Intelligent regulation and control device and system for multi-branch thermodynamic system

Abstract

The invention discloses an intelligent regulation and control device and system for a multi-branch thermodynamic system, particularly relates to the technical field of heat supply systems, and is used for solving the problem of hydraulic congestion caused by the fact that a heat source output instruction exceeds the conveying capacity of a pipe section. A heat source output instruction and pipe network pressure and flow data are obtained through a data acquisition module, a state deduction module deduces the hydraulic transport capacity based on pipe network topology analysis and state change collaboration, and a topology recognition module constructs parameter space point cloud and recognizes conflicting pipe sections through topology data analysis. The risk early warning module analyzes the flow path change influence and identifies the hidden congestion risk pipe section, the instruction reconstruction module recalculates the heat source output instruction by combining the conflict and risk pipe sections, and the instruction issuing module issues a new instruction to realize the safe and stable operation of the system.

Description

Technical Field

[0001] This invention relates to the field of heating system technology, and more specifically, to an intelligent control device and system for multi-branch heating systems. Background Technology

[0002] In large-scale thermal systems with multi-branch structures, the use of multi-heat source coupling for energy supply is a typical configuration for improving system reliability and operational economy. In existing technologies, such systems are typically built based on a cost-optimal load distribution model, and force commands are issued to each heat source through a central control platform to achieve coordination between heat sources. At the same time, the pipeline network side is regulated according to traditional hydraulic calculation models or preset operating curves to ensure the safe and stable heat transmission and distribution process.

[0003] However, existing technical solutions fail to fully coordinate the dynamic coupling relationship between optimization decisions and physical systems when achieving multi-heat source coordination and economic optimization. Because the load allocation model does not embed the hydraulic transport capacity constraints of the pipe network in real time, the issued heat source output commands may exceed the actual transport limits of some pipe sections under the current operating conditions, making it difficult to execute control commands at the physical level. This leads to the risk of local hydraulic congestion and overall control failure, limiting the further exploration of the optimization potential of complex thermal systems. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent control device and system for multi-branch thermal systems to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: An intelligent control system for multi-branch thermal systems includes: The data acquisition module is used to acquire in real time the output commands of each heat source in the multi-heat source system and the pressure and flow data of key nodes in the pipeline network; The state simulation module is used to analyze the coordinated state changes between related pipe segments based on pressure and flow data and the pipeline network topology, and to deduce the hydraulic transport capacity of each pipe segment. The topology identification module is used to construct a parametric spatial point cloud of flow demand and hydraulic transport capacity for each pipe segment, extract topological invariant features through topological data analysis algorithms, and identify conflicting pipe segments based on their persistence. The risk warning module is used to analyze the potential impact of changes in flow paths on related pipe sections during the control process, based on conflict pipe sections and pipe network topology, and to identify pipe sections with hidden congestion risks. The command reconfiguration module is used to recalculate the output commands of each heat source by combining conflict pipe sections and pipe sections with hidden congestion risk. The command issuing module is used to issue the recalculated output command to the corresponding heat source in the multi-heat source system.

[0006] Furthermore, real-time acquisition of output commands from each heat source in the multi-heat-source system and pressure and flow data from key nodes in the pipeline network, including: Receive output commands from each heat source from the central control platform; Pressure and flow data are collected by sensors deployed at key nodes of the pipeline network.

[0007] Furthermore, based on pressure and flow data, and considering the changes in the coordinated state between related pipe segments according to the pipeline network topology, the hydraulic transport capacity of each segment is deduced, including: Based on the pipeline network topology, pipe segments with hydraulic relationships are identified to form a group of associated pipe segments. Calculate the correlation coefficient of pressure gradient between adjacent pipe segments in a group of associated pipe segments and the synchronicity index of flow rate changes; Coordinated state assessment parameters are constructed based on the pressure gradient correlation coefficient and the flow rate change synchronicity index; The hydraulic transport capacity of each pipe section is inferred based on the comparison results between the collaborative status assessment parameters and historical normal operation data.

[0008] Furthermore, a parametric spatial point cloud of flow demand and hydraulic transport capacity for each pipe segment is constructed. Topological invariant features are extracted using topological data analysis algorithms, and conflicting pipe segments are identified based on their persistence, including: The flow demand and hydraulic transport capacity of each pipe segment are mapped to data points in a multi-dimensional parameter space to form a parameter space point cloud; Topological data analysis algorithms are applied to parameter space point clouds to calculate persistent homology groups and extract topological invariant features. The persistence measure of topological invariant features is analyzed. When the persistence is lower than a preset threshold, the corresponding pipe segment is identified as a conflicting pipe segment.

[0009] Furthermore, applying topological data analysis algorithms to parameter space point clouds to calculate persistent homology groups and extract topological invariant features includes: constructing a filtered complex sequence of parameter space point clouds and calculating persistent homology groups, and extracting Betti numbers from the persistent homology groups as topological invariant features.

[0010] Furthermore, based on conflicting pipe sections and network topology, the potential impact of flow path changes on related pipe sections during regulation is analyzed, and pipe sections with hidden congestion risks are identified, including: Identify the associated pipe segment group that has hydraulic connections with the conflicting pipe segment based on the pipe network topology; Changes in flow distribution within the associated pipe segment group after a change in flow path during simulation control process; Calculate the change in pressure gradient of the associated pipe segment group after the change in flow distribution; Hidden congestion risk segments were identified based on the comparison between the pressure gradient change and the historical normal pressure gradient range.

[0011] Furthermore, considering conflict pipe sections and pipe sections with hidden congestion risks, the output commands for each heat source are recalculated, including: Establish a load redistribution model with the constraints of eliminating conflicting pipe sections and pipe sections with hidden congestion risk, and with the optimization objective of minimizing the total output adjustment; The output adjustment priority of each heat source is determined based on the hydraulic sensitivity coefficient of each heat source and the conflict pipe section and the pipe section with hidden congestion risk. The load redistribution model is transformed into a sequential optimization problem based on the output adjustment priority, and then solved sequentially. New output commands for each heat source are generated based on the solution results of the sequence optimization problem.

[0012] Furthermore, the load redistribution model is transformed into a sequential optimization problem according to the output adjustment priority and solved sequentially. This includes: decomposing the load redistribution model into a sequence of subproblems according to the output adjustment priority, solving the output adjustment amount of each heat source in order of priority, and fixing the adjustment results of the heat sources that have been solved in the solution process.

[0013] Furthermore, the recalculated output command is sent to the corresponding heat sources in the multi-heat-source system, including: Verify the compatibility between the recalculated output command and the current operating status of each heat source; The priority sequence for issuing instructions is determined based on the hydraulic distance between each heat source and the conflicting pipe section in the pipeline network topology. According to the priority sequence of the instructions, the recalculated output instructions are sent to the corresponding heat sources through a secure communication protocol. Receive confirmation feedback from each heat source regarding the output command and record the command execution status.

[0014] On the other hand, the present invention provides an intelligent control device for multi-branch thermal systems, the device comprising an intelligent control system for multi-branch thermal systems.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. By constructing a complete technology chain encompassing data acquisition, state simulation, topology identification, risk warning, command reconstruction, and command issuance, the system achieves deep coupling between the physical operating state of the thermal system and optimized control strategies. First, the data acquisition module acquires real-time heat source output commands and pipeline hydraulic data, providing an accurate input foundation for subsequent analysis. The state simulation module analyzes the coordinated state changes between related pipe segments based on the pipeline topology, dynamically predicting the hydraulic transport capacity of each segment, overcoming the limitations of traditional static hydraulic calculation models. The topology identification module constructs a parameter space point cloud of flow demand and transport capacity. A topological data analysis algorithm is used to extract topological invariant features that characterize the system's operating state, enabling an innovative method for identifying potential conflict pipe sections from complex data. The risk warning module further analyzes the potential impact of changes in flow paths during regulation on associated pipe sections, identifying latent congestion risks that are difficult to detect using traditional methods. The command reconstruction module integrates information from conflict pipe sections and latent congestion risk pipe sections to recalculate the output commands of each heat source, ensuring that the regulation commands meet both economic requirements and the physical constraints of the pipeline network. Finally, the command issuance module safely and reliably transmits the verified commands to each heat source, forming a complete intelligent regulation closed loop.

[0016] 2. Regarding system safety, real-time state simulation and topology feature identification enable early detection of local congestion risks that are difficult to detect using traditional methods, effectively preventing system failures caused by control commands exceeding the pipe segment's transport capacity. In terms of operational economy, based on precise physical state perception and risk warning, the optimization potential of multi-heat source collaboration can be fully explored while ensuring system safety, achieving more refined load allocation. Regarding control reliability, a full-process guarantee mechanism from state perception and risk identification to command reconfiguration has been established, significantly improving the success rate and adaptability of system control under complex operating conditions. In particular, this system innovatively applies topology data analysis methods to the assessment of thermal system operating states. By extracting topological invariants that characterize the global features of the system, it achieves a new approach to fundamentally understanding the system's operating state, providing a new technological paradigm for the intelligent control of complex thermal systems. Attached Figure Description

[0017] Figure 1 This is a flowchart of an intelligent control system for multi-branch thermal systems according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: Figure 1 A schematic diagram of an intelligent control system for multi-branch thermal systems according to the present invention is provided. The intelligent control system for multi-branch thermal systems includes the following modules: The data acquisition module is used to acquire in real time the output commands of each heat source in the multi-heat source system and the pressure and flow data of key nodes in the pipeline network; The state simulation module is used to analyze the coordinated state changes between related pipe segments based on pressure and flow data and the pipeline network topology, and to deduce the hydraulic transport capacity of each pipe segment. The topology identification module is used to construct a parametric spatial point cloud of flow demand and hydraulic transport capacity for each pipe segment, extract topological invariant features through topological data analysis algorithms, and identify conflicting pipe segments based on their persistence. The risk warning module is used to analyze the potential impact of changes in flow paths on related pipe sections during the control process, based on conflict pipe sections and pipe network topology, and to identify pipe sections with hidden congestion risks. The command reconfiguration module is used to recalculate the output commands of each heat source by combining conflict pipe sections and pipe sections with hidden congestion risk. The command issuing module is used to issue the recalculated output command to the corresponding heat source in the multi-heat source system.

[0020] Real-time acquisition of output commands from each heat source in a multi-heat-source system and pressure and flow data from key nodes in the pipeline network, specifically implemented as follows: The system acquires output commands from each heat source in a multi-heat source system and pressure and flow data from key nodes in the pipeline network in real time. Command acquisition is achieved by receiving output commands from each heat source from a central control platform. The central control platform establishes connections with each heat source control system using an interface based on a standard network communication protocol, such as TCP / IP. The frequency of receiving output commands is set according to the system's real-time requirements, for example, once per second, to ensure timely delivery. The data structure of the output command includes multiple fields, such as a heat source identifier to uniquely identify each heat source, output values ​​expressed in power units, such as megawatts, and a timestamp recording the precise time the command was generated. If a network anomaly is detected during reception, such as network interruption or delay, the system will initiate an automatic retry mechanism. The retry interval is set according to network conditions, for example, 5 seconds, and the maximum number of retries is set according to reliability requirements, for example, 3 times, to ensure the integrity and availability of command reception.

[0021] Pressure and flow data are collected by sensors deployed at key nodes in the pipeline network. The selection of key nodes is determined based on pipeline topology analysis, such as the location of branch points and confluence points, to ensure coverage of critical areas of hydraulic change. Pressure data is collected using piezoelectric pressure sensors, and flow data is collected using electromagnetic flowmeters. The deployment location of the sensors is optimized through hydraulic calculation models, such as based on flow distribution simulation results, to maximize data representativeness. During the acquisition process, the sampling frequency of pressure data is set according to dynamic response requirements, for example, 10 times per second, with the unit being megapascals (MPA). The sampling frequency of flow data is also set according to system requirements, for example, 10 times per second, with the unit being cubic meters per hour (m³ / h). Sensor data is transmitted to the data acquisition unit via a standard bus protocol, such as RS-485 bus. Error detection algorithms, such as CRC checksums, are applied during transmission to identify data errors. Erroneous data is automatically discarded and re-acquisition is triggered to ensure data quality.

[0022] Real-time integrity checks are performed on the collected pressure and flow data to ensure data continuity and completeness. Integrity checks include temporal continuity and numerical reasonableness checks. Temporal continuity is verified by checking if the timestamp interval between adjacent data points equals the preset sampling interval (e.g., 0.1 seconds). If the timestamp interval deviation exceeds the allowable threshold (e.g., 0.1 seconds), the data is considered discontinuous. Numerical reasonableness checks are performed by comparing data values ​​with preset reasonable ranges. These reasonable ranges are set based on the system's physical characteristics. For example, the reasonable range for pressure data is determined based on the system's design pressure, with a minimum of 0 MPa and a maximum of 2 MPa. The reasonable range for flow data is determined based on the pipeline network's design flow rate, with a minimum of 0 cubic meters per hour and a maximum of 100 cubic meters per hour. If a data value exceeds the reasonable range, the data is considered abnormal. Missing data is supplemented using interpolation methods, such as linear interpolation. The required historical data length for interpolation is set based on data stability (e.g., 10 sampling points). Integrity check results are recorded in real-time. Abnormal data is marked as invalid and triggers an alarm mechanism, such as through log recording or notification to operators.

[0023] The validated pressure data, flow data, and output commands are integrated into a unified dataset. The integration process aligns different data sources based on timestamps. The maximum allowed time deviation for timestamp alignment is set according to data synchronization requirements, for example, 0.05 seconds. If the time deviation exceeds this value, an interpolation method is used to adjust the data time point, such as nearest neighbor interpolation. The integrated dataset is stored in a structured format, containing fields such as heat source identifier, output value, pressure value, flow value, and timestamp. The dataset is managed through a storage system, such as a memory database for temporary storage. The storage period is set according to data processing requirements, for example, 1 hour, and then transferred to a persistent storage system. A data deduplication mechanism is applied during the integration process to avoid duplicate records and ensure the uniqueness and consistency of the dataset. At the same time, the data integration considers system scalability, such as supporting the dynamic addition of new data sources.

[0024] Based on pressure and flow data, and considering the pipeline network topology, the coordinated state changes among related pipe segments are analyzed to deduce the hydraulic transport capacity of each segment. The specific implementation is as follows: Based on pressure and flow data, the collaborative state changes between related pipe segments are analyzed according to the pipeline network topology, and the hydraulic transport capacity of each segment is deduced. Initial analysis is achieved by identifying hydraulically related pipe segments and forming related pipe segment groups based on the pipeline network topology. The pipeline network topology is represented by a graph theory model, where nodes represent pipe connection points and edges represent pipe segments. The determination of hydraulic association is based on the connection relationship of pipe segments in the topology and the direction of water flow. For example, directly connected pipe segments or pipe segments connected through no more than three intermediate nodes are considered to have hydraulic association. The scope of related pipe segment groups is set according to the system scale. For example, the number of pipe segments included in each related pipe segment group is controlled between 3 and 10 to ensure the operability of the analysis. The characteristics of the pipeline network branch structure are considered during the identification process. For example, continuous pipe segments on the same branch path are preferentially assigned to the same related pipe segment group. At the same time, the division of related pipe segment groups avoids overlap to ensure that each pipe segment belongs to only one related pipe segment group. The division criteria also include the physical characteristics of the pipe segments, such as the similarity of pipe diameter and length, to enhance the hydraulic consistency of pipe segments within the group.

[0025] The pressure gradient correlation coefficient and flow rate change synchronicity index between adjacent pipe segments in a group of associated pipe segments are calculated. The pressure gradient correlation coefficient is obtained by analyzing the correlation between the pressure data of adjacent pipe segments over time. The calculation uses a continuous time window sequence of pressure data, for example, a time window length of 10 minutes and a pressure data sampling interval of 1 second. The pressure gradient correlation coefficient is calculated using the statistical correlation coefficient method, by calculating the ratio of the product of the covariance and standard deviation of the pressure gradient values ​​of two adjacent pipe segments. The pressure gradient value is calculated as the ratio of the pressure difference between adjacent measurement points to the pipe segment length, with the unit being megapascals per meter. The flow rate change synchronicity index is obtained by comparing the consistency of the flow rate change trends of adjacent pipe segments. The calculation uses a flow rate data sequence within the same time window, by calculating the sign matching ratio of the flow rate change rates of two pipe segments. The flow rate change rate is calculated as the ratio of the difference between the current flow rate value and the previous flow rate value to the time interval, with the unit being cubic meters per hour per second. During the calculation process, the data sequence needs to be smoothed to eliminate noise, for example, by using a moving average method with a window size of 5 data points, to ensure the stability of the index calculation.

[0026] Coordination status assessment parameters are constructed based on the pressure gradient correlation coefficient and the flow change synchronicity index. These parameters are obtained by weighted combination of the pressure gradient correlation coefficient and the flow change synchronicity index. The weight allocation is determined based on the degree of influence of each index on system stability. For example, the weight of the pressure gradient correlation coefficient is set to 0.6, and the weight of the flow change synchronicity index is set to 0.4. The weight values ​​are determined through historical data analysis and optimization. For example, regression analysis is used to determine the relationship between each index and the system failure rate. The calculation of the coordination status assessment parameters adopts a linear weighted summation method to ensure that the parameter values ​​are between 0 and 1. The closer the value is to 1, the better the coordination status. During the calculation process, the input indicators are standardized. For example, the pressure gradient correlation coefficient and the flow change synchronicity index are normalized to the range of 0 to 1. The normalization method adopts minimum-maximum standardization. The minimum and maximum values ​​required for standardization are extracted from historical data, such as using the extreme values ​​of the indicators in the past month.

[0027] The hydraulic transport capacity of each pipe section is extrapolated based on the comparison between the coordinated state assessment parameters and historical normal operation data. Historical normal operation data consists of the coordinated state assessment parameter values ​​and their corresponding actual hydraulic transport capacity data recorded during the system's stable operation period. For example, data from the past 30 days of fault-free system operation is selected as historical normal operation data. The comparison process is achieved by calculating the deviation between the current coordinated state assessment parameters and historical normal values. The deviation calculation uses a relative difference method, such as dividing the difference between the current parameter value and the median of the historical normal value range by the width of the historical normal value range. The extrapolation of hydraulic transport capacity is adjusted based on the magnitude of the deviation. If the deviation is less than 0.1, the hydraulic transport capacity is considered to be at a normal level. If the deviation is between 0.1 and 0.3, the estimated hydraulic transport capacity is reduced proportionally. If the deviation is greater than 0.3, the lowest historical hydraulic transport capacity value is used. The simulation results are expressed as the maximum allowable flow rate of the pipe section, with the unit being cubic meters per hour. The physical characteristics of the pipe section are also taken into account, such as the limitations of pipe diameter and material on transport capacity. The simulation process also incorporates real-time operating conditions, such as the effect of ambient temperature on fluid viscosity. The transport capacity value is adjusted by a correction factor, which is set based on experimental data. For example, for every 10 degrees Celsius increase in temperature, the transport capacity decreases by 1%.

[0028] A parametric spatial point cloud of flow demand and hydraulic transport capacity for each pipe segment is constructed. Topological invariant features are extracted using topological data analysis algorithms, and conflicting pipe segments are identified based on their persistence. The specific implementation is as follows: A parameter space point cloud of flow demand and hydraulic transport capacity for each pipe segment is constructed. Topological invariant features are extracted using topological data analysis algorithms, and conflicting pipe segments are identified based on their persistence. Data preparation is achieved by mapping the flow demand and hydraulic transport capacity values ​​of each pipe segment to data points in a multi-dimensional parameter space. The flow demand values ​​are derived from real-time load forecasts of the system, while the hydraulic transport capacity values ​​are calculated based on the physical characteristics and operating status of the pipe segments. The dimension of the parameter space is determined according to the analysis requirements; for example, in a 2-dimensional space, the flow demand value is used as the horizontal axis and the hydraulic transport capacity value as the vertical axis. The coordinate values ​​of the data points are standardized to eliminate the influence of dimensions. The standardization method uses min-max normalization to linearly transform the original values ​​to the range of 0 to 1. The construction of the parameter space point cloud considers the time dimension; for example, data from 10 consecutive sampling periods constitute a dynamic point cloud, with each sampling period lasting 5 minutes. The size of the point cloud is determined according to the total number of pipe segments in the system, for example, containing 100 to 500 data points. The spatial distribution characteristics of the data points are preliminarily analyzed using density clustering to identify abnormal distribution patterns.

[0029] Topological data analysis algorithms are applied to parametric point clouds to calculate persistent homology groups and extract topological invariant features. The algorithm extracts topological features by constructing a filtered complex sequence of the parametric point cloud and calculating the persistent homology group. The construction of the filtered complex sequence is based on the distance relationships between data points in the point cloud, for example, using Euclidean distance as a metric. A series of simple complexes are generated by gradually increasing distance parameters. The range of distance parameters is set according to the point cloud distribution density, for example, starting from 0 and increasing to 2 in steps of 0.1. The calculation of the persistent homology group is achieved by analyzing the generation and disappearance process of topological features in the filtered complex sequence. The distance parameter value at which each topological feature appears is recorded as the birth time, and the distance parameter value at which it disappears is recorded as the death time. Betti numbers are extracted from the persistent homology group as topological invariant features. The Betti number represents the number of topological holes in each dimension; for example, the 0-dimensional Betti number represents the number of connected components, and the 1-dimensional Betti number represents the number of ring structures. Simultaneously, the lifetime length in persistent barcodes is recorded. The lifetime length is calculated by the difference between the death time and the birth time, characterizing the stability of the topological features.

[0030] The persistence metric of topological invariant features is analyzed. When the persistence falls below a preset threshold, the corresponding pipe segment is identified as a conflicting pipe segment. The persistence metric is based on the lifecycle length of the topological feature. The preset threshold is determined according to the system stability requirements. For example, the average lifecycle length of the topological feature under normal operating conditions is obtained through historical data analysis, and 70% of this value is used as the threshold. The specific threshold setting considers the distribution characteristics of the pipe segment in the parameter space. When the lifecycle length of the topological feature corresponding to a data point of a certain pipe segment is lower than the threshold, the pipe segment is determined to have an operational conflict. The identification of conflicting pipe segments also combines the type of topological invariant features for comprehensive analysis. For example, when the lifecycle of the ring structure corresponding to the 1-dimensional Betti number is too short, it indicates that there is a local circulation flow anomaly in the system, and the corresponding pipe segment is marked as a conflicting pipe segment. The synergistic effect of multiple topological features is considered during the identification process. For example, the confidence of the conflict determination is improved when multiple low persistence features appear at the same time. The final determination of the conflicting pipe segment needs to be verified through a continuous time window. For example, a conflict state is confirmed only when the persistence is detected to be lower than the threshold for three consecutive sampling periods.

[0031] The implementation of the topology data analysis algorithm also includes data preprocessing and postprocessing steps. Data preprocessing includes point cloud denoising and missing value handling. Denoising uses the local outlier detection method to identify and remove points with abnormal distances from other points. Missing value handling uses the nearest neighbor interpolation method. Postprocessing includes the classification and archiving of topological features. The extracted topological invariant features are classified and stored according to dimensionality and persistence. A feature library is established for historical comparison. The algorithm parameters are dynamically adjusted according to the system scale. For example, when the point cloud scale is large, the step size of the distance parameter is appropriately increased to improve the computational efficiency. All calculation processes are implemented on a dedicated computing platform for hydraulic systems. The platform adopts a distributed architecture to process large-scale point cloud data, ensuring that the timeliness of the analysis meets the real-time control requirements.

[0032] Based on conflicting pipe sections and pipeline topology, the potential impact of flow path changes on related pipe sections during regulation is analyzed, and pipe sections with hidden congestion risks are identified. The specific implementation is as follows: Based on conflicting pipe segments and the network topology, this study analyzes the potential impact of changes in flow paths on related pipe segments during regulation, identifies pipe segments with hidden congestion risks, and defines the scope of influence by determining a group of related pipe segments with hydraulic connections to the conflicting pipe segments according to the network topology. The network topology is represented as a directed graph, where nodes represent pipe connection points and edges represent pipe segments and flow directions. The determination of hydraulic connections is based on connectivity analysis in graph theory. For example, starting from the conflicting pipe segment, all directly or indirectly connected pipe segments are searched along the flow direction. The search depth is set according to the system complexity, for example, limited to pipe segments within 3 hops. The composition of the related pipe segment group considers the functional characteristics of the pipe segments. For example, pipe segments within the same heating branch are preferentially grouped into the same group, while duplicate calculations of pipe segments already marked as conflicting are excluded. The size of the related pipe segment group is limited by the maximum number of pipe segments, for example, each group does not exceed 15 pipe segments, to ensure the feasibility of subsequent analysis.

[0033] The simulation of flow distribution changes in a group of associated pipe segments after a change in flow path during the control process is based on a hydraulic model of the pipe network. The hydraulic model includes basic elements such as pipe segment resistance characteristics and node flow balance equations. The simulation of flow path changes is achieved by adjusting the flow distribution weights of conflicting pipe segments, for example, by reducing the flow distribution coefficient of conflicting pipe segments by a certain proportion, and at the same time, redistributing the flow to adjacent pipe segments according to the topology. The flow distribution calculation adopts an iterative solution method, such as using the Hardy-Cross method to solve the pipe network node equations. The convergence condition of the iteration is set as the flow difference between two adjacent calculation results is less than 0.1 cubic meters per hour. The simulation considers the temporal characteristics of the control operation, for example, the flow path change process is discretized into multiple time steps, each with a length of 30 seconds, simulating the flow change process within the next 5 minutes. The flow distribution results include the instantaneous flow values ​​and change trends of each pipe segment.

[0034] The pressure gradient change of the associated pipe segment group after the flow distribution change is calculated. The pressure gradient change is obtained by comparing the pressure distribution before and after the regulation. The pressure distribution is obtained by solving the pipeline pressure equation. The pressure gradient is calculated as the ratio of the pressure difference between adjacent nodes to the pipe segment length, with the unit being kPa per meter. The pressure gradient change is calculated by selecting the pressure gradient value at a specific time point, such as the pressure gradient data at the 3rd minute after the regulation begins. The change is calculated using the relative change rate method, which is the ratio of the difference between the pressure gradient after regulation and the reference pressure gradient to the reference pressure gradient. The reference pressure gradient is taken from the average value of the stable operating state before regulation. The calculation process considers the spatial position relationship of the pipe segments. For example, the cumulative effect of the pressure gradient change is calculated for continuous pipe segments with the same flow direction. When the pressure gradient change of continuous pipe segments increases in the same direction, they are marked as key areas of concern.

[0035] Hidden congestion risk pipe sections are identified based on the comparison between pressure gradient changes and historical normal pressure gradient ranges. Historical normal pressure gradient ranges are obtained by statistically analyzing pressure gradient data during stable system operation; for example, the range of pressure gradient values ​​occurring for 95% of the past 30 days is used as a normal reference. The comparison process employs segmented threshold judgments. For instance, when the pressure gradient change exceeds the upper limit of the historical normal range by 10%, it is marked as primary risk; exceeding 20% ​​as intermediate risk; and exceeding 30% as high risk. The threshold settings for risk levels consider the importance of the pipe section in the system; for example, the threshold setting for main pipes is stricter than that for branch pipes. The final determination of hidden congestion risk pipe sections requires meeting a persistence condition; for example, a risk state is only confirmed after two consecutive sampling cycles. The identification results are verified in conjunction with flow distribution change characteristics; for example, when the pressure gradient change increases simultaneously with abnormal flow distribution, the confidence level of the risk determination is increased. All identified hidden congestion risk pipe sections have their risk level and characteristic parameters recorded for subsequent control decisions.

[0036] Based on the conflict pipe sections and the pipe sections with hidden congestion risks, the output commands of each heat source are recalculated, and the specific implementation is as follows: By combining conflicting pipe sections and pipe sections with latent congestion risks, the output commands of each heat source are recalculated. A load redistribution model is established with the elimination of conflicting and latent congestion risk pipe sections as constraints and the minimization of total output adjustment as the optimization objective. The constraints are set based on the hydraulic characteristic parameters of conflicting and latent congestion risk pipe sections. For example, the flow rate of conflicting pipe sections is limited to the safe operating range, and the pressure gradient change of latent congestion risk pipe sections is controlled within the historical normal fluctuation range. The total output adjustment in the optimization objective is represented by the sum of the absolute values ​​of the adjustment of each heat source, with the unit being megawatts. The load redistribution model also includes basic constraints on system operation, such as upper and lower limits of heat source output, flow balance constraints at pipe network nodes, and heat source regulation rate constraints. The model is solved using linear programming, transforming each constraint into a system of linear inequalities, and the objective function into a linear function form for optimization.

[0037] The output adjustment priority of each heat source is determined based on the hydraulic sensitivity coefficients of each heat source, conflicting pipe sections, and pipe sections with latent congestion risk. The hydraulic sensitivity coefficients are obtained by analyzing the impact of heat source output changes on the hydraulic parameters of the pipe sections. Numerical differentiation methods are used in the calculation, such as calculating the weighted sum of the flow rate change in conflicting pipe sections and the pressure gradient change in pipe sections with latent congestion risk when the heat source output changes by a unit. The weight allocation is determined based on the importance of the pipe sections. For example, the weight of the main pipe is set to 0.7, and the weight of the branch pipe is set to 0.3. The calculation of the hydraulic sensitivity coefficients takes into account the influence of the pipe network topology. For example, the connection relationship between heat sources and pipe sections is established through the node-branch association matrix. The output adjustment priority is sorted according to the size of the hydraulic sensitivity coefficient. For example, heat sources with larger hydraulic sensitivity coefficients are adjusted first. The priority division adopts a hierarchical method. For example, heat sources with hydraulic sensitivity coefficients greater than 0.8 are listed as first priority, those with coefficients between 0.5 and 0.8 are listed as second priority, and those with coefficients less than 0.5 are listed as third priority. The priority setting also considers the actual adjustment capacity of the heat source. For example, heat sources with a large adjustment range are given a higher priority.

[0038] The load redistribution model is transformed into a sequential optimization problem based on the output adjustment priority. The transformation process decomposes the load redistribution model into a sequence of subproblems, each corresponding to a priority level of heat source output optimization. These subproblems are solved sequentially from highest to lowest priority. For example, the output adjustment of the first priority heat source is solved first, and the adjustment result of the first priority heat source is fixed when solving the second priority heat source, and so on until all priority levels are solved. An iterative algorithm is used to solve the sequential optimization problem. The optimization method for each subproblem follows that of the load redistribution model. During the solution process, equality constraints are introduced for heat sources with fixed adjustment amounts to ensure that subsequent optimizations do not change the determined adjustment amounts. Convergence conditions are set for iterative solutions, such as a total output adjustment change of less than 0.1 MW between two adjacent iterations. Sequential optimization also considers coordination between different priority levels; for example, the optimization order of lower-level heat sources is automatically adjusted when higher-level optimizations cannot meet the constraints.

[0039] Based on the solution to the sequence optimization problem, new output commands are generated for each heat source. The new output command is obtained by adding the current output value to the optimized output adjustment amount. The generation of the new output command considers the feasibility and safety of the command. For example, command range verification is performed to ensure that the new output command is within the allowable operating range of the heat source. At the same time, command change rate verification is performed to ensure that the change amplitude between adjacent commands does not exceed the maximum adjustment rate of the heat source. The new output command also undergoes system-level verification. For example, hydraulic simulation is used to verify whether the status of conflicting pipe sections and pipe sections with hidden congestion risks under the new command achieves the expected improvement target. If the verification fails, the optimization parameters are readjusted. The output format of the new output command includes fields such as heat source identifier, output value, and effective time. The output value is accurate to 0.1 MW, and the effective time is set with a reasonable delay to adapt to the system response characteristics. For example, it is set to take effect at the exact minute after 5 minutes.

[0040] The recalculated output command is then sent to the corresponding heat source in the multi-heat-source system. Specifically, this is implemented as follows: The recalculated output command is sent to the corresponding heat source in the multi-heat source system. The compatibility of the command is ensured by verifying its compatibility with the current operating status of each heat source. The verification process includes output command value range check and output change rate check. The output command value range check is achieved by comparing the new output command with the technical parameters of the heat source equipment. For example, the new output command value should be within 50% to 100% of the rated output of the heat source. The output change rate check is achieved by calculating the difference between the new output command and the current output value and comparing it with the maximum adjustment rate of the heat source. For example, the difference should not exceed the maximum adjustment capacity of the heat source per minute. The real-time status of the heat source equipment is also considered during the verification process. For example, when the heat source is in maintenance status, the command issuance is automatically suspended. The verification results are divided into three levels: fully compatible, requiring adjustment, and unexecutable. For cases requiring adjustment, the command is automatically fine-tuned. For example, the output value that is out of range is adjusted to the closest allowable value.

[0041] The priority sequence for issuing commands is determined based on the hydraulic distance between each heat source and conflicting pipe segment in the pipeline network topology. The hydraulic distance is calculated using the shortest path length between the heat source and the conflicting pipe segment in the pipeline network topology. The path length is expressed as the equivalent length of the pipe segment in meters. The hydraulic distance calculation takes into account the influence of pipe segment diameter and roughness. For example, the actual pipe segment is converted into an equivalent length using the Darcy-Weisbach formula. The command issuance priority sequence is arranged in ascending order of hydraulic distance. For example, the heat source with the smallest hydraulic distance is given priority. When multiple heat sources have the same hydraulic distance, the priority is determined by referring to the heat source capacity. For example, the heat source with the larger capacity is given priority. The determination of the priority sequence also considers the correlation between the heat source and the pipe segment with latent congestion risk. For example, the heat source that is close to both the conflicting pipe segment and the pipe segment with latent congestion risk is given an appropriate higher priority. The priority sequence is recalculated every fixed period, for example, updated every 5 minutes to adapt to changes in system status.

[0042] According to the priority sequence of the instructions, the recalculated output instructions are sent to the corresponding heat sources through a secure communication protocol. The secure communication protocol adopts a transport layer security protocol to ensure data transmission security, such as using the TLS 1.3 protocol to establish an encrypted channel. Before the instructions are sent, the data is encapsulated. The encapsulation format includes an instruction header and a payload. The instruction header includes information such as the heat source identifier, instruction sequence number, and timestamp. The payload includes the output instruction value and the effective time. The sending process adopts a message queue mechanism to ensure the orderly transmission of instructions. For example, an independent message queue is set for each priority, and the instructions of each heat source are processed in sequence according to the priority sequence. If the sending fails, a retransmission mechanism is initiated. For example, the maximum number of retransmissions is set to 3, and the retransmission interval is gradually increased. The first retransmission interval is 2 seconds, and the interval doubles for each subsequent retransmission.

[0043] The system receives confirmation feedback from each heat source regarding output commands and records the command execution status. Confirmation feedback is received through a secure communication protocol feedback channel. Feedback information includes fields such as heat source identifier, command reception status, and estimated execution time. Real-time monitoring of command execution status is achieved through heat source operation data acquisition. For example, it monitors the deviation between the actual output value and the command value of the heat source. For instance, when the deviation between the actual output value and the command value continues to exceed 5% for 1 minute, an abnormal alarm is triggered. The command execution status record contains a complete historical trajectory, including time nodes such as command sending time, receiving time, start execution time, and completion time. The recorded data is stored in a distributed database for subsequent analysis and auditing. At the same time, a command execution timeout mechanism is established. For instance, when a heat source fails to start executing the command within the predetermined time, a backup control strategy is automatically triggered.

[0044] Example 2: An intelligent control device for multi-branch thermal systems, the device including an intelligent control system for multi-branch thermal systems.

[0045] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0046] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0047] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0048] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0049] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0050] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0051] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent control system for multi-branch thermal systems, characterized in that, include: The data acquisition module is used to acquire in real time the output commands of each heat source in the multi-heat source system and the pressure and flow data of key nodes in the pipeline network; The state simulation module is used to analyze the coordinated state changes between related pipe segments based on pressure and flow data and the pipeline network topology, and to deduce the hydraulic transport capacity of each pipe segment. The topology identification module is used to construct a parametric spatial point cloud of flow demand and hydraulic transport capacity for each pipe segment, extract topological invariant features through topological data analysis algorithms, and identify conflicting pipe segments based on their persistence. The risk warning module is used to analyze the potential impact of changes in flow paths on related pipe sections during the control process, based on conflict pipe sections and pipe network topology, and to identify pipe sections with hidden congestion risks. The command reconfiguration module is used to recalculate the output commands of each heat source by combining conflict pipe sections and pipe sections with hidden congestion risk. The command issuing module is used to issue the recalculated output command to the corresponding heat source in the multi-heat source system.

2. The intelligent control system for multi-branch thermal systems according to claim 1, characterized in that, Real-time acquisition of output commands from each heat source in a multi-heat-source system and pressure and flow data from key nodes in the pipeline network, including: Receive output commands from each heat source from the central control platform; Pressure and flow data are collected by sensors deployed at key nodes of the pipeline network.

3. The intelligent control system for multi-branch thermal systems according to claim 1, characterized in that, Based on pressure and flow data, and considering the pipeline network topology, the coordinated state changes among related pipe segments are analyzed to infer the hydraulic transport capacity of each segment, including: Based on the pipeline network topology, pipe segments with hydraulic relationships are identified to form a group of associated pipe segments. Calculate the correlation coefficient of pressure gradient between adjacent pipe segments in a group of associated pipe segments and the synchronicity index of flow rate changes; Coordinated state assessment parameters are constructed based on the pressure gradient correlation coefficient and the flow rate change synchronicity index; The hydraulic transport capacity of each pipe section is inferred based on the comparison results between the collaborative status assessment parameters and historical normal operation data.

4. The intelligent control system for multi-branch thermal systems according to claim 1, characterized in that, A parametric spatial point cloud of flow demand and hydraulic transport capacity for each pipe segment is constructed. Topological invariant features are extracted using topological data analysis algorithms, and conflicting pipe segments are identified based on their persistence, including: The flow demand and hydraulic transport capacity of each pipe segment are mapped to data points in a multi-dimensional parameter space to form a parameter space point cloud; Topological data analysis algorithms are applied to parameter space point clouds to calculate persistent homology groups and extract topological invariant features. The persistence measure of topological invariant features is analyzed. When the persistence is lower than a preset threshold, the corresponding pipe segment is identified as a conflicting pipe segment.

5. The intelligent control system for multi-branch thermal systems according to claim 4, characterized in that, The application of topological data analysis algorithms to parameter space point clouds to calculate persistent homology groups and extract topological invariant features includes: constructing a filtered complex sequence of parameter space point clouds and calculating persistent homology groups, and extracting Betti numbers from the persistent homology groups as topological invariant features.

6. The intelligent control system for multi-branch thermal systems according to claim 1, characterized in that, Based on conflicting pipe sections and network topology, this study analyzes the potential impact of flow path changes on related pipe sections during regulation and control, identifying pipe sections with hidden congestion risks, including: Identify the associated pipe segment group that has hydraulic connections with the conflicting pipe segment based on the pipe network topology; Changes in flow distribution within the associated pipe segment group after a change in flow path during simulation control process; Calculate the change in pressure gradient of the associated pipe segment group after the change in flow distribution; Hidden congestion risk segments were identified based on the comparison between the pressure gradient change and the historical normal pressure gradient range.

7. The intelligent control system for multi-branch thermal systems according to claim 1, characterized in that, Based on the conflict pipe sections and the pipe sections with hidden congestion risks, the output commands of each heat source are recalculated, including: Establish a load redistribution model with the constraints of eliminating conflicting pipe sections and pipe sections with hidden congestion risk, and with the optimization objective of minimizing the total output adjustment; The output adjustment priority of each heat source is determined based on the hydraulic sensitivity coefficient of each heat source and the conflict pipe section and the pipe section with hidden congestion risk. The load redistribution model is transformed into a sequential optimization problem based on the output adjustment priority, and then solved sequentially. New output commands for each heat source are generated based on the solution results of the sequence optimization problem.

8. The intelligent control system for multi-branch thermal systems according to claim 7, characterized in that, The load redistribution model is transformed into a sequential optimization problem according to the output adjustment priority and solved sequentially. This includes: decomposing the load redistribution model into a sequence of subproblems according to the output adjustment priority, solving the output adjustment amount of each heat source in order of priority, and fixing the adjustment results of the solved heat sources during the solution process.

9. The intelligent control system for multi-branch thermal systems according to claim 1, characterized in that, The recalculated output command is sent to the corresponding heat source in the multi-heat-source system, including: Verify the compatibility between the recalculated output command and the current operating status of each heat source; The priority sequence for issuing instructions is determined based on the hydraulic distance between each heat source and the conflicting pipe section in the pipeline network topology. According to the priority sequence of the instructions, the recalculated output instructions are sent to the corresponding heat sources through a secure communication protocol. Receive confirmation feedback from each heat source regarding the output command and record the command execution status.

10. An intelligent control device for multi-branch thermal systems, characterized in that, The device includes an intelligent control system for multi-branch thermal systems as described in any one of claims 1-9.

Images