CN-122018308-A - Heat supply two-network dynamic balance intelligent regulation and control method based on fluid mechanics and AI coupling

CN122018308ACN 122018308 ACN122018308 ACN 122018308ACN-122018308-A

Abstract

The invention provides a dynamic balance intelligent regulation and control method of a heat supply two-network based on fluid mechanics and AI coupling, which relates to the technical field of data processing, and comprises the steps of acquiring pressure, temperature and flow data of all nodes in real time through an Internet of things sensor network deployed on a heat supply two-network, and constructing a multi-dimensional space-time characteristic data set; based on the established two-network hydraulic working condition space-time coupling dynamic model, training and optimizing the model by a machine learning algorithm to obtain an optimized model for performing second-level high-precision dynamic deduction on the pressure gradient and flow distribution of the pipeline. The invention realizes the full chain intellectualization of the two heating networks from perception, deduction and regulation.

Inventors

LIU YUBIN
CHEN YUDI
YANG YAMEI
FU QIANG
MEI DEFANG
WU SHENGJUN
YANG GAISHUN
YAN DONGXU
ZHANG JIANHUA
ZHAO NAN
YIN XIANZHEN

Assignees

北京京能热力股份有限公司

Dates

Publication Date: 20260512
Application Date: 20251226

Claims (10)

1. The utility model provides a two network dynamic balance intelligent regulation and control methods of heat supply based on fluid mechanics and AI coupling which characterized in that, the method includes: The method comprises the steps of acquiring pressure, temperature and flow data of all nodes in real time through an Internet of things sensor network deployed in a heat supply secondary pipe network, and constructing a multi-dimensional space-time characteristic data set; Based on the established two-network hydraulic working condition space-time coupling dynamic model, training and optimizing the model through a machine learning algorithm to obtain an optimized model for performing second-level high-precision dynamic deduction on the pressure gradient and flow distribution of the pipeline network; Based on the real-time deduction result, the operation data is subjected to deep analysis through a neural network algorithm, and the flow mutation at the user side and the abnormal working condition of pipe network leakage are intelligently identified so as to obtain a real-time regulation strategy; converting the real-time regulation strategy into a control instruction, driving the variable frequency pump and the electric regulating valve actuating mechanism to act, and obtaining pipe network state feedback data after the actuating mechanism acts; The method comprises the steps of collecting pipe network state feedback data after the action of an executing mechanism, using the feedback data to self-learn and optimize control parameters of a neural network algorithm to obtain optimized control parameters, and feeding the optimized control parameters back to a recognition and strategy generation process to realize self-adaptive closed-loop regulation.
2. The intelligent regulation and control method for dynamic balance of a heat supply two-network based on hydrodynamic and AI coupling of claim 1, wherein the pressure, temperature and flow data of all nodes are collected in real time through an internet of things sensor network deployed in a heat supply two-network, and a multi-dimensional space-time characteristic data set is constructed, comprising: the method comprises the steps of receiving a full-node data set which is uploaded by an Internet of things sensor network deployed in a heat supply secondary pipe network and contains pressure, temperature and flow data and three-dimensional space coordinates of each sensor, abstracting a pipeline structure into a space topology geometric structure in a three-dimensional space based on a space topology relation of the heat supply secondary pipe network, and carrying out position verification and data validity verification on the sensor space coordinates according to a three-dimensional space dotted line relation judging algorithm to obtain a verified effective full-node data set; Carrying out space-time alignment and data cleaning treatment on the verified effective full-node data to obtain regular space-time sequence data; And fusing the extracted pressure gradient, flow rate change rate and temperature distribution characteristics to obtain a multi-dimensional space-time characteristic data set.
3. The intelligent regulation and control method for the dynamic balance of the two networks for heat supply based on the hydrodynamic and the AI coupling of the invention according to claim 2 is characterized in that based on the constructed multidimensional space-time characteristic data set, a two-network hydraulic working condition space-time coupling dynamic model is constructed through a hydrodynamic continuity equation and a Bernoulli dynamic equation, based on the established two-network hydraulic working condition space-time coupling dynamic model, the model is trained and optimized through a machine learning algorithm to obtain an optimized model for carrying out second-level high-precision dynamic deduction on the pressure gradient and the flow distribution of the pipeline network, and the method comprises the following steps: Based on the obtained multi-dimensional space-time characteristic data set, constructing an initial space-time coupling dynamic model of a two-network hydraulic working condition through a hydrodynamic continuity equation and a Bernoulli dynamic equation; Performing iterative training on the constructed initial space-time coupling dynamic model by using historical data in the multi-dimensional space-time characteristic data set as a training sample through a machine learning algorithm to obtain an optimized model for performing second-level dynamic deduction on the pipeline network pressure gradient and flow distribution; based on the optimized model, the obtained optimized model is verified and parameters are finely adjusted through data acquired in real time, so that model deduction accuracy is maintained.
4. The intelligent regulation and control method for dynamic balance of two heat supply networks based on fluid mechanics and AI coupling according to claim 3, wherein the method is characterized in that the optimized model is used as an intelligent center to perform deduction calculation of real-time hydraulic state of the pipe network to obtain real-time deduction result, and comprises the following steps: loading the optimized model to an intelligent center through the obtained optimized model to obtain a loaded optimized model; Inputting the multidimensional space-time characteristic data set acquired in real time into a loaded optimized model; Performing second-level hydraulic calculation on the input real-time data through the optimized model, and deducting to obtain real-time hydraulic state data comprising pressure gradient and flow distribution of the whole pipe network; and carrying out abnormal working condition characteristic extraction analysis based on the deduced real-time hydraulic state data to generate a real-time deduction result comprising the real-time hydraulic state and the abnormal recognition result of the pipe network.
5. The intelligent regulation and control method for dynamic balance of two heat supply networks based on fluid mechanics and AI coupling according to claim 4, wherein based on real-time deduction results, the operation data is subjected to deep analysis by a neural network algorithm, and user side flow mutation and abnormal pipe network leakage conditions are intelligently identified to obtain a real-time regulation and control strategy, comprising: carrying out abnormal working condition characteristic analysis on the received real-time deduction result, and extracting characteristic vectors of user side flow mutation and pipe network leakage; inputting the extracted feature vector into a pre-trained neural network anomaly identification model, intelligently diagnosing specific anomaly condition types and severity levels, and obtaining anomaly diagnosis results; based on the obtained abnormal diagnosis result, combining with a reinforcement learning control strategy library to generate a preliminary regulation instruction set aiming at the variable frequency pump and the electric regulating valve; And (3) carrying out hydraulic safety and stability verification on the generated preliminary regulation instruction set, and outputting a final executable real-time regulation strategy.
6. The intelligent regulation and control method for dynamic balance of two heat supply networks based on fluid mechanics and AI coupling according to claim 5, wherein converting the real-time regulation and control strategy into control instructions to drive the variable frequency pump and the electric regulating valve actuator to act, obtaining pipe network state feedback data after the actuation of the actuator, comprises: Analyzing and converting the received real-time regulation strategy into a standardized control instruction which can be identified by an edge controller; The generated standardized control instruction is issued to a corresponding variable frequency pump and an electric control valve executing mechanism; the execution state of the issued control instruction is monitored in real time, and pressure, temperature and flow data of the pipe network after the action of the execution mechanism are collected and used as pipe network state feedback data; And summarizing and preprocessing the collected pipe network state feedback data to obtain normalized feedback data for controlling parameter self-learning.
7. The intelligent regulation and control method for dynamic balance of two heat supply networks based on hydrodynamic and AI coupling of claim 6, wherein the method is characterized by collecting pipe network state feedback data after the action of an executing mechanism, using the feedback data to self-learn and optimize control parameters of a neural network algorithm to obtain optimized control parameters, feeding the optimized control parameters back to an identification and strategy generation process to realize self-adaptive closed-loop regulation and control, and comprising the following steps: based on the received normalized feedback data set, performing self-learning optimization on the control parameters of the neural network anomaly identification model and the reinforcement learning control strategy library to obtain an optimized control parameter set; The obtained optimized control parameter set is fed back to an abnormal working condition characteristic extraction and analysis process and a neural network abnormal recognition model so as to obtain updated control parameters; Based on the updated control parameters, continuous self-adaptive optimization of the control strategy is realized in the process of real-time deduction and strategy generation, and closed-loop regulation and control are formed.
8. A two-network dynamic balance intelligent regulation system for heat supply based on hydrodynamic and AI coupling, which realizes the method as set forth in any one of claims 1 to 7, comprising: The acquisition module is used for acquiring pressure, temperature and flow data of all nodes in real time through an Internet of things sensor network deployed in the heat supply secondary pipe network, and constructing a multi-dimensional space-time characteristic data set; The construction module is used for constructing a two-network hydraulic working condition space-time coupling dynamic model based on the constructed multi-dimensional space-time characteristic data set through a hydrodynamic continuity equation and a Bernoulli dynamic equation, training and optimizing the model through a machine learning algorithm based on the established two-network hydraulic working condition space-time coupling dynamic model to obtain an optimized model for performing second-level high-precision dynamic deduction on the pipeline pressure gradient and the flow distribution; The optimizing module is used for taking the optimized model as an intelligent center, carrying out deduction calculation on the real-time hydraulic state of the pipe network to obtain a real-time deduction result, carrying out deep analysis on the operation data through a neural network algorithm based on the real-time deduction result to intelligently identify the flow mutation at the user side and the abnormal pipe network leakage working condition so as to obtain a real-time regulation strategy, converting the real-time regulation strategy into a control instruction, and driving a variable frequency pump and an electric regulating valve executing mechanism to act to obtain pipe network state feedback data after the executing mechanism acts; the processing module is used for collecting pipe network state feedback data after the action of the executing mechanism, using the feedback data for self-learning and optimizing control parameters of the neural network algorithm to obtain optimized control parameters, and feeding the optimized control parameters back to the identification and strategy generation process to realize self-adaptive closed-loop regulation and control.
9. A computing device, comprising: one or more processors; Storage means for storing one or more programs which when executed by the one or more processors cause the one or more processors to implement the method of any of claims 1 to 7.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein a program which, when executed by a processor, implements the method according to any of claims 1 to 7.

Description

Heat supply two-network dynamic balance intelligent regulation and control method based on fluid mechanics and AI coupling Technical Field The invention relates to the technical field of data processing, in particular to a dynamic balance intelligent regulation and control method for a heat supply two-network based on fluid mechanics and AI coupling. Background In the field of urban central heating secondary pipe network regulation and control, the traditional regulation and control mode is limited by the double technical bottlenecks of manual experience dependence and static hydraulic model, dynamic complexity of pipe network operation is difficult to adapt, and in 2024-2025 heating season, due to the fact that a courtyard network lacks intelligent automatic control devices and data-driven analysis capability, workers need to manually measure temperature, read pressure and flow data on site every day, basic regulation and control are realized through manual regulation balance valves, a complete regulation period is as long as 3 weeks, the limitation of manual experience is limited, the regulation effect can only be maintained for 2-3 days, obvious thermodynamic imbalance problems that a near-heat building needs to be windowed for heat dissipation at a temperature of 26 ℃ and a far-end building needs to be only 16 ℃ can occur quickly, and more serious, when the pipe network leaks, the traditional mode completely depends on manual inspection and check, fine parameter changes caused by leakage cannot be quickly captured, and the heating of a 3 building is stopped for 10 hours once, so that a user is seriously influenced. The method has the core defects that firstly, the dynamic modeling capability based on fluid mechanics is lacking, the dynamic characteristics of the change of the pressure, the flow and the user load of a pipe network cannot be reflected by a traditional steady-state model, the overestimated error of the heat loss can reach 27.4%, the accurate regulation and control are difficult to support, secondly, the AI technology, particularly the support of a neural network algorithm, cannot be used for carrying out deep analysis on multidimensional space-time data acquired by a pipe network full-node sensor, the traditional method cannot extract the fine characteristics of abnormal working conditions such as user side flow mutation, pipe network leakage and the like through the algorithm such as a convolutional neural network and the like, the abnormal recognition logic driven by the data cannot be constructed by the neural network, the manual trial-and-error regulation and the periodic inspection can only be relied on, the regulation and control efficiency is low, the precision is not enough, the real-time recognition and the dynamic response of the abnormal working conditions cannot be realized, the repeated occurrence of the heat unbalance and the serious energy waste are caused, and the core requirements of modern heat supply on the accurate regulation and the efficient energy-saving operation are difficult to meet. Disclosure of Invention The technical problem to be solved by the invention is to provide the intelligent regulation and control method for the dynamic balance of the two heat supply networks based on the coupling of the fluid mechanics and the AI, so that the full chain intellectualization from sensing, deduction and regulation of the two heat supply networks is realized. In order to solve the technical problems, the technical scheme of the invention is as follows: in a first aspect, a method for intelligently regulating and controlling dynamic balance of a heat supply two-network based on fluid mechanics and AI coupling, the method comprises the following steps: The method comprises the steps of acquiring pressure, temperature and flow data of all nodes in real time through an Internet of things sensor network deployed in a heat supply secondary pipe network, and constructing a multi-dimensional space-time characteristic data set; Based on the established two-network hydraulic working condition space-time coupling dynamic model, training and optimizing the model through a machine learning algorithm to obtain an optimized model for performing second-level high-precision dynamic deduction on the pressure gradient and flow distribution of the pipeline network; Based on the real-time deduction result, the operation data is subjected to deep analysis through a neural network algorithm, and the flow mutation at the user side and the abnormal working condition of pipe network leakage are intelligently identified so as to obtain a real-time regulation strategy; converting the real-time regulation strategy into a control instruction, driving the variable frequency pump and the electric regulating valve actuating mechanism to act, and obtaining pipe network state feedback data after the actuating mechanism acts; The method comprises the steps of collecting pipe network state feedback data aft