Method and system for operation and maintenance management of a heating system

Abstract

This invention relates to the field of heating technology and discloses a method and system for the operation, maintenance, and management of a heating system. The method includes acquiring sensor data, repair work order data, and the location and status of maintenance teams from the heating network; calculating the degradation degree of network sections based on the sensor data, marking source nodes, and generating aggregated fault groups; calculating the time cost and value recovery parameters for each team within the fault group to determine the optimal matching team; based on the estimated arrival time of the optimal team, performing a throttling operation on the upstream electric regulating valve when the triggering conditions are met to proactively intervene in the fault development; terminating the throttling operation after receiving a repair confirmation signal, and disbanding the aggregated fault group after confirming that the degradation degree has decreased and the efficiency has met the standards. This invention integrates multi-source data and optimizes maintenance resource allocation through optimal scheduling and proactive intervention, thereby improving fault response efficiency and system operational safety.

Description

Technical Field

[0001] This invention relates to the field of heating technology, specifically to a method and system for the operation, maintenance and management of a heating system. Background Technology

[0002] Heating systems are an important part of urban infrastructure, and the efficiency and reliability of their operation, maintenance and management directly affect heating quality and energy consumption.

[0003] In existing heating system operation and management, the physical condition monitoring system of the pipeline network and the customer service system on the user side are usually independent. This separation leads to fault response relying mainly on user reports or alarms based on single physical parameter thresholds, lacking the comprehensive diagnostic capability to pinpoint the root cause of the fault. Maintenance team scheduling and allocation are often based on experience or simple geographical proximity principles, failing to quantitatively assess and optimally match the urgency of maintenance tasks, economic impact, and the real-time operational status of the teams, thus affecting the efficiency of maintenance resource utilization.

[0004] Furthermore, there is a time delay between the confirmation of a fault and the arrival of maintenance personnel on-site to address it. During this period, abnormal conditions such as leaks or pressure loss in the pipeline network continue to worsen, and existing management methods generally lack automated, proactive risk intervention mechanisms, making it difficult to effectively control the further development of the fault and posing safety hazards. Therefore, how to integrate multi-source information to achieve accurate diagnosis, optimized scheduling, and proactive risk control is an urgent problem to be solved in the field of heating system operation and maintenance. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for the operation and maintenance management of heating systems. This solves the problem that in the existing operation and maintenance management model of heating systems, monitoring data on the physical status of the pipeline network and business repair data from the user side are usually processed independently, resulting in delayed fault response, low scheduling efficiency, and a lack of proactive prevention and control measures.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] The first aspect of this invention provides a method for the operation, maintenance, and management of a heating system, comprising:

[0008] Acquire sensor data from the physical monitoring nodes of the heating network, repair work order data generated by the customer service module, location coordinates of the mobile terminal belonging to the maintenance team, and the work status of the maintenance team;

[0009] The degradation value of the pipeline section is calculated based on the sensor data. When the degradation value exceeds the set baseline, the source node to be repaired is marked, and an aggregated fault group is generated by combining the repair work order data.

[0010] For the aggregated fault group, and in combination with the location coordinates and the operation status, calculate the time cost parameters and value recovery parameters of each available maintenance team, and calculate the maintenance benefit-time ratio based on the time cost parameters and value recovery parameters, so as to find the optimal matching team.

[0011] Extract the estimated arrival time from the time cost parameters of the optimal matching shift, and based on the estimated arrival time, perform a physical throttling operation on the electric regulating valve upstream of the source node to be repaired when the triggering condition is met;

[0012] After receiving a repair confirmation signal from the mobile terminal, the physical throttling operation is terminated, and the aggregated fault group is disbanded when the degradation value of the source node to be repaired falls back to within a preset range.

[0013] Preferably, the step of calculating the degradation degree of a pipeline section involves extracting the actual physical parameters and theoretical operating parameters of the pipeline section, and calculating based on the deviation between the two using a computational model that includes independent deviation terms and cross-coupling terms. This model can comprehensively evaluate anomalies in both thermal and hydraulic dimensions, and amplify the characterization of concurrent faults through cross-coupling terms, thereby quantifying concurrent faults more accurately.

[0014] Preferably, the time cost parameter is derived by superimposing the estimated remaining time determined based on the work status, the estimated travel time calculated based on the positioning coordinates, and the preset standard operation time; the value recovery parameter is derived by summing the physical heat loss recovery value calculated based on the sensor data and the business default penalty avoidance value calculated based on the repair work order data included in the aggregated fault group. This parameter construction method provides a comprehensive quantitative basis for scheduling decisions.

[0015] Preferably, the step of determining whether the triggering condition is met is achieved by predicting the physical state parameters of the source node to be repaired at the time of the shift's arrival, based on the estimated arrival time, and comparing them with a preset minimum safe operating threshold. When the predicted value is lower than the minimum safe operating threshold, the triggering condition is determined to be met, thereby initiating active intervention.

[0016] Preferably, the physical throttling operation is carried out by combining the difference between the current physical state parameters of the source node to be repaired and the minimum safe operating threshold, as well as the real-time operating status such as the fluctuation rate of the pipeline pressure, to construct a hydraulic buffer constraint model, and dynamically solve the target opening command of the electric regulating valve based on the model to generate a smooth regulating command to suppress the water hammer effect caused by excessively rapid regulation.

[0017] Preferably, before terminating the physical throttling operation, the method further includes initiating a pressure holding verification after receiving the repair confirmation signal. This verification objectively determines whether the physical repair meets the standard by calculating the pressure drop gradient of the repaired area within a preset pressure holding duration and comparing it with a pre-stored allowable leakage rate threshold.

[0018] Preferably, the step of terminating the physical throttling operation is to gradually restore the electric regulating valve to its standard opening degree according to the exponential dynamic unlocking curve after determining that the physical repair has met the standard, so as to smoothly restore the pipeline flow and reduce the pressure impact on the pipeline system.

[0019] Preferably, the step of disbanding the aggregated fault group involves extracting the associated end-point heating parameters for each repair work order within the group to calculate the heat recovery efficiency index. Only when the heat recovery efficiency index of a specific repair work order reaches a preset efficiency threshold is that specific repair work order decoupled from the aggregated fault group, thereby verifying that the heating effect on the user side has been restored to the service standard.

[0020] Preferably, the step of acquiring sensor data from physical monitoring nodes of the heating network further includes performing timestamp alignment on the acquired sensor data, and performing zero-order preservation or linear interpolation to complete the data that fails to be aligned by the timestamp, thereby providing time-consistent and data-complete input parameters for subsequent analysis and calculation.

[0021] A second aspect of the present invention provides an operation and maintenance management system for a heating system, used to implement the method described in any of the preceding claims, comprising:

[0022] On the operational side, there are sensor networks used to collect physical parameters of the heating network, and actuators equipped with electric regulating valves driven by programmable logic controllers.

[0023] The information terminal includes a customer service module for receiving repair work order requests from users, a mobile terminal for maintenance teams, and a scheduling engine for performing data processing and logical operations.

[0024] A communication interface is established between the information terminal and the operation terminal to enable bidirectional data flow between the physical world and the information system.

[0025] This invention provides a method and system for the operation, maintenance, and management of a heating system. It has the following beneficial effects:

[0026] 1. This invention optimizes the matching of available maintenance teams by constructing a maintenance benefit-time ratio evaluation index. It quantifies the current operating status, location information, estimated travel time, and standard operation time of each maintenance team into time costs, while simultaneously quantifying recoverable physical heat loss and avoidable business default penalties into value returns. The system makes scheduling decisions with optimal overall benefits, achieving dynamic optimization of maintenance resource allocation, shortening overall fault response and handling time, and reducing operating costs.

[0027] 2. This invention enhances system operational safety by introducing a proactive intervention mechanism based on state prediction. After determining the optimal matching shift, the system uses its estimated arrival time to predict the physical state parameters of the fault node at that moment. If the predicted value is lower than a preset minimum safe operating threshold, a physical throttling operation is automatically performed on the upstream valve before maintenance personnel arrive. This prevents the fault from worsening en route and reduces the risk of secondary accidents such as pipeline depressurization or pipe bursts.

[0028] 3. This invention integrates physical sensor data of the pipeline network, service repair work orders, and maintenance team status data to form a full-process management system from fault discovery to service closure. By aggregating fault groups, the source fault at the physical level is associated with multiple repair requests from the user side. After repair, dual verification is performed through physical-level pressure maintenance verification and service-level thermal recovery efficiency index to ensure that both physical repair and service restoration meet the standards before the fault group is disbanded, thereby improving the quality and completeness of maintenance work. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the system architecture of the present invention;

[0030] Figure 2 This is a schematic diagram of the method flow of the present invention;

[0031] Figure 3 This is a graph showing the evaluation effect of business work order clustering and redundancy compression in this invention.

[0032] Figure 4 This is a comparison chart of the dynamic hydraulic restraint and smooth recovery control curves of the present invention;

[0033] Figure 5 This is a comparison chart of the cross-dimensional scheduling benefits of the work team according to the present invention. Detailed Implementation

[0034] The technical solutions in 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 embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] See attached document Figure 1 This invention provides an operation and maintenance management system for a heating system, comprising:

[0036] On the operation side, it includes a data monitoring module, a sensor network, and an actuator. The sensor network is located at each physical monitoring node of the heating pipe network, and the actuator is equipped with a programmable logic controller and an electric regulating valve driven by it.

[0037] The information end includes a scheduling engine, a customer service module, and a mobile terminal. The customer service module receives and parses the business work order requests initiated by the user, and the scheduling engine performs data processing and logical operations.

[0038] The communication interface is established between the information end and the operation end, supporting the convergence and transmission of physical parameters to the information end and the reverse transmission of control commands to the operation end.

[0039] See attached document Figure 2 This invention provides a method for the operation, maintenance, and management of a heating system, comprising the following steps:

[0040] S100 collects and integrates multi-source status data. The scheduling engine pulls sensor data from each physical monitoring node through the communication interface at a set period, synchronously reads the repair work order data generated in the customer service module, and obtains the location coordinates of the mobile terminal.

[0041] S200 quantifies node degradation and performs spatial clustering. The scheduling engine extracts the actual physical parameters and theoretical operating parameters of the pipeline section, calculates the corresponding degradation value, and marks the degradation value of a specific section as the source node to be repaired according to the pipeline directed graph topology. The downstream repair work orders affected by the fluid kinematics of the node are clustered and aggregated fault groups are output.

[0042] S300 evaluates the maintenance benefit-time ratio and outputs the matching results. For the generated aggregated fault group, the scheduling engine calculates the time cost parameters and value recovery parameters of each available maintenance team to perform the repair task. Based on the aforementioned parameters, it calculates the maintenance benefit-time ratio, performs optimization calculation under preset constraints to determine the optimal matching team, and issues the work instruction to the mobile terminal to which the team belongs.

[0043] S400, reverse injection of time parameters and execution of dynamic hydraulic restraint, the scheduling engine extracts the estimated arrival time of the optimal matching team to the source node to be repaired, and transmits the estimated arrival time back to the data monitoring module. The data monitoring module compares the estimated arrival time with the on-site observation time threshold, and when it determines that the triggering condition is met, it performs physical throttling operation on the electric regulating valve upstream of the source node to be repaired in combination with the preset hydraulic buffer constraint model.

[0044] S500 verifies the physical state recovery and performs business decoupling. After receiving the repair confirmation signal returned by the mobile terminal, the data monitoring module terminates the control operation of the electric regulating valve and restores its standard opening. The scheduling engine continuously monitors the degradation value of the source node to be repaired. When it determines that the sampled values ​​of multiple consecutive cycles have fallen back to the preset operating range, it disbands the corresponding aggregated fault group and triggers the archiving procedure of the associated repair work order.

[0045] In this embodiment, the physical sensing and service work order data acquisition process in step S100 includes the following sub-steps to achieve the extraction and alignment of multi-dimensional heterogeneous data:

[0046] S101, the scheduling engine retrieves sensor data from each physical monitoring node of the heating network at set time intervals via the communication interface. The sensor network is located at the main branch nodes and heat inlet points of the heating network. Pressure transmitters, temperature sensors, and flow meters are deployed within the physical monitoring nodes. As a preferred approach, the communication interface is specifically implemented as an industrial protocol conversion gateway in the hardware data acquisition scenario. Its main function is to encapsulate the low-level register data of field actuators and programmable logic controllers into high-level network packets. For the low-level packet assembly and parsing process involved in the industrial protocol conversion gateway, those skilled in the art can select common protocols such as OP CUA or Modbus TCP for parameter configuration according to the field hardware environment. Its network communication handshake and verification mechanisms are well-known technologies in the field and will not be elaborated upon here.

[0047] To ensure the timing consistency of input data during system analysis, the scheduling engine retrieves data at set intervals. For specific physical monitoring nodes Constructing physical state vectors Based on the principles of fluid mechanics and thermodynamics, this state vector intuitively reflects the energy and mass boundary conditions of a specific cross-section of the pipeline network under transient conditions, serving as a core benchmark for subsequent assessment of pipeline network degradation. The specific model definition is: In the formula, For physical monitoring nodes At any moment The actual pressure collected; This corresponds to the inlet water temperature; This corresponds to the outlet water temperature; The instantaneous flow rate of the pipeline cross-section at the current moment; symbol This is the transpose operation of a matrix.

[0048] The scheduling engine not only performs simple data reading but also timestamp alignment on the received raw data, filtering out delayed data points that exceed the preset clock tolerance window. To avoid runtime dead zones caused by missing input in subsequent algorithms due to filtering of individual data points, for data dimensions that fail to be timestamped, the system uses the values ​​from the previous sampling period to perform zero-order hold or performs linear interpolation to fill in the gaps based on historical trends. This ensures that all physical parameters are within the same transient time slice when performing thermo-hydraulic coupling analysis. The preset clock tolerance window is typically dynamically calibrated based on pipeline data transmission delays and fluid thermal inertia, and its value can be set between 2 and 5 seconds.

[0049] S102, while acquiring the underlying physical network status, the scheduling engine synchronously reads the repair work order data generated in the customer service module and extracts business attribute features. The customer service module aggregates fault report records from the user side, transforming discrete report events into structured business sequences. In the software business interaction scenario, the communication interface is specifically implemented as an application programming interface (API), enabling the scheduling engine to cross system boundaries and pull a set of work orders in the pending allocation state from the customer service module's main database via the standard Hypertext Transfer Protocol (HTTP). By constructing the communication interface as a combination of an API and an industrial protocol conversion gateway, a bidirectional data interaction channel between the physical operating environment and the business operation environment can be effectively established.

[0050] For the extracted repair work orders The data parsing module inside the scheduling engine converts the unstructured text into two sets of key variables that participate in the core algorithm's calculations. The first set of variables is the spatial coordinates of the repair point. The coordinates represent the latitude and longitude mapping of the reported physical address in the standard geographic information system. The second set of variables represents the service level agreement weight parameters. Service Level Agreement Weighting Parameters It is a dimensionless penalty multiplier, generated by the customer service module based on a preset user service level and the urgency of the fault. Specifically, this weighting parameter is obtained by multiplying the base customer level coefficient by the fault urgency coefficient, and its value is typically set between 1.0 and 5.0. For example, when a repair order involves a large-scale pipe burst or a heating outage in a key protected area, The value will be assigned close to the upper limit of the interval. Through the above analysis process, the system can transform routine business requirements into numerical boundary conditions that the subsequent system needs to identify for executing the global matching optimization algorithm based on time and value, thereby providing a reliable basis for calculation.

[0051] In this embodiment, after the initial fusion of physical and business data is completed, the implementation process of step S100 is further extended to the execution domain to dynamically extract the status of maintenance resources within the system. This process specifically includes the following sub-steps:

[0052] S103, the scheduling engine retrieves the location coordinates of mobile terminals belonging to each available maintenance team in real time via the communication interface. The mobile terminals have built-in satellite positioning hardware modules configured to send geographic location data to the information terminal at a preset upload frequency. For the underlying message parsing and base station-assisted positioning processes involved in the satellite positioning module, those skilled in the art can choose a conventional Global Positioning System or BeiDou Navigation Satellite System to implement them based on the actual network coverage environment. The network communication handshake and coordinate transformation mechanisms are well-known technologies in this field and will not be elaborated upon here.

[0053] To construct the spatial distance constraint boundary for subsequent optimization algorithms, the scheduling engine slices the data at the same time. Below, regarding the first [item] within the heating service grid For each available maintenance team, extract the real-time spatial coordinates of their mobile terminals. Considering the potential for satellite signal loss due to building obstruction or positioning drift caused by multipath effects in complex urban environments, the system incorporates a data cleaning mechanism at the input. When the instantaneous jump in positioning coordinates exceeds the theoretical limit calculated based on the physical speed of conventional vehicles, the scheduling engine discards the coordinates of the anomalous jump. As a preferred approach, this theoretical limit distance is set by multiplying the current sampling time interval by the maximum speed limit on urban roads (e.g., 60 to 80 km / h). After discarding the anomalous points, the system replaces them with the valid coordinates from the last high signal-to-noise ratio condition, or performs inertial estimation by combining road network topology and historical vehicle speeds, thereby reducing the probability of severe distortion in subsequent distance-based time cost parameters.

[0054] S104, the scheduling engine synchronously obtains the current task allocation status and resource idle status of each available maintenance team. While uploading spatial coordinates, the mobile terminal also transmits the current team's work status tag via application layer protocol. Based on this, the scheduling engine extracts and constructs a dynamic resource status set for the entire operations and maintenance team. To provide a more refined scheduling basis, the variable values ​​reflecting the team status are not limited to the conventional logic of idle or busy; this embodiment further introduces the expected time dimension. The specific status model is defined as follows: In the formula, For the first Each maintenance team is at all times The comprehensive state set; This is a discrete task status scalar. A value of 0 indicates that the team is currently idle and can be dispatched, while a value of 1 indicates that the team is performing other emergency repair tasks. For when When the value is 1, it represents the estimated remaining time required for the work team to complete the current emergency repair task. If the value is 0, the time parameter will be automatically assigned a value of 0.

[0055] Estimated remaining time The mobile terminal dynamically updates and reports the time taken for the current task's standard operation, combining this time with the system's recorded job time. Specifically, the system uses a formula for calculation. Quantification is performed, among which The standard operating time for this type of fault, The boundary constraint mechanism of the external function effectively avoids negative dead zones in the remaining time due to task timeouts, assuming the actual start time of the task. To handle logical anomalies such as interruptions in the underlying communication gateway or power outages of field personnel's equipment leading to untimely status updates, the scheduling engine has a timeout forced blocking mechanism in the control layer. When the status and location data of a certain shift fails to be successfully retrieved for several consecutive synchronization cycles (usually set to 3 to 5 cycles), the system will temporarily revoke the shift's scheduling participation rights and forcibly lock its status scalar as unassignable until the scheduling engine receives a heartbeat recovery packet initiated by the mobile terminal again.

[0056] In this embodiment, the scheduling engine, after acquiring the underlying multi-dimensional state data, enters the logic operation module to perform deep parameter analysis. This process specifically includes the following sub-steps to quantify and characterize the abnormal operating states of the pipeline fluid:

[0057] S201, the scheduling engine extracts the actual operating parameters of a specific pipeline section and synchronously matches them with the corresponding theoretical design parameters. For each pipeline section in the heating network composed of adjacent physical monitoring nodes, the scheduling engine calculates the actual enthalpy drop of the section based on the collected inlet and outlet water temperatures of the nodes. To simplify complex calculations on-site, the specific value of the actual enthalpy drop is usually obtained by directly multiplying the isobaric specific heat capacity of the pipeline fluid by the temperature difference of the section, where the temperature difference is obtained by subtracting the actual collected temperatures from the physical monitoring nodes at both ends of the pipeline section. While acquiring the thermodynamic state parameters, the scheduling engine calculates the actual pressure gradient using the actual pressure difference between nodes and the pre-stored physical pipeline length of the specific section. The actual enthalpy drop and the actual pressure gradient together constitute characteristic parameters reflecting the current objective pipeline state.

[0058] As a preferred approach, the theoretical design parameters used for comparison are not static fixed values, but are dynamically generated by the system's built-in hydraulic and thermal simulation calculation module based on current outdoor meteorological data, heat source output, and pipeline physical topology. The theoretical design parameters include the theoretical enthalpy drop and theoretical pressure gradient that the pipeline section should have under ideal, fault-free conditions, providing a dynamic benchmark for subsequent anomaly detection. For the solution process of the pipeline fluid dynamics equations involved in the hydraulic and thermal simulation module, those skilled in the art can use conventional graph theory node analysis methods or finite element numerical simulation methods. The basic computational principles are well-known technologies in this field and will not be elaborated upon here.

[0059] In S202, the scheduling engine constructs an algebraic computation model based on the extracted parameter set to calculate the degradation degree of each section. When a substantial leak or blockage occurs in the pipeline network, it typically manifests as a simultaneous abrupt change in thermodynamic and hydrodynamic characteristics. The determination of exceeding limits for a single parameter is easily affected by sensor drift or local turbulence interference, leading to false alarms. Therefore, the system establishes a computational model that includes independent deviation terms and cross-coupling terms. To prevent algorithmic dead zones caused by extreme garbled values ​​output by underlying sensor hardware failures, the scheduling engine incorporates a physical extreme value clamping procedure before parameter input, forcibly constraining actual enthalpy drop and actual pressure gradient exceeding the conventional physical range of hydrodynamics within set reasonable boundaries. Subsequently, the scheduling engine substitutes the actual and theoretical parameters into the model to obtain the degradation degree value used to characterize the degree of anomaly. The specific mathematical expression is as follows:

[0060] ;

[0061] In the formula, This represents the actual enthalpy drop of the pipeline section; This corresponds to the theoretical enthalpy drop; This represents the actual pressure gradient. This corresponds to the theoretical pressure gradient; This is the thermal deviation weighting coefficient; This is the hydraulic deviation weighting coefficient; To characterize the coupling amplification factor for concurrent anomalies, and to avoid the computational dead zone caused by division-to-zero anomalies due to theoretical enthalpy drop or theoretical pressure gradient approaching zero under pipeline pump shutdown or specific operating conditions, a minimal positive real constant is introduced into the model denominator. As a value protection buffer, this constant is typically set to 10. -6 level.

[0062] The values ​​of the coefficients in the above formula need to be calibrated based on historical pipeline operation data. Generally, the weighting coefficients satisfy... Due to the incompressible nature of fluids in heating pipe networks, the propagation speed of pressure waves within the system is much greater than the conduction speed of thermal parameters. Therefore, the pipe network is more sensitive to hydraulic imbalances. Based on this principle, the hydraulic deviation weighting coefficient... The value range is generally set to 0.6 to 0.8. Coupling amplification factor. An empirical constant greater than 1 is set, and its value can be set between 1.5 and 3.0 in a conventional urban heating network. The first two terms of the formula quantify the independent temperature field degradation and pressure field degradation, respectively, while the third cross-coupling term constitutes the core mechanism of this calculation model.

[0063] When the temperature and pressure fields deviate simultaneously, the cross-coupling terms produce a product amplification effect, increasing the final output degradation value. It exhibits nonlinear amplification. This algebraic structure helps to suppress data noise caused by the drift of a single sensor to a certain extent, and improves the system's ability to detect concurrent faults such as actual physical damage to the pipeline network, providing more reliable data support for subsequently locating the source fault node.

[0064] In this embodiment, after obtaining the degradation values ​​of each pipeline section, the logic operation module needs to further perform limit violation judgment and spatial association mapping to achieve cross-domain integration from physical fault location to business work order clustering. This process specifically includes the following sub-steps:

[0065] S203, the logic operation module compares the real-time calculated degradation value with the system's dynamic baseline and marks the source node to be repaired when an out-of-limit condition is triggered. During normal operation, the pipeline network is inevitably affected by environmental noise such as changes in pipe wall roughness, minor local scaling, or normal fluid pulsation. Using a fixed threshold can easily lead to frequent false alarms. As a preferred approach, the system constructs an adaptive dynamic baseline. This serves as a reference benchmark for assessing the degree of degradation. This dynamic baseline combines the statistical characteristics of degradation in the same area under similar historical meteorological conditions and similar load conditions. These historical corresponding load conditions are typically obtained by jointly searching and matching historical databases with a set outdoor temperature tolerance range (e.g., ±2°C) and water supply load range. The specific computational model is constructed as follows: In the formula, This is the moving average of the deterioration value of this specific pipeline section within the corresponding historical operating condition period; The standard deviation corresponds to the historical period and is used to quantify the degree of dispersion and fluctuation of data under normal conditions. This is the baseline tolerance coefficient. The value of this coefficient is usually manually calibrated by maintenance personnel in combination with the overall aging of the pipeline network. In a typical municipal heating network, its reasonable value range is set between 2.0 and 3.0.

[0066] When the scheduling engine detects a certain segment at the current time Degradation value Strictly greater than the dynamic baseline at that moment At this time, a suspected physical anomaly is determined to have occurred in this section. To prevent water hammer phenomena caused by the switching of circulating water pumps in the heat exchange station or the instantaneous action of regulating valves from causing a brief spike in the calculated degradation degree exceeding the limit, the system introduces a time window anti-jitter mechanism at this point. Only when... The scheduling engine only formally identifies a segment as a faulty segment if this inequality condition holds continuously within a set time window (typically covering 3 to 5 consecutive data fetching cycles). If data loss occurs due to individual communication packet drops within this time window, the interpolation compensation mechanism described above is used to maintain the continuity of the determination. After identifying the faulty segment, the system marks the upstream endpoint of the segment as the source node to be repaired based on the fluid flow direction.

[0067] S204, the scheduling engine traces the fluid dynamics impact range of the physical source based on the directed graph model of the pipeline network and merges scattered repair work orders into aggregated fault groups. Due to the physical connectivity of the pipeline fluid system, hydraulic imbalance or heat leakage at a single node usually propagates to the downstream network, causing varying degrees of decline in heating quality for multiple end users, thus generating a large number of seemingly isolated repair work orders in the customer service module. To avoid repeatedly dispatching field resources to multiple apparent work orders caused by the same physical fault, the system needs to perform spatial-level aggregation processing.

[0068] The logic operation module internally establishes a directed graph model that maps to the actual physical topology of the pipeline network. Among them, the vertex set For various physical nodes such as valves, heat exchangers, and water distributors, the side assembly The system uses physical pipes connecting the vertices, with edges aligned with the designed fluid flow direction. After identifying the source node to be repaired, the logic module performs a breadth-first traversal with that node as the root of the directed graph. During the traversal, the system, in conjunction with the hydraulic and thermal simulation module, extracts the theoretical attenuation gradient of the pressure drop or heat loss at that node along the distance in the pipe network. When the abnormal fluid conduction attenuation on a certain traversal path falls below the system's preset perceptible impact threshold, the traversal of that branch is stopped. This perceptible impact threshold is typically characterized by the critical physical quantity at which the radiator at the fluid's end cannot maintain the designed room temperature. In specific applications, it can be set as a water supply temperature attenuation of 2°C to 3°C, or a drop in available pressure head of 10 kPa to 15 kPa. By setting the above quantitative boundaries, it is ensured that the traversal impact range will not diverge indefinitely within the topology model. All traversed and unblocked downstream nodes constitute a subset of the fluid dynamics impact range of the source fault. .

[0069] After constructing the physical impact area, the scheduling engine retrieves the spatial coordinates of each repair point to be assigned. System call spatial projection function This coordinate is mapped to the service node in the directed graph model that is closest to it by Euclidean distance and whose physical home radius is less than the set maximum (e.g., 500 meters). If the nearest distance of the projection exceeds the assigned radius, the system will treat the work order as abnormal data caused by coordinate entry errors and temporarily suspend it to prevent cross-regional erroneous aggregation. For the underlying algorithm matching process of this spatial projection and nearest neighbor search, those skilled in the art can use R-tree spatial index or KD-tree data structure for implementation. The basic retrieval logic is well-known in the field and will not be elaborated here.

[0070] Ultimately, the scheduling engine establishes an aggregated fault group. The mathematical expression for its aggregation condition is: In the formula, For independent repair work orders in the customer service module ; The graph model service node mapped to this work order; This refers to a subset of the fluid dynamics impact range calculated above. Through this step, the system identifies and separates the independent business attributes of the work order, binding it as a whole to the single physical source that triggered these requests.

[0071] In this embodiment, after identifying the source node to be repaired and generating the corresponding aggregated fault group, the scheduling engine needs to further combine the dynamic state set of the maintenance team to extract and quantify evaluation parameters in both the time and economic value dimensions, providing basic data for subsequent global optimization. This process specifically includes the following sub-steps:

[0072] S301, the scheduling engine combines field resource coordinates and urban road network data to quantify the time cost parameters for each available maintenance team to reach the target node. In the scheduling evaluation system, time cost includes not only the physical travel time but also the team's current downtime and on-site handling time. Regarding the first [item / item] within the system... The number of available maintenance teams and the identified first One source node to be repaired, and the time and cost parameters for building the logic operation module. The calculation formula is as follows: In the formula, This value is zero if the shift is idle, to estimate the remaining time. For the work team, from the current real-time coordinates Drive to the coordinates of the node to be repaired at the source The estimated travel time; The standard operation time preset for a specific fault type at this source node is usually obtained by looking up historical maintenance records in the system ledger. To prevent lookup failures due to new fault types not recorded in the ledger, the system sets a default operation time (e.g., 120 minutes) as... The minimum parameters.

[0073] As a preferred method, the estimated travel time is... Typically, the solution is obtained by calling the application programming interface (API) of an external geographic information system, inputting the coordinates of the origin and destination, and combining them with real-time traffic conditions. To prevent computational dead zones caused by timeouts in external road network API responses or network interruptions, the system incorporates a degradation calculation mechanism in the control layer. When the API call duration exceeds a set tolerance threshold (e.g., 2 seconds), the logic operation module automatically switches to local calculation mode. It calculates the Manhattan distance between the two coordinate points and divides it by the system's pre-stored average urban vehicle speed for that time period to estimate the backup travel time. Before performing this division operation, the system verifies the average vehicle speed data. If the vehicle speed approaches zero due to anomalies in the underlying data, it forces the use of a set minimum physical vehicle speed (e.g., 15 km / h) as the denominator, thereby ensuring the continuity and robustness of subsequent scheduling algorithms.

[0074] S302, the scheduling engine synchronously quantifies the value recovery parameters that can be generated by repairing the source node. The operation and maintenance of heating systems are highly economically sensitive; the impact of failures at different nodes on the overall network energy efficiency and user satisfaction varies. Therefore, the system expands scheduling evaluation from a simple time dimension to an economic value dimension. Regarding the first... One source node awaiting repair; logical operation module constructs value recovery parameters. This parameter is composed of two parts: physical heat loss recovery and business default avoidance, and its mathematical expression is: In the formula, To recover value from physical heat loss; To avoid value in business default penalties.

[0075] Value in recovering physical heat loss Its calculation is based on repairing the energy loss that the node can cut off. The logic operation module constructs a specific quantization sub-model: In the formula, The instantaneous heat loss rate is extracted by subtracting the abnormal heat flow and the theoretical normal heat flow at the node under deterioration conditions. To evaluate the time window, a standard scheduling assessment period (e.g., 2 to 4 hours) is typically set to estimate the heat loss that would continue to occur within that period without intervention. The pricing is based on the economic cost per unit of heat, which is usually fixed based on the cost of gas or coal consumption at the local heat source plant.

[0076] Value of avoiding business default penalties The system retrieves aggregated fault groups. Extract all independent repair orders within the group. Based on the service level agreement weighting parameters of each order, calculate the weighted total of penalties avoided from being paid to users due to timely repairs. The specific formula is as follows:

[0077] In the formula, For the first fault in the aggregated fault group Service level agreement weighting parameters for each repair work order; This serves as the system's preset base for late payment penalties. This base can be set based on the actual operating refund standards of the heating company, typically fluctuating between tens and hundreds of RMB. If, in the early stages of a physical anomaly in the pipeline network, the dispatch engine proactively detects it before the customer service module receives a user's repair request, then the fault group is aggregated. This is an empty set. Under this condition, the summation term in the above formula naturally becomes zero. Assigning a value of zero, the evaluation of the entire value recovery parameter will primarily be based on the physical level. This additive operation structure not only helps to balance network hardware protection and user service experience to a certain extent, but also reduces the risk of algorithm logic dead zones in preventive maintenance scenarios by handling empty set states, providing an objective numerical benchmark for quantifying the overall urgency of fault repair.

[0078] In this embodiment, after quantifying and obtaining the time cost parameters and value recovery parameters of each available maintenance team, the logic operation module needs to further construct a cross-dimensional comprehensive evaluation index and perform a global matching operation for task allocation based on this index. This process specifically includes the following sub-steps:

[0079] S303, the logic operation module combines value recovery parameters and time cost parameters to calculate the maintenance benefit-time ratio for each work team for each repair node. After obtaining independent physical time and business value parameters, the system needs to establish a unified metric that can balance emergency repair response speed and economic recovery benefits. The logic operation module targets the... The available maintenance team and the first The combination of several source nodes awaiting repair can improve the efficiency and time-saving of construction and maintenance. Its specific mathematical expression is: In the formula, Parameters for recovering the value of the source node; This is a time cost parameter. This ratio parameter represents the comprehensive recovery value that a specific work group can generate within a unit of time cost. By constructing this ratio parameter, the system effectively reduces the dimensionality of the original multi-objective scheduling evaluation process, which required simultaneously considering the shortest time consumption and the greatest recovery value, into a maximum value optimization problem for a single indicator. Considering that extreme network latency or abnormal underlying state parameters may cause the calculated time cost to approach zero, thus triggering anomalies in division-by-zero calculations, the logic operation module has a lower limit threshold check mechanism before performing division operations. If the parameter... If the time consumed is less than the minimum effective time set by the system (e.g., 10 minutes), the system will force the use of the minimum effective time as the denominator to prevent the risk of algorithm logic collapse while maintaining the engineering rationality of the evaluation indicators.

[0080] S304, the scheduling engine constructs an integer programming model based on the maintenance benefit-time ratio matrix, performs global optimization, and generates scheduling instructions. After completing the combined evaluation of all available resources and nodes to be repaired, the system integrates the discrete evaluation indicators into a global allocation matrix. To maximize overall operational efficiency, the logic operation module constructs an integer linear programming model with the sum of global maintenance benefit-time ratios as the optimization objective, and its objective function is expressed as: In the formula, The overall efficiency-time ratio score of the global scheduling scheme; Let be a Boolean decision variable. When the system assigns the first... The maintenance team went to the first When a source node awaiting repair executes its task... The value is 1 if the condition is met, and 0 otherwise. To conform to the physical constraints of actual field operations, boundary conditions are applied to the model simultaneously, i.e., limitations are imposed. as well as By setting the above constraint equations, it is helpful to ensure that within a single scheduling cycle, a work team is assigned at most one repair task, and a faulty node is handled by at most one work team. For any remaining nodes to be repaired that cannot be assigned to an available work team within this cycle due to resource constraints, the system will mark their status as suspended and automatically push them into the preferred waiting queue for the next scheduling cycle. If the suspension time of a specific task exceeds the set safety tolerance limit, the scheduling engine will trigger a cross-regional field support request to the upper-level application.

[0081] For the optimization process of solving the above integer linear programming model, those skilled in the art can use the conventional branch and bound method or call a commercial optimization solver for iterative calculations. The underlying matrix dimensionality reduction and optimization convergence principles are well-known technologies in this field and will not be elaborated upon here. As a preferred approach, to prevent the global optimization algorithm from entering a long-term solution stagnation state due to a surge in the number of nodes to be repaired under extreme disaster conditions, the system deploys a timeout blocking and algorithm degradation mechanism at the control layer. When the solver's single calculation time exceeds the set maximum tolerance time (usually set to 3 to 5 seconds), the logic operation module will automatically interrupt the current planning solution and downgrade to a heuristic greedy matching algorithm. The downgraded execution logic will prioritize extracting and locking the largest value in the matrix. The corresponding task allocation pairs are then iteratively removed from already allocated rows and columns until resources are exhausted. This is to avoid the occurrence of multiple equal maximum values ​​during the greedy matching process. Numerical values ​​can cause the algorithm to get stuck in a random selection dead zone, so the system introduces a two-level sorting rule. Specifically, when encountering a tie with the same ratio, the scheduling engine will prioritize extracting the value recovery parameter. Tasks with larger absolute values ​​are locked. Finally, the scheduling engine generates structured field dispatch instructions based on the optimized allocation matrix obtained from the solution, and pushes them down to the mobile terminals of the corresponding maintenance teams in real time through the communication interface, thereby realizing the coordinated connection between the underlying physical pipeline network perception and the upper-level operation and maintenance dispatch instructions.

[0082] In this embodiment, after the scheduling engine completes global optimization and issues scheduling instructions, considering that the pipeline fluid system is still in a state of continuous physical deterioration before the maintenance team arrives on site, the logic operation module needs to further perform deduction calculations of the evolution of the pipeline physical state to prevent secondary safety accidents. This process specifically includes the following sub-steps:

[0083] S401, the logic operation module extracts the estimated time of the assigned tasks and injects it back into the hydraulic and thermal simulation calculation module to perform time-domain extrapolation of the pipeline network's physical state. After obtaining the time cost parameters of the matching shifts, to prevent further deterioration of the pipeline network's state during shift commutes, the system uses this time parameter as the upper limit of the integration domain. Combined with the current degradation rate of the fault source node, a dynamic dimensionality-reduced extrapolation model for the target area is constructed. Taking the hydraulic pressure field, which is critical in pipeline network operation, as an example, the predicted pressure value of the target node at the estimated arrival time... The calculation expression is set as follows: In the formula, For the current scheduling time target node The actual pressure value collected; For time cost parameters; For pressure decay dynamic gradient function This is the local standard atmospheric pressure constant. The integral term in the formula quantifies the total pressure loss that is expected to occur in the pipeline during the maintenance team's commuting and waiting periods.

[0084] As a preferred approach, the pressure decay dynamic gradient function Typically, the system extracts the actual collected pressure sequence within the previous sliding time window (e.g., 15 to 30 minutes) and extrapolates it using linear or exponential fitting with the least squares method. To prevent the predicted pressure from having a negative dead zone that violates the laws of fluid mechanics due to excessively large integral calculation results, the outermost layer of the model contains a maximum value function, which forcibly constrains the lower limit of the theoretical extrapolation to the atmospheric pressure level.

[0085] To prevent excessively long estimated time due to extreme road conditions, which could lead to computational dead zones and boundary divergence in the simulation model, the system incorporates a time truncation mechanism before performing the aforementioned integral derivation. If the input time cost parameter exceeds the system's set effective derivation horizon (e.g., 120 minutes), the upper limit of integration is forcibly clamped to that effective boundary, thereby ensuring the stability of the numerical solution. For the numerical solution process of the dynamic equations involved in the hydraulic and thermal simulation module, those skilled in the art can employ conventional Runge-Kutta methods or implicit difference methods. The underlying iterative approximation mechanisms are well-known technologies in the field and will not be elaborated upon here.

[0086] S402, the logic operation module compares the predicted pipeline state parameters with the system's preset safety baseline and generates an automated physical containment command when an over-limit condition is triggered. If localized water loss or pressure drop in the heating pipeline is allowed to develop unchecked, it can lead to systemic operational disruptions such as circulating water pump vaporization or large-scale air intrusion. Therefore, the logic operation module establishes active containment judgment conditions, the mathematical logic of which is as follows: In the formula, The minimum safe operating threshold for maintaining normal fluid circulation in the area where the target node is located; To prevent accidental activation, a protection margin is provided. Minimum safe operating threshold. It is not arbitrarily designated, but rather calculated rigorously based on the critical saturated vapor pressure at the highest elevation within the pipeline network area where the fluid does not undergo phase change vaporization, or in conjunction with the necessary net positive suction head (NPSH) of the system's main circulating water pump. (Anti-maloperation protection margin) The calibration is usually performed by combining the measurement accuracy range and data drift characteristics of the underlying pressure sensor. In conventional urban heating systems, the value range is generally set to 5 kPa to 10 kPa.

[0087] When the above conditions are met, it indicates that the physical condition of the pipeline network will likely fall below the safe operating threshold before the maintenance team arrives on site. At this point, the dispatch engine no longer relies solely on delayed manual intervention, but instead immediately triggers the underlying automated control protocol, issuing a containment intervention command to the nearest remotely operated electric regulating valve group or distributed variable frequency pump to the faulty node. The specific containment intervention action typically does not employ a transient full shutdown operation that could easily cause strong pipeline oscillations, but rather executes cascade pressure reduction or bypass diversion logic based on the subset of the hydrodynamic influence range generated by the source node in S204.

[0088] In this embodiment, after determining that an automated physical restraint command needs to be triggered, the scheduling engine does not directly output a fixed switching signal. Instead, it delegates the task to the logic operation module, which combines the real-time hydraulic characteristics of the pipeline network to calculate the continuous dynamic adjustment step size and executes closed-loop control with self-correcting feedback. This process specifically includes the following sub-steps:

[0089] S403, the logic operation module constructs a hydraulic buffer constraint model based on the real-time operating status of the pipeline network, and calculates the dynamic adjustment step size of the target actuator. In the actual control scenario of the heating pipeline network, sudden and large changes in valve opening or pump frequency can easily cause a drastic redistribution of fluid momentum, thereby impacting weak links in the pipeline. To balance containment efficiency and physical safety, the system constructs a nonlinear adjustment model with a volatility penalty mechanism. The logic operation module calculates the target opening command for the next control cycle for the remotely operated electric regulating valve upstream of the target node. The specific algebraic operation model is as follows: In the formula, For the current moment The actual physical opening percentage of the valve; The set control execution cycle is typically set to 2 to 5 seconds; The basic adjustment step size coefficient is generally set to a range of 1% to 5%; This represents the current actual data collection pressure; This represents the minimum safe operating threshold. This is a smoothing scaling factor used to adjust the output sensitivity of the hyperbolic tangent function. In conventional water pressure regulation scenarios, its reasonable value range is usually set to 0.05MPa to 0.15MPa. This is the hydraulic buffer constraint function. Utilizing the range characteristics of the hyperbolic tangent function tanh, when the actual pressure is much higher than the safety threshold, the output tends to saturate to achieve rapid pressure reduction; while when the actual pressure approaches the safety threshold, the adjustment step size will decrease smoothly, thereby preventing pressure overshoot.

[0090] The calculation logic for dynamically suppressing and adjusting the pressure range during drastic fluctuations in the pipeline network is as follows: In the formula, The actual rate of change of pressure at the current moment is usually obtained by extracting the absolute value of the pressure difference within three adjacent control cycles; This is the maximum allowable safe pressure fluctuation rate threshold for the pipe network material. As a preferred approach, this threshold needs to be determined in conjunction with the service life of the pipe network and the allowable stress of the pipe material. In conventional steel heating pipe networks, it is usually set between 0.1 MPa / s and 0.2 MPa / s.

[0091] By introducing a hydraulic buffer constraint function, when the pressure fluctuation in the pipeline network approaches the safety limit... The value will approach zero, thus forcibly stopping the valve's operation. To prevent high-frequency electromagnetic interference noise collected by the underlying sensors from causing artificially high differential calculation results, which could lead to a logical dead zone where the regulation algorithm is falsely locked for a long time, the system incorporates a first-order low-pass filter preprocessing step before performing the above volatility calculation. Specifically, the system uses a recursive equation... The raw acquired values ​​are smoothed, where the filter coefficients are... The sensor sampling frequency is typically between 0.1 and 0.3.

[0092] In S404, the scheduling engine issues a sequence of physical actions based on the solved control commands and establishes a compensation and degradation mechanism for deviations in the actuator's state. After acquiring continuous dynamic target opening commands, the system needs to convert the calculation results from the digital layer into mechanical displacements at the physical layer. The scheduling engine encapsulates the target commands into standard communication messages through an IoT gateway and pushes them down to the corresponding programmable logic controller or remote terminal unit. For the conventional communication addressing process of the underlying control algorithm or controller, those skilled in the art can implement it using industrial Ethernet or standard fieldbus protocols. The basic command encapsulation and data handshake mechanism are well-known technologies in the field and will not be elaborated here.

[0093] Long-term pipeline operation often results in valve stem scaling or motor aging, which can easily lead to a disconnect between the actual response of the actuator and the theoretical command. To address this objective physical obstacle, the logic operation module establishes a command tracking and self-correcting feedback loop. Within a set delay window after the command is issued, the system continuously reads the actual position measurement value fed back by the valve. When a persistent gap exists between the actual displacement and the target command exceeding the system's set mechanical tolerance dead zone (e.g., opening deviation exceeding 3%), the system does not employ the conventional logic of continuously increasing the drive current for forced adjustment. Instead, it directly activates the abnormal condition degradation contingency plan. At this point, the scheduling engine determines that the main control valve has a risk of mechanical jamming and automatically switches the control target to the upstream secondary control valve or bypass pressure relief valve pre-associated in the directed graph topology model. If it further detects that the secondary actuator still times out or has reached its adjustment limit, the system triggers a final safety measure: the distributed variable frequency pumps in the faulty area perform a stepped frequency reduction operation, and a maintenance work order is dispatched to the maintenance team responsible for that area. This closed-loop verification mechanism, from theoretical step size constraints to on-site physical feedback, helps mitigate the risk of the entire hydraulic containment strategy stalling due to the mechanical failure of a single control node, providing underlying redundancy for the continuous and reliable operation of the system under complex hardware degradation conditions.

[0094] In this embodiment, after the maintenance team completes the emergency repair work on the target node, the system needs to perform multi-dimensional status verification to confirm that the physical fault has been effectively handled and smoothly release the previous automated physical restraint commands. This process specifically includes the following sub-steps:

[0095] S501, the logic operation module initiates pressure holding verification based on the field feedback signal, quantifying the hydraulic tightness of the repair area. After receiving the repair completion signal from the maintenance team via mobile terminal, the scheduling engine does not immediately restore the normal operation of the pipeline network, but instead coordinates with the underlying control network to enter a local pressure holding observation state. For the source node to be repaired, the logic operation module extracts the repair completion time. Calculate the pressure drop gradient of the pipeline network based on the actual pressure time series up to the end of the pressure holding verification window. Its specific algebraic expression is: In the formula, The steady-state pressure of the pipeline network was collected at the initial moment of completion; The acquisition pressure at the end of the pressure holding verification window; This is the preset pressure holding time. As a preferred method, the pressure holding time is usually dynamically matched based on the pipe diameter and design operating pressure, and is generally set to 15 to 30 minutes in conventional heating branch networks. The logic operation module compares the calculated pressure drop gradient with the allowable leakage rate threshold pre-stored in the system ledger. A comparison is performed. This permissible leakage rate threshold is typically set between 0.01 MPa / h and 0.05 MPa / h, based on the relevant provisions in the "Code for Construction and Acceptance of Urban Heating Pipeline Network Engineering" and the aging coefficient of the pipeline network. If the actual reduction gradient... If the physical repair is deemed satisfactory, the system confirms from the business logic level that the source of the fault has been eliminated.

[0096] To prevent the verification algorithm from stalling due to lack of pressure input caused by accidental activation during on-site repairs or the intrusion of high-temperature moisture, the system employs a bypass cross-validation mechanism based on quality conservation at the data link layer. When the pressure sequence validity verification of the target node fails, the logic operation module automatically retrieves data from the ultrasonic flowmeters of adjacent upstream and downstream nodes in the area, calculating the instantaneous flow difference between the inflow and outflow. If this difference is lower than the set lower limit of instrument measurement tolerance (e.g., 1% to 3% of the overall range), it is considered a successful repair verification. Furthermore, if the flow cross-validation link is also unavailable, the system will downgrade to a manual verification mode, requiring the field team to upload photos of mechanical pressure gauge readings and on-site confirmation signatures via mobile terminals, serving as a backup physical proof to break the verification process lockout. For the data acquisition and digital filtering process of the underlying flowmeters, those skilled in the art can implement it using conventional moving average or median filtering algorithms. The basic data cleaning logic is well-known in the field and will not be elaborated here.

[0097] In S502, the scheduling engine constructs a state recovery simulation model based on the repair confirmation results and executes a smooth release of automated physical restraints. After confirming that the physical repair has met the standards, the system needs to cancel the previously issued physical restraint commands and gradually restore the designed transport capacity of the pipeline network. To prevent hydraulic imbalances in the pipeline network and impacts on downstream equipment caused by transient and significant adjustments in valve openings, the logic operation module constructs an exponential dynamic unlocking curve to calculate the dynamic target opening of the target valve during the recovery phase. The specific mathematical model is as follows: In the formula, This represents the percentage of pressure drop opening maintained by the system under restraint conditions. This is the design reference opening for this node under normal heating conditions; The initial trigger moment for the containment release; This represents the recovery time constant. In the initial recovery phase, when the pipeline pressure differential is large, this model can provide relatively rapid opening adjustment; however, as the actual opening gradually approaches the normal reference, the adjustment rate decays exponentially, thus meeting the buffering requirements of fluid inertia. The aforementioned recovery time constant... The value is not a fixed configuration, but is dynamically calculated by the logic operation module in combination with the fluid volume of the pipe section and the rated flow of the downstream user. It is usually limited to the range of 30 seconds to 120 seconds, which helps to ensure that the acceleration of the flow increment in the pipeline network is always lower than the hydraulic impact limit of the pipe material.

[0098] As the actuators gradually open according to the dynamic unlocking curve, the logic operation module simultaneously performs real-time monitoring of the hydraulic conditions of the entire network. If, during the network restoration process, the pressure fluctuation rate of a local network, as reported by the underlying sensors, touches or exceeds the maximum safe pressure fluctuation rate threshold again, the scheduling engine will immediately trigger the control suspension protection mechanism. Under this mechanism, the system forcibly suspends the valve's continued opening command, maintaining the current actual opening until the network hydraulic parameters return to the steady-state range. To prevent the hydraulic parameters from failing to return to normal due to potential hidden secondary leaks in the network, thus causing the system to remain in a suspended half-open state for an extended period, the logic operation module is configured with a timeout wake-up timer. If the suspension period exceeds the set safe waiting limit (e.g., 10 minutes), the system will automatically terminate the automatic unlocking process, maintain the current safe opening, and send a manual intervention warning to the upper-level scheduling center. This control strategy, which couples on-site physical seal verification with exponential smooth curve recovery, helps to achieve a smooth transition from abnormal containment conditions to normal heating conditions, reducing the risk of secondary water pipe rupture damage during the system restart phase at the underlying mechanism level.

[0099] In this embodiment, after confirming that the physical state of the pipeline network has been stably restored, the scheduling engine needs to map the improvement of the underlying physical parameters to the upper-layer business system, completing the closed loop of the previously aggregated repair work order status. This process specifically includes the following sub-steps:

[0100] S503, the logic operation module performs business status decoupling and closed-loop verification of aggregated fault groups based on the dynamic changes of terminal thermal parameters. The previously generated aggregated fault groups contain multiple independent user repair work orders based on spatial and fluid topology associations. The completion of physical source repair does not mean that the heating quality of all associated terminals will synchronously return to standard operating conditions. To reasonably determine the actual resolution status of each work order, the logic operation module extracts the first fault order from the group... A thermal recovery verification model is constructed based on the terminal heating parameters bound to each user's work order. The specific algebraic verification logic is set as follows: In the formula, This refers to the user's thermal recovery efficiency index; To be completed by the time of physical restoration Starting from this point, the actual indoor temperature or secondary network return water temperature of the user is extracted. The actual temperature at the current verification time; This refers to the standard heating temperature set based on the user's service level agreement. Considering that in real-world scenarios, if the user's actual collected temperature has already reached or exceeded the standard heating temperature at the time of physical repair completion (i.e., ...), ... At this point, the denominator in the formula will approach zero or become negative, leading to a division-by-zero error. Therefore, the system includes a pre-processing state check logic before substituting the above formula, directly assigning a performance index to handle such cases. Maximum value: 1.

[0101] The logic operation module will calculate the result. With the set performance target threshold A comparison is then performed. As a preferred approach, this threshold is typically set between 0.8 and 0.9, in conjunction with human thermal comfort perception standards. If... If the physical fault corresponding to the work order has been substantially resolved, the system will decouple it from the aggregated fault group and mark it as pending archiving.

[0102] To prevent verification from stalling due to battery depletion or offline status of individual user indoor temperature monitoring nodes causing data breakpoints, the system employs a spatial proxy mechanism in the data link. When target node data is missing for more than a preset time window, the logic operation module automatically addresses the temperature sequence of adjacent nodes on the same floor or within the same heating branch as a substitute parameter in the performance calculation, thereby ensuring the integrity of the decoupled logic.

[0103] In S504, the scheduling engine performs service level agreement (SLA) settlement for the decoupled work orders and generates structured business archive data. For verified work orders, the system needs to further quantify the service quality of this maintenance event. The logic operation module combines the timestamps within the work order lifecycle to calculate the actual business response time and physical repair time, and retrieves the SLA weight parameters to perform quantitative calculations for penalty waivers or compensation. After completing the calculation, the scheduling engine integrates the fault waveforms of the underlying sensors, the on-site maintenance logs of the maintenance team, and the thermal recovery verification results, and encapsulates them into a standard-format maintenance work order voucher. For the asynchronous communication and persistent storage of the above work order data in the enterprise service bus or distributed database, those skilled in the art can implement it using conventional microservice architecture data processing techniques. The basic concurrency control and transaction consistency guarantee mechanisms are well-known technologies in the field and will not be elaborated here.

[0104] To generate data iterations for operational experience, the scheduling engine extracts key operational characteristic data that triggered the aggregated alarm (such as the material of the faulty pipe section, initial leakage pressure drop rate, repair time, and material consumption) while archiving business work orders, and updates it to the system's internal equipment operation log. If, during the decoupling of the aggregated fault group, some work orders consistently have a thermal recovery efficiency index below the target threshold within the set maximum verification period (e.g., 4 to 6 hours after repair), the scheduling engine will determine that these work orders involve independent secondary faults or deep-seated local hydraulic imbalance problems. For such isolated work orders, the system will automatically generate a secondary dispatch instruction, re-importing them as independent events into the initial anomaly monitoring and matching process. This closed-loop verification and anomaly stripping mechanism helps avoid overlooking secondary hazards caused by the superposition of primary and secondary faults, providing underlying continuous support for the quality of operation and maintenance throughout the entire lifecycle of the heating network.

[0105] Specific application examples:

[0106] Suppose that during the peak winter heating season in a large northern city (outdoor temperature -12℃), a sudden physical rupture occurs in the #4 branch pipe of the secondary heating network in Xingfu Community, resulting in hot water leakage.

[0107] Step-by-step implementation simulation:

[0108] At 14:00, the system retrieved data every 3 seconds. The inlet and outlet water pressure of physical monitoring node #4 suddenly dropped from the normal 0.60MPa to 0.48MPa. Simultaneously, the customer service module received repair requests from 12 households reporting cold heating systems, and each repair work order... The urgency level is high, so the system assigns its Service Level Agreement weight parameter to it. Automatically assigned the highest level 5.0.

[0109] The logic operation module extracts the actual operating parameters of pipe section #4. The current actual pressure drop deviates from the theoretical value by 300%, and the actual temperature drop deviates by 20%. The system extracts the calibrated weighting coefficients (thermal). Hydraulics Coupling amplification factor Substituting these values ​​into the thermo-hydraulic coupling degradation formula, extreme values ​​are calculated:

[0110] ;

[0111] Substituting the above deviation ratio, the current time can be calculated. Simultaneous dynamic baseline of the system The calculated value is .because The system determined that node #4 is the source node requiring repair. (See attached document.) Figure 3 The system clusters 12 scattered work orders within the radius into one aggregated fault group.

[0112] The system detected three available work groups nearby (A, B, and C). The system first considered the base penalty amount (…). (in yuan) and heat loss value estimation, substitute into the value recovery parameter formula Calculate the total recoverable value of the fault. Yuan. Next, the system calculates the maintenance efficiency time ratio for the three work groups. (i.e., the economic value that can be recovered per unit of time):

[0113] ;

[0114] Group A: Minutes, substituting into the formula, yields... Yuan / minute.

[0115] Group B: Minutes, substituting into the formula, yields... Yuan / minute.

[0116] Class C: Minutes, substituting into the formula, yields... Yuan / minute. See attached document. Figure 5 The quantitative demonstration shows that although the individual time costs of each shift differ, the system locks in the time cost based on the global integer programming model. This is the maximum point. Based on this, the scheduling engine makes the optimal dispatch decision, instantly assigning the work order to the B shift team with the highest overall efficiency.

[0117] System integral simulations indicate that the pressure will drop below 25 minutes later. The safety baseline. The scheduling engine issues a containment command to the upstream valve of node #4. To prevent water hammer caused by instantaneous shut-off, the system calls the nonlinear adjustment step size formula (assuming a base step size). scaling factor ):

[0118] With actual pressure As the pressure gradually approaches the bottom line of 0.30 MPa, the output value of the hyperbolic tangent function tanh gradually and smoothly decreases from 1 to 0. The valve opening gently hovers at 35% from 100% in an S-shaped curve, firmly holding the pipeline pressure at 0.35 MPa (retaining a physical buffer margin of 0.05 MPa).

[0119] After Team B completes the work, refer to the attached document. Figure 4 The valve recovered according to the exponential curve. One hour later, the system retrieved a data from a user at the end of pipe section #4. The room temperature data. The basic room temperature of this household. It has dropped to Target standard temperature Current measured temperature rebounded to Substitute into the thermal recovery verification model: Since the calculation result is 0.917 With the set threshold of 0.9, the system determines that the user's thermal characteristics have been substantially restored, automatically decouples and archives the user from the aggregated fault group, and the entire process is conducted without human intervention.

[0120] Experimental verification and effect comparison:

[0121] To verify the effectiveness of the system of the present invention, the research and development team conducted a comparative experiment for one heating season (120 days) on the heating network (covering an area of ​​5 million square meters and including 85 heat exchange stations).

[0122] Control group: The traditional SCADA manual monitoring alarm and customer service manual dispatch mode was used (previous historical data).

[0123] Experimental group: Fully deploy the heating system operation, maintenance and management system of this invention.

[0124] 1. Comparison Table of Core Indicators

[0125] Evaluation indicators Traditional monitoring and dispatching model (control group) The system of this invention (experimental group) Optimization range Average time to isolation and intervention for sudden failures 35.4 minutes (depending on manual judgment and manual valve closure) 4.2 minutes (automatic simulation triggers hydraulic containment) Shortened by 88.1% Average total field response time per fault 68.5 minutes (dispatching orders based on the order of reported repairs, prioritizing the nearest location). 42.3 minutes (optimization based on maintenance benefit-time ratio) Shortened by 38.2% Invalid dispatch rate due to same source fault 18.5% (Multiple repair teams dispatched to the same leak point) 0% (Spatial clustering generates aggregated fault groups) Reduce by 100% Number of times the water pump was damaged due to water loss / cavitation 7 times / heating season 0 times / heating season (pre-installed dynamic hydraulic containment protection) eliminate Compensation for breach of contract due to delayed emergency repairs during the heating season Approximately 245,000 yuan Approximately 32,000 yuan (introducing value recovery parameter guidance) Reduced by 86.9%

[0126] 2. Conclusion of Effect Analysis

[0127] To prevent redundant scheduling: In the control group, due to the temperature drop in the area caused by hydraulic imbalance, customer service would generate dozens of scattered orders and distribute them to different work groups. This invention, through directed graph topological clustering in step S200, traces the apparent low room temperature back to the physical cause of pipeline leakage, achieving one order per source and significantly saving human resources.

[0128] Balancing timeliness and economy: Traditional algorithms only consider who is closest and who gets the job. The innovative maintenance benefit-time ratio mechanism in step S300 of this invention ensures that high-value faults (those with high thermal loss and impact on high-priority customers) can overcome physical distance limitations and receive higher priority resource allocation.

[0129] The invention eliminates the dead zone of secondary hardware damage: traditional manual emergency valve closing is prone to water hammer effect, which can easily cause old pipelines to burst. The hydraulic buffer constraint (based on the tanh function) and exponential recovery curve introduced in steps S400 and S500 of this invention have not resulted in pressure peaks exceeding the allowable stress in actual measurements, thus achieving a soft landing.

Claims

1. A method for operation, maintenance and management of a heating system, characterized in that, include: Acquire sensor data from the physical monitoring nodes of the heating network, repair work order data generated by the customer service module, location coordinates of the mobile terminal belonging to the maintenance team, and the work status of the maintenance team; The degradation value of the pipeline section is calculated based on the sensor data. When the degradation value exceeds the set baseline, the source node to be repaired is marked, and an aggregated fault group is generated by combining the repair work order data. For the aggregated fault group, and in combination with the location coordinates and the operation status, calculate the time cost parameters and value recovery parameters of each available maintenance team, and calculate the maintenance benefit-time ratio based on the time cost parameters and value recovery parameters, so as to find the optimal matching team. Extract the estimated arrival time from the time cost parameters of the optimal matching shift, and based on the estimated arrival time, perform a physical throttling operation on the electric regulating valve upstream of the source node to be repaired when the triggering condition is met; After receiving a repair confirmation signal from the mobile terminal, the physical throttling operation is terminated, and the aggregated fault group is disbanded when the degradation value of the source node to be repaired falls back to within a preset range.

2. The operation, maintenance and management method for a heating system according to claim 1, characterized in that, Calculating the degree of deterioration of a pipeline section includes the following steps: Extract the actual physical parameters and theoretical operating parameters of the pipeline section; Based on the deviation between the actual physical parameters and the theoretical operating parameters, the degradation degree is calculated using a computational model that includes independent deviation terms and cross-coupling terms.

3. The operation, maintenance and management method for a heating system according to claim 1, characterized in that, For the aggregated fault group, and in conjunction with the location coordinates and the work status, the time cost parameters and value recovery parameters for each available maintenance team are calculated, including the following steps: The time cost parameter is derived by superimposing the estimated remaining time determined by the operation status, the estimated travel time calculated by the positioning coordinates, and the preset standard operation time. The value recovery parameter is derived by summing the physical heat loss recovery value calculated based on the sensor data and the business penalty avoidance value calculated based on the repair work order data included in the aggregated fault group.

4. The operation, maintenance and management method for a heating system according to claim 1, characterized in that, Based on the estimated arrival time, the determination that the triggering condition is met includes the following steps: Based on the estimated arrival time, predict the physical state parameters of the source node to be repaired at the time of the shift's arrival; The physical state parameters are compared with a preset minimum safe operating threshold. When the predicted value is lower than the minimum safe operating threshold, the triggering condition is determined to be met.

5. The operation, maintenance and management method for a heating system according to claim 4, characterized in that, The physical throttling operation includes the following steps: The hydraulic buffer constraint model is constructed by combining the difference between the current physical state parameters of the source node to be repaired determined based on the sensor data and the minimum safe operation threshold, the fluctuation rate of the pipeline pressure calculated based on the sensor data, and the real-time operation status of the pipeline extracted from the sensor data. The target opening command of the electric regulating valve is dynamically solved based on the hydraulic buffer constraint model.

6. The operation, maintenance and management method for a heating system according to claim 1, characterized in that, Before terminating the physical throttling operation, the following steps are also included: Upon receiving the repair confirmation signal, pressure holding verification is initiated, and the pressure drop gradient of the repair area within the preset pressure holding duration is calculated based on the sensor data. When the pressure drop gradient is not greater than the system's pre-stored allowable leakage rate threshold, the physical repair is deemed to have met the standard.

7. The operation, maintenance and management method for a heating system according to claim 1, characterized in that, Terminating the physical throttling operation includes the following steps: After determining that the physical repair has met the standards, the electric regulating valve is gradually restored to its standard opening degree according to the exponential dynamic unlocking curve.

8. The operation, maintenance and management method for a heating system according to claim 1, characterized in that, Dissolving the aggregated fault group includes the following steps: For each repair work order within the aggregated fault group, extract the end-point heating parameters bound to the repair work order and derived from the sensor data to calculate the thermal recovery efficiency index; When the thermal recovery efficiency index of a specific repair work order reaches a preset efficiency threshold, the specific repair work order is decoupled from the aggregated fault group.

9. The operation, maintenance and management method for a heating system according to claim 1, characterized in that, The steps for acquiring sensor data from the physical monitoring nodes of the heating network also include the following steps: The collected sensor data is timestamped, and data that fails to be timestamped is either zero-order preserved or linearly interpolated.

10. An operation and maintenance management system for a heating system, characterized in that, A method for operation, maintenance and management of a heating system according to any one of claims 1-9, comprising: On the operational side, there are sensor networks used to collect physical parameters of the heating network, and actuators equipped with electric regulating valves driven by programmable logic controllers. The information terminal includes a customer service module for receiving repair work order requests from users, a mobile terminal for maintenance teams, and a scheduling engine for performing data processing and logical operations. The communication interface is established between the information terminal and the operation terminal.

Images