an even number of points are truncated from back to front to obtain the first sequence for transmission.

[0051] In several embodiments involving at least specific compression, truncation, and alignment methods, for monitoring terminals with pre-configured spatial matching relationships on a cloud server... , The real-time water pressure sequences collected within the same specific first time window are all of length N. These sequences are then compressed using a linear fitting lossy compression algorithm in real time to obtain compressed sequences. and , As an example, with a compression ratio of approximately 100:1, the original sequence data length is approximately 6k words, and the compressed data length is approximately 120 words. A specified number of segments (s) are extracted from the end to the beginning. and This serves as the first sequence sent to the cloud server at a designated transmission time based on its on-chip clock. Since the length of the compressed summary obtained by the compression algorithm is not fixed, when s is greater than m or greater than n, the excess portion is padded with a default value. In some specific NB-IoT demonstrations, a fixed length of 100 characters is extracted from the end of the compressed data as the first sequence to be sent. This is then placed into the PayLoad segment of the message according to the CoAP protocol specification and packaged and sent at a preset transmission time. The total message length can be controlled within 400 bytes.

[0052] For received on the cloud service side and If the receiving time of each device is within a specific second time window based on the network standard time, thus satisfying the time matching condition, then it is considered that the cumulative clock drift between the first time window based on the on-chip clock and the set first time window is within an acceptable range, and subsequent effective time series alignment operations can be performed to obtain the difference sequence located in the aligned part. For the difference sequence after deducting the static pressure head, the difference sequence of time alignment can be uniformly deducted according to formula (3) or similar engineering estimation. This deduction does not affect the difference sequence in subsequent analysis when different monitoring terminals are matched in a single pair. However, in multiple continuous pipelines or pipelines with correlation, such as E1-E2 matching and E2-E3 matching, if the receiving time of the first sequence of the E1 to E3 pipeline segment is in the same second time window, then the two difference sequences can be summed. This summation has a calculable error range.

[0053] In one aspect, in some specific examples of decompressing the first sequences of two matched monitoring terminals and aligning them over time to obtain a difference sequence, the steps S100 to S400 are included:

[0054] S100: Obtain the two first sequences returned by the query and their data identifiers. The data identifiers are used to query the sequence information of the first sequence stored on the cloud server. The sequence information includes the sampling monitoring terminal identifier, the first sampling time window, the sampling interval, the sampling delay, and the transmission time. It is easy to understand that the sampling monitoring terminal identifier is used to match spatial relationships to obtain relevant static pressure difference information; other parameters cannot be directly used due to on-chip clock errors in the monitoring terminal.

[0055] S200, select the difference mode. The difference mode includes whether to subtract the static pressure difference, whether to directly filter the first sequence, and whether to perform pre-alignment, etc.

[0056] S300, according to the configuration in step S200, the first sequence or the second sequence generated after direct filtering of the first sequence is decompressed or resampled to obtain a time series for alignment. If the first sequence has padded data, the padded data is not considered in the processing, and the calculation starts from the non-padded data in the sequence.

[0057] S400, according to the configuration in step S200, use the alignment algorithm to obtain the aligned part of the two output time series in step S300, and output the difference of the corresponding values ​​at each time point in the aligned part as a difference sequence.

[0058] In multiple implementations, pre-alignment includes pre-alignment operations before and after decompression or resampling in step S300.

[0059] A sample time alignment operation in S300 includes steps S3100 to S3102:

[0060] S3100, querying the first sequence returned by the cloud service at the same time interval. and Resampling was performed separately to generate sequences. and Resampling includes decompression performed at a fixed sampling interval on the monitoring terminal side according to the original settings. and The resampling interval is redefined. It is easy to understand that this resampling satisfies the maximum error condition during sampling compression.

[0061] S3101, for the sequence and Alignment is performed using dynamic time planning to obtain the optimal warping path. Examples, such as Figure 5 In an alignment operation, Time series and The distance matrix is ​​measured in Euclidean distance, where:

[0062]

[0063] make Distance matrix The dynamic time-bending path on, where It is the first of the paths There are n elements, and they satisfy boundary, continuity, and monotonicity constraints:

[0064]

[0065] It's easy to understand that there are multiple dynamic time-warped paths on the distance matrix that satisfy the conditions. Among all dynamic time-warped paths, the sum of DTW distances, for example, when using Euclidean distance as a metric, is... The path that recursively minimizes the curve is the optimal curved path, which this disclosure refers to as the aligned path. The sum of the DTW distances along this path can be denoted as:

[0066]

[0067] Generally, there is a relatively stable difference between the two first sequences. This difference includes the static pressure difference and the dynamic pressure difference related to the flow velocity in the long-distance pipeline. The distance in formula (4) can also be measured using the Manhattan distance metric after deducting the static pressure difference, for example:

[0068]

[0069] in, To provide The location of the monitoring terminal points to the provided The static pressure difference at the monitoring terminal location is such that the actual sampling duration corresponding to the first sequence received by the cloud server is basically close to the first time window preset by the monitoring terminal. The duration of the signal is generally several times the actual maximum water hammer wave period in the field. Therefore, the actual alignment is based on the delay transmission of the low-frequency pressure wave in the pipeline between the matching monitoring terminals.

[0070] S3102, trim and remove the "gap" portions with a slope of 0 or 1 at the beginning and end of the aligned path to obtain a continuous path. This continuous path is the first sequence. and The aligned portion obtained by time series alignment corresponds to the duration of the first sequence. and Based on the duration corresponding to the resampling time interval. This part has a defined duration and corresponding relationship, depending on the pipeline design flow direction, such as... The collection location is Upstream of the acquisition location, the first difference sequence of the aligned portion of the two sequences is represented as:

[0071]

[0072] When the first difference sequence needs to be subtracted from the static pressure difference, it can be represented as:

[0073]

[0074] It is easy to understand that time dynamic programming algorithms are existing technologies. If the alignment effect is adjusted to reduce the alignment error or reduce the complexity of the alignment algorithm, the above steps such as the distance matrix can be improved, such as by using the Needleman-Wunsch algorithm, which has a better fitting effect on multiple line segments. Figure 6 Show another continuous path The existence of multiple alignment examples with slopes of 0 or 1, representing local one-to-many or many-to-one "gaps," suggests that these gaps may contain resampled, unseparated high-frequency components. Since multiple consecutive high-frequency waves may cause long gaps along a continuous path, a sliding window approach can also be used to obtain the aligned portions of two first sequences for the purpose of aligning low-frequency waveforms. An example time alignment operation in S300 includes steps S3200 to S3203:

[0075] S3200, querying the first sequence returned by the cloud service at the same time interval. and Resampling was performed separately to generate sequences. and .

[0076] S3201, Select the range of sliding window length Traverse the sequence over its entire length. and The pursuit Previous child window and Previous child window Normalized cross-correlation coefficient:

[0077]

[0078] in, and These are the means of the two sub-windows, respectively. The closer the value is to 1, the more similar the two sub-windows are. This operation is equivalent to dividing a window of length L into separate sequences. and Swipe up to see more. Each starting position in ,and Each starting position in Perform pairing scoring. It's easy to understand that this example uses normalized cross-correlation coefficients as the similarity score. Other examples may use metrics such as cosine similarity as the alignment score. Some examples using deep learning models for alignment may also set conventional long-range or short-range cues to achieve time series alignment.

[0079] S3202, find among all calculated normalized cross-correlation coefficients that... The largest Triples, at this time the window subsequence in and This refers to the aligned portion. Referring to step S3102, the first difference sequence of the aligned portion of the two sequences is represented as:

[0080]

[0081] In some instances where alignment is guaranteed, a threshold can be used to evaluate the significance of the alignment because a lower limit for the length of the sliding window is set. Simultaneously, based on the verification of the data acquisition principle, the time window that satisfies the maximum score is... If i or j is significantly close to 1, i+L is significantly close to m, or j+L is significantly close to n. It is easy to understand that these methods, primarily using long-period water hammer waves for alignment, are suitable for obtaining the water pressure drop in the pipeline between monitoring nodes. After alignment, the difference sequence of the aligned portion can reflect smaller long-term changes in the pipe section.

[0082] Since the purpose of subsequent data analysis is not limited to displaying the waveform of the difference sequence, some examples involve filtering the first sequence received from the cloud server before performing the alignment operation. For example, for the first sequence obtained using the PLA compression algorithm based on the maximum error condition, median filtering can be used directly to eliminate the influence of low-frequency or high-frequency components on the sampled sequence and improve the alignment effect. An example time alignment operation in S300 includes steps S3300 to S3302:

[0083] S3300 uses a digital filter to process the first sequence returned by the cloud service query. and The second sequence was obtained by performing filtering processes separately. and Compared to conventional resampling or decompression filtering, it satisfies the maximum error criterion ( The PLA compression summary allows for direct filtering in the compressed state. An example is the median filter based on the maximum error criterion provided in Chinese publication CN112487880B. This filter processes the original signal sequence P through a sequence information window of specified width w, treating the longer wavelength water hammer wave as a baseline drift of hydrostatic pressure, thus preserving high-frequency quantities. In some instances, the second sequence can be output separately, using the median of the filtered sequence as the absolute water pressure reading at that time point. The preserved high-frequency water hammer waveform can then be used in conjunction with the time series for subsequent water hammer analysis.

[0084] S3301, for the second sequence and Resample at equal time intervals. To preserve the high-frequency quantities used for alignment, the time interval should be less than or equal to the actual sampling interval.

[0085] S3302 performs an alignment operation on the two resampled time series to obtain the aligned portion, and obtains the second difference sequence based on the upstream and downstream relationship. In some instances, the static pressure difference obtained from the query is subtracted from the second difference sequence during calculation.

[0086] It is easy to understand that, for the supply and return water pressure difference of the monitoring node, the high-frequency water hammer brought by the upstream and downstream regulating valves arrives at basically at the same time. Therefore, by aligning the short-period waveforms of the time series obtained by sampling, a relatively stable supply and return water pressure difference of the monitoring node can be obtained. The pressure difference fluctuation is significantly smaller than the mean difference after filtering.

[0087] It should be noted that although the configuration of the monitoring terminal in the various system embodiments of this disclosure is described using computer methods, those skilled in the art can clearly understand from the above description of the implementation methods that each implementation method can be implemented using software plus a general-purpose hardware platform. For example... Figure 7In one example, the monitoring terminal hardware platform includes a power management unit 1, a storage unit 2, a sensing unit 3, a processing unit 4, and a network unit 5. The power management unit 1 provides power management functions such as device power supply and low power consumption. The storage unit 2 stores compressed data and readable / writable configuration parameters. The sensing unit 3 collects physical quantities. The processing unit 4 uses a low-power MCU to execute relevant instructions, such as the STM32L071 or ESP32-C3 series. The network unit 5 is a radio frequency module for implementing specific NB-IoT access, such as the Quectel BC95 series or the Gosuncn ME3612. However, this does not mean that the configuration method in this example can only be implemented on the aforementioned discrete architecture modules. The existing Quectel BC95 already integrates an AD sampling circuit. Any hardware platform combining these functional modules according to existing technology can also be configured to provide the monitoring terminal disclosed herein. Depending on the characteristics of different hardware platforms, those skilled in the art can implement the specific method steps of this disclosure in various unit combinations. In a hardware platform that integrates AD functionality using the Quectel BC95 class, network unit 5 covers part of the functionality of sensing unit 3, and processing unit 4 can call network unit 5 to perform AD conversion.

[0088] Demonstration Figure 7In an example of a device configuration method for a monitoring terminal hardware platform, the power management unit 1 is configured to provide a global on-chip clock to provide corresponding timed wake-up hardware interrupts, as well as corresponding operating voltage and current. The hardware interrupts include a first interrupt for initiating sampling within a first time window and a second interrupt for initiating radio frequency transmission after the first time window ends. The processing unit 4 is configured to, upon responding to the first interrupt, read the last real-time water pressure data record required for compression in the first buffer of the storage unit 2, and the compressed sequence data and status in the second buffer, and initiate water pressure acquisition and real-time compression, storing them in their respective buffers before going into sleep mode. Upon responding to the second interrupt, it starts a transmission timer, checks the compressed sequence data and status in the second buffer, prepares the application data packets to be transmitted, and waits for the transmission timer to overflow. Then, it outputs a pin to notify the power management module 1 to switch power supply modes and simultaneously wakes up the network module 5 to activate its radio frequency window. After completing transmission and instructing the network module 5 to go back to sleep, it also goes back to sleep. It is easy to understand that the first interrupt is used to achieve continuous acquisition of multiple real-time water pressure data at equal intervals, and the parameters of the storage unit 2 are used to configure the duration of the first time window in conjunction with the instructions of the processing unit 4 to achieve acquisition within the specified time and length of the first time window. Processing unit 4 is mainly used to store and execute instructions for acquisition and compression algorithms and communication flow control. The second interrupt and transmission timer ensures that the end time of the first time window and the transmission time have a basically fixed on-chip clock interval. In some improvements, the basically fixed on-chip clock interval between the end time of the first time window and the transmission time is set to satisfy the maximum value of the transmission preparation time advance (TA) and data preparation time. The time source of the time advance (TA) is derived from the timestamp of network signaling, measured by the base station and sent to the terminal, with its reference being the base station's reception time window. This is used to indicate how much time the terminal needs to prepare to transmit uplink signals in advance to compensate for propagation delay and ensure that the uplink signals of all terminals are aligned on the base station side. In some improvements, during the RF window activation period, processing unit 4 also receives downlink instructions from the server and, after processing, such as performing specified read / write operations on the storage unit, enters a sleep state.

[0089] On one hand, the system embodiments disclosed in this disclosure provide several aspects of a long-distance heating pipeline monitoring system that can essentially achieve the acquisition of the true time difference without significant time synchronization. Although the system tolerance is extremely large, it still cannot avoid the significant time drift error between the on-chip times of monitoring terminals that have been in long-term dormant operation. Specific methods used in some implementations of this aspect are used to solve this problem. These methods include the step of: when the first sequence of two matching monitoring terminals returns empty or the duration of the aligned portion obtained from the two returned first sequences is less than a threshold, adjusting the on-chip clocks of the two matching monitoring terminals so that their respective first time windows are substantially the same standard time. On the other hand, these method examples provide technical solutions for addressing the timing and target of significant drift. Clearly, when the server responds to internal or external query requests for the first sequences of two matched monitoring terminals, it can conditionally assess the approximate or accurate alignment quality of the returned first sequences. Specifically, this can be evaluated based on the time length of the aligned portion. Since the segments of the first sequences contain temporal information, and the time interval between adjacent time sequences is predefined, at least a method such as Direct Dynamic Time Planning (DTW) without decompression or a small-scale self-attention mechanism considering short-range cosine similarity, or a learnable deep neural network, can be used to estimate the approximate continuous alignment duration, i.e., the time length of the aligned portion. For cases exceeding a given threshold, such as when the time window width is less than half, the cloud server can send a time synchronization command to the two monitoring terminals via the downlink channel, instructing them to complete a network time synchronization to adjust the on-chip clocks of the two matched monitoring terminals so that their respective first time windows are approximately the same standard time. This is clearly a local time refresh strategy and will not affect the normal uplink procedures of other monitoring terminals in the system.

[0090] This disclosure provides several embodiments demonstrating sampling compression at system monitoring terminals and query and alignment algorithms on cloud servers. In another application embodiment based on these demonstrations, a method for displaying information about a long-distance heating pipeline network is included. Based on the system embodiment, a data display service is configured in the application server. This service sends query requests for two specified monitoring terminals to the cloud data server according to refresh parameters. The query request provides a pre-agreed second time window. When the cloud data server returns a non-empty first sequence of matching two monitoring terminals, the application server calculates the water pressure readings and / or water pressure difference readings of the two monitoring terminals based on the returned first sequence. The updated, most recent data is then pushed to the connected client in the data display service to refresh the client display. These application servers may not be part of the monitoring system; they obtain data from the long-distance heating pipeline network monitoring system of this disclosure through a data coupling interface.

[0091] In other application embodiments of the system solutions disclosed herein, the platform-layer data server or the application-layer application server directly provides users with readings of one or more physical quantities, including water pressure at monitoring nodes, supply and return water pressure difference at monitoring nodes, and water pressure drop in pipelines between monitoring nodes, through a human-computer interaction terminal; these methods include channel subscription or push notifications. Although the human-computer interaction terminal is not considered a necessary part of the system, if the data it obtains is the dynamic execution result of the system configuration solution provided herein, it should be considered as using the methods provided herein.

[0092] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-described device configuration and execution steps and system application methods.

[0093] It is easy to understand that the various specific implementations disclosed in this document include a long-distance heating pipeline monitoring system and a long-distance heating pipeline monitoring method. In different specific implementations, based on the emphasis of the described technology, other contents that can be easily understood or known based on the existing technology or the content disclosed in this document have been omitted.

Claims

1. A long-distance heating pipeline monitoring system, comprising a cloud server and a monitoring terminal that transmits data to the cloud server via a low-power wide area network, characterized in that: The monitoring terminals, deployed on the supply and return water pipelines of each monitoring node in the long-distance heating pipeline network, are configured as follows: Within the first time window, based on the on-chip clock, multiple real-time water pressure data are continuously collected at equal intervals. Obtain a compressed sequence of real-time water pressure data within the first time window, and extract a first sequence of fixed length from the compressed sequence from the end to the beginning. The first sequence is sent to the cloud server at a specified transmission time based on the on-chip clock; The cloud server is configured as follows: Receive the first sequence, and, A first sequence of two matching monitoring terminals is found, wherein the matching relationship satisfies that the reception time of the first sequence, based on standard time, is within a second time window, and the monitoring terminals are respectively located in the water supply pipeline and return pipeline of the same monitoring node or in adjacent upstream and downstream monitoring nodes of the same pipeline; Wherein, the end time of the first time window has a basically fixed on-chip clock interval with the transmission time, and the length of the second time window is less than the interval between two adjacent transmission times configured by the monitoring terminal.

2. The long-distance heating pipeline monitoring system according to claim 1, characterized in that: The monitoring terminal is also configured to send the water temperature of its deployed pipeline corresponding to the first time window to the cloud server.

3. The long-distance heating pipeline monitoring system according to claim 1 or 2, characterized in that: When two matched monitoring terminals are located at adjacent monitoring nodes of the same pipeline, the cloud server is also used to find the static pressure difference of the pipeline between the adjacent monitoring nodes, the static pressure difference including the static pressure difference caused by the height difference between adjacent monitoring nodes.

4. The long-distance heating pipeline monitoring system according to claim 3, characterized in that: The cloud server is also used to first decompress the first sequences of the two matched monitoring terminals respectively, and then perform time series alignment to obtain the first difference sequence of the aligned part.

5. The long-distance heating pipeline monitoring system according to claim 3, characterized in that: The cloud server is also used to directly filter the first sequence received from the monitoring terminal to generate a second sequence.

6. The long-distance heating pipeline monitoring system according to claim 5, characterized in that: First, the second sequences of the two matched monitoring terminals are decompressed, and then time series alignment is performed to obtain the second difference sequence of the aligned part.

7. The long-distance heating pipeline monitoring system according to claim 4 or 6, characterized in that: The difference in the difference sequence is the dynamic pressure difference after deducting the static pressure head of the corresponding pipeline.

8. The long-distance heating pipeline monitoring system according to claim 1, characterized in that: The compression method for obtaining the compressed sequence of real-time water pressure data within the first time window is a linear fitting lossy compression algorithm with the maximum error condition.

9. A method for monitoring long-distance heating pipelines, applied to the long-distance heating pipeline monitoring system described in claims 1 to 8, characterized in that, The steps include: when the first sequence of the two matching monitoring terminals returns empty or the duration of the aligned portion obtained by the two returned first sequences is less than a threshold, the on-chip clocks of the two matching monitoring terminals are adjusted so that their respective first time windows are basically the same standard time.

10. A method for monitoring a long-distance heating pipeline network, based on the long-distance heating pipeline network monitoring system described in claims 1 to 8, characterized in that: When the first sequence of two matching monitoring terminals returns non-empty, the physical quantity readings provided to the user are calculated and / or updated through the human-computer interaction terminal. The physical quantity readings include the readings of one or more physical quantities among the water pressure at the monitoring node of the heating long-distance pipeline monitoring system, the supply and return water pressure difference at the monitoring node, and the water pressure drop in the pipeline between the monitoring nodes.

Images