Heat supply trusted intelligent control method and system based on full trust creation

Abstract

The invention relates to a heat supply credible intelligent control method and system based on full trust creativity, and the method comprises the steps: collecting heat supply parameters in real time through Internet of Things equipment, and carrying out the preprocessing of the heat supply parameters through an edge calculation node, and obtaining the preprocessing data; performing thermal load prediction on the pre-processed data through a bidirectional LSTM neural network model based on a full trust creation platform, and outputting a thermal load prediction curve in a future time period; generating a dynamic scheduling instruction based on the thermal load prediction curve in combination with a real-time equipment state and an energy consumption optimization strategy; and the dynamic scheduling instruction is issued to the edge computing node, the operation parameters of the heat source, the pipe network and the user side are automatically adjusted, and when it is detected that the equipment is abnormal, standby equipment is automatically switched or a heat supply path is adjusted, and unmanned regulation and control of the heat supply system are achieved. According to the application, edge computing, the Internet of Things, trusted transmission and industrial control safety capabilities are integrated, so that full-flow coverage of heat supply scene data acquisition and analysis is realized, and the unattended scheduling capability of heat supply operation is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent heating technology, and in particular to a trusted intelligent heating control method and system based on full-innovation. Background Art

[0002] The heating business refers to the whole process of generating heat energy through heat sources (such as cogeneration units, coal-fired / gas-fired boilers, waste heat recovery devices, etc.), transporting it to heat exchange stations through the pipeline system (primary network and secondary network), and then transferring the heat to residential, commercial and industrial user terminals through heat exchange equipment to meet winter heating needs. It involves key links such as heat source production, pipeline distribution, and user heat use.

[0003] Traditional heating services have problems such as complicated operating procedures, a large variety of spare parts and accessories involved and high usage, and long single operations. At the same time, the monitoring methods for the entire business process still rely on traditional methods, which are subject to information lags and make it difficult to achieve real-time tracking and dynamic updates of business processes. As the scale of heating provided by heating companies continues to expand and the demand for heating services from heat users becomes increasingly diversified, the existing information system is gradually showing bottlenecks in data processing capabilities and service scalability: on the one hand, the system is unable to bear the explosive growth in computing resource demands; on the other hand, there are obvious data barriers between business systems, and the cross-system collaborative analysis capabilities are insufficient to support refined operational decisions. This results in a limited level of system automation, and key links still require manual intervention based on historical experience, resulting in insufficient dynamic matching between heating supply and actual demand.

[0004] In response to the above problems, this application provides a reliable intelligent control method and system for heating based on full trust and innovation. Summary of the Invention

[0005] In order to improve the problem that the system automation level is limited and key links still require manual intervention based on historical experience, resulting in insufficient dynamic matching between heating supply and actual demand, this application provides a trusted intelligent control method and system for heating based on full trust innovation.

[0006] In the first aspect, the present application provides a reliable intelligent control method for heating based on full trust innovation, which adopts the following technical solutions: including:

[0007] The IoT devices deployed at the heat source, pipe network, and user end collect heating parameters in real time, and pre-process the heating parameters through the edge computing node to obtain pre-processed data;

[0008] Based on the all-innovation platform, the heat load of the pre-processed data is predicted by a bidirectional LSTM neural network model, and a heat load prediction curve for the future time period is output;

[0009] Based on the heat load prediction curve, combined with real-time equipment status and energy consumption optimization strategy, dynamic scheduling instructions for heat source output, pipe network flow and user-end water supply temperature are generated;

[0010] The dynamic scheduling instructions are sent to the edge computing node to automatically adjust the operating parameters of the heat source, pipeline network and user end. When equipment abnormalities are detected, the backup equipment is automatically switched or the heating path is adjusted to realize unmanned control of the heating system.

[0011] Preferably, the real-time collection of heating parameters by IoT devices deployed at the heat source, pipe network and user end, and preprocessing of the heating parameters by edge computing nodes to obtain preprocessed data include:

[0012] Temperature sensors, pressure sensors, and combustion efficiency monitoring equipment are deployed at the heat source end; flow sensors, pressure transmitters, and valve status monitoring equipment are deployed at the pipe network end; and room temperature sensors and heat dissipation equipment operation status collection devices are deployed at the user end. These sensors collect parameters such as heat source outlet temperature, pipe network flow, user indoor temperature, and user behavior, and generate multi-dimensional heating parameters.

[0013] The collected heating parameters are cleaned to remove abnormal jump values, time aligned, and normalized, and missing data are filled with linear interpolation or historical mean to form pre-processed data with continuous time and consistent dimensions.

[0014] Preferably, the heat load prediction of the pre-processed data is performed based on the full-innovation platform through a bidirectional LSTM neural network model, and a heat load prediction curve for a future time period is output, including:

[0015] The pre-processed multi-dimensional time series data is organized into time windows of fixed length in chronological order, wherein the pre-processed multi-dimensional time series data includes: historical heating parameters, environmental parameters of the synchronization section, and historical load data, to obtain initial input data;

[0016] Calculating the similarity between the user behavior parameter at each time point and the historical load data at the corresponding time point, and normalizing the similarity to obtain the attention coefficient of the user behavior parameter at each time point;

[0017] Multiplying the user behavior parameter by the corresponding attention coefficient to obtain a dynamically adjusted user behavior feature;

[0018] The dynamically adjusted user behavior characteristics are combined with the heating parameters and the historical load data in chronological order to form final input data integrating the dynamic weights of user behavior;

[0019] Based on the final input data, the forward LSTM layer in the bidirectional LSTM neural network model analyzes the data in chronological order, and the reverse LSTM layer analyzes the data in reverse chronological order, capturing the front-to-back dependency of the time series and merging the hidden states extracted in the two time series directions to obtain a time series feature that integrates the front-to-back dependency and the dynamic weight of user behavior;

[0020] Based on the time series characteristics, a heat load prediction value in a future time period is output through linear transformation to form a continuous heat load prediction curve.

[0021] Preferably, after performing heat load prediction on the pre-processed data through a bidirectional LSTM neural network model based on the all-innovation platform and outputting a heat load prediction curve for a future time period, the method further includes:

[0022] When there are multiple heat sources, a multi-objective optimization model is constructed based on the marginal cost of each heat source and the heat load prediction curve;

[0023] Minimizing the total operating cost is used as the objective function, and the constraints include the upper limit of the pipe network pressure and the maximum output of each heat source;

[0024] The optimal heat source output allocation scheme is solved using the particle swarm optimization algorithm.

[0025] Preferably, after sending the dynamic scheduling instruction to the edge computing node and automatically adjusting the operating parameters of the heat source, the pipe network and the user end, the method further includes:

[0026] Build user behavior patterns by analyzing user historical usage data;

[0027] When the predicted heat load matches the user behavior pattern, the user-side water supply temperature is automatically adjusted.

[0028] Preferably, when an equipment abnormality is detected, the backup equipment is automatically switched or the heating path is adjusted to achieve unmanned control of the heating system, including:

[0029] If an abnormality is detected in the heat source equipment, the backup unit with the same heat source will be activated first;

[0030] If an abnormality is detected in key node equipment of the pipe network, the abnormal section valve will be closed and the bypass pipeline will be activated to maintain normal heating in other areas;

[0031] If an abnormality is detected in the user-end equipment, the temperature loss will be compensated by increasing the heating parameters of adjacent users.

[0032] In the second aspect, the present application provides a heat supply trustworthy intelligent control device based on full trust creation, which adopts the following technical solutions, including:

[0033] The acquisition module is used to collect heating parameters in real time through IoT devices deployed at the heat source, pipe network, and user end, and pre-process the heating parameters through the edge computing node to obtain pre-processed data;

[0034] A prediction module is used to perform heat load prediction on the preprocessed data through a bidirectional LSTM neural network model based on the all-innovation platform, and output a heat load prediction curve for a future time period;

[0035] An optimization module is used to generate dynamic scheduling instructions for heat source output, pipe network flow and user-end water supply temperature based on the heat load prediction curve, combined with real-time equipment status and energy consumption optimization strategy;

[0036] The scheduling module is used to send the dynamic scheduling instructions to the edge computing node, automatically adjust the operating parameters of the heat source, pipeline network and user end, and automatically switch to backup equipment or adjust the heating path when equipment abnormality is detected, to achieve unmanned control of the heating system.

[0037] In a third aspect, the present application further provides a control device, comprising:

[0038] It includes a memory and a processor, and the memory stores a computer program that can be loaded and executed by the processor, such as the above-mentioned trusted intelligent control method for heating based on full trust and innovation.

[0039] In a fourth aspect, the present application also provides a computer-readable storage medium storing a computer program that can be loaded and executed by a processor such as the above-mentioned trusted intelligent control method for heating based on full trust innovation.

[0040] To summarize, in this application, multi-dimensional heating parameters of heat sources, pipelines and user terminals are collected in real time through IoT devices, and data quality is improved through edge computing preprocessing; relying on the bidirectional LSTM model of the all-innovation platform, the dynamic weight of user behavior parameters is combined to optimize the heat load prediction, and accurately capture the association between time dependence and user heating patterns; based on the prediction results and energy consumption optimization strategy, dynamic scheduling instructions are generated to realize intelligent regulation of heat source output, pipeline flow and user terminal temperature; when the equipment is abnormal, it automatically switches to the backup or adjusts the heating path to ensure continuous operation of the system; at the same time, the national secret algorithm is used to encrypt transmission data, authority control and anomaly detection to ensure that the entire process is autonomous, controllable, safe and reliable, thereby achieving a dynamic match between the heat supply and actual demand, and realizing unmanned and precise management of the heating system from data collection, prediction to regulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of a trustworthy intelligent control method for heating based on full trust creation.

[0042] Figure 2It is a structural block diagram of a heating trustworthy intelligent control device based on full trust creation. DETAILED DESCRIPTION

[0043] The following combination Figure 1 - Figure 2 This application is described in further detail.

[0044] The all-innovation architecture is an information technology application innovation system built based on independent innovation technology, covering the entire domestic ecosystem from underlying hardware to upper-level software. The core includes domestic hardware, basic software, application software and security system. Through the coordinated optimization and adaptation of software and hardware, it achieves independent controllable technology and secure and reliable data, avoiding dependence on foreign technology in key areas. It is widely used in core scenarios such as energy, finance, and government affairs, providing stable, secure and efficient digital support for critical information infrastructure.

[0045] This application is based on a fully trusted innovation architecture, integrating edge computing, Internet of Things, trusted transmission and industrial control security capabilities to achieve full process coverage of heating scene data security collection, intelligent analysis to trusted transmission protection, and improve the unattended scheduling and operation capabilities of heating production operations.

[0046] Reference Figure 1 , the embodiment of the present application includes at least steps S10 to S40.

[0047] S10, collects heating parameters in real time through IoT devices deployed at heat sources, pipe networks and user terminals, and preprocesses the heating parameters through edge computing nodes to obtain preprocessed data.

[0048] S20, based on the all-innovation platform, uses a bidirectional LSTM neural network model to predict the heat load of preprocessed data and outputs the heat load prediction curve for the future time period.

[0049] S30, based on the heat load prediction curve, combined with the real-time equipment status and energy consumption optimization strategy, generates dynamic scheduling instructions for heat source output, pipe network flow and user-end water supply temperature.

[0050] S40 sends dynamic scheduling instructions to edge computing nodes, automatically adjusts the operating parameters of the heat source, pipeline network and user end, and automatically switches to backup equipment or adjusts the heating path when equipment abnormalities are detected, realizing unmanned control of the heating system.

[0051] Specifically, the heating parameters of heat sources, pipe networks and user terminals are collected in real time through IoT devices, and pre-processed by edge computing to improve data quality and provide reliable input for subsequent analysis; the bidirectional LSTM model based on the full-innovation platform predicts the heat load of the pre-processed data, outputs the load curve of the future cycle, and accurately captures the timing dependency; dynamic scheduling instructions are generated by combining equipment status and energy consumption optimization strategies to guide the intelligent adjustment of heat source output, pipe network flow and user terminal temperature; in the event of an abnormality, standby equipment is automatically switched or the heating path is adjusted to ensure the continuous operation of the system, ultimately realizing unmanned and precise management of the heating system from data collection and prediction to regulation, balancing energy efficiency and user heating demand, and improving the reliability and intelligence of the heating system.

[0052] In some embodiments, step S10 specifically includes the following steps: deploying temperature sensors, pressure sensors and combustion efficiency monitoring equipment at the heat source end, deploying flow sensors, pressure transmitters and valve status monitoring equipment at the pipeline end, and deploying room temperature sensors and heat dissipation equipment operation status acquisition devices at the user end, respectively collecting parameters of the heat source outlet temperature, pipeline flow, user indoor temperature and user behavior to generate multi-dimensional heating parameters; performing data cleaning on the collected heating parameters to eliminate abnormal jump values, time alignment, and normalization processing, and using linear interpolation or historical mean filling for missing data to form time-continuous and dimensionally consistent pre-processed data.

[0053] Specifically, multiple types of sensors and monitoring equipment are deployed at the heat source, pipeline network, and user end to comprehensively collect multi-dimensional heating parameters such as heat source outlet temperature, pipeline network flow, user indoor temperature, and user behavior, covering the operating status of all links of the system; abnormal jump values ​​are eliminated through data cleaning to avoid noise interference, time alignment is used to unify the time dimension of the data, normalization is used to eliminate dimensional differences, and linear interpolation or historical mean is used to fill in missing data, ultimately forming pre-processed data that is continuous in time and consistent in dimension, providing a standardized and high-quality input basis for subsequent heat load prediction models, ensuring the accuracy of the analysis results and the reliability of the control strategy.

[0054] In some embodiments, step S20 specifically includes the following steps: organizing the preprocessed multidimensional time series data into a time window of fixed length in chronological order, the preprocessed multidimensional time series data including: historical heating parameters, environmental parameters of the synchronization segment and historical load data, to obtain initial input data; calculating the similarity between the user behavior parameters at each time point and the historical load data at the corresponding time point, and normalizing the similarity to obtain the attention coefficient of the user behavior parameters at each time point; multiplying the user behavior parameters by the corresponding attention coefficient to obtain the dynamically adjusted user behavior characteristics; splicing the dynamically adjusted user behavior characteristics with the heating parameters and historical load numbers in chronological order to form the final input data of the final fusion user behavior dynamic weight; based on the final input data, analyzing the data in chronological order through the forward LSTM layer in the bidirectional LSTM neural network model and analyzing the data in reverse chronological order through the reverse LSTM layer to capture the forward and backward dependencies of the time series, and merging the hidden states extracted in the two time series directions to obtain the time series characteristics of the fusion of the forward and backward dependencies and the dynamic weights of the user behavior; outputting the heat load prediction value in the future time period through linear transformation based on the time series characteristics to form a continuous heat load prediction curve.

[0055] Specifically, the preprocessed multi-dimensional time series data is organized as the initial input of a fixed time window to ensure time series continuity; the similarity between user behavior parameters and historical loads is calculated and normalized to generate the attention coefficient, and the weights of user behavior features are dynamically adjusted to highlight the impact of key user behaviors on the load; the final input data of the fused dynamic weights is spliced ​​together, and the forward (in chronological order) and reverse (in reverse chronological order) analysis of the bidirectional LSTM is combined to comprehensively capture the forward and backward dependencies of the time series, and extract the characteristics of the fused time series association and the dynamic impact of user behavior; finally, a continuous heat load forecast curve is output through linear transformation, providing data support for accurately predicting future heat demand and formulating dynamic scheduling strategies.

[0056] In some embodiments, the corresponding processing steps are as follows: when there are multiple heat sources, a multi-objective optimization model is constructed based on the marginal cost and heat load prediction curve of each heat source; minimizing the total operating cost is used as the objective function, and the constraints include the upper limit of the pipeline pressure and the maximum output of each heat source; and the optimal heat source output distribution plan is solved by the particle swarm optimization algorithm.

[0057] Specifically, in the multi-heat source scenario, a multi-objective optimization model is constructed by combining the marginal cost of each heat source with the heat load forecast curve, with minimizing the total operating cost as the objective function. At the same time, constraints such as the upper limit of pipeline pressure and the maximum output of the heat source are incorporated to ensure the safety and feasibility of the allocation plan; the particle swarm optimization algorithm is used to solve the optimal heat source output allocation plan, dynamically balancing the output ratio of each heat source, which not only reduces the overall operating cost of the system, but also avoids the operational risks caused by excessive load of a single heat source or excessive pipeline pressure, and realizes the economy and reliability of coordinated regulation of multiple heat sources.

[0058] In some embodiments, the corresponding processing steps are as follows: constructing a user behavior pattern by analyzing the user's historical heat usage data; and automatically adjusting the user-end water supply temperature when the predicted heat load matches the user behavior pattern.

[0059] Furthermore: if an abnormality is detected in the heat source equipment, the backup unit with the same heat source will be activated first; if an abnormality is detected in the key node equipment of the pipeline network, the normal heating supply to other areas will be maintained by closing the abnormal section valve and activating the bypass pipeline; if an abnormality is detected in the user-end equipment, the temperature loss will be compensated by increasing the heating parameters of adjacent users.

[0060] Specifically, by analyzing the user's historical heat usage data to build a behavioral pattern, combining the predicted heat load with the pattern matching results, the user-side water supply temperature is automatically adjusted to achieve precise adaptation of temperature control to the user's actual heat demand, improving heat comfort and energy efficiency; for different abnormal scenarios, graded response - when the heat source equipment is abnormal, the same heat source standby unit is activated to ensure heat supply; when the key node of the pipeline network is abnormal, the abnormal section valve is closed and the bypass pipeline is activated to maintain regional heating; when the user-side equipment is abnormal, the heating parameters of adjacent users are increased to compensate for temperature loss, quickly blocking the spread of abnormal impacts, and maximizing the guarantee of continuous and stable operation of the heating system.

[0061] In some embodiments, the corresponding processing steps are as follows: use the national secret algorithm to perform signature verification and hash verification on the transmission data between the cloud and the edge computing node; limit the user's operation rights to key equipment through role authority control, and identify pipeline leaks, equipment failures or illegal heat use based on predictive residual analysis and isolation forest algorithm, generate alarm information and push it to the operation and maintenance terminal.

[0062] Specifically, the use of national secret algorithms to perform signature verification and hash verification on data transmitted between the cloud and edge computing nodes can ensure the integrity, authenticity and tamper-resistance of data transmission, and enhance communication security; through role-based authority control, the operating permissions of key equipment can be restricted to avoid system risks caused by unauthorized or unauthorized operations; combined with predictive residual analysis and isolation forest algorithm, it can efficiently identify abnormal behaviors such as pipeline leakage, equipment failure or illegal use of heat, generate alarm information and push it to the operation and maintenance terminal, realize timely warning and rapid disposal of abnormalities, comprehensively improve the data security protection capabilities, operational compliance and abnormal response efficiency of the heating system, and ensure stable operation of the system.

[0063] The implementation principle of a trusted intelligent control method for heating based on full-innovation in the embodiment of the present application is as follows: in this application, multi-dimensional heating parameters of heat sources, pipelines and user terminals are collected in real time through Internet of Things devices, and data quality is improved through edge computing preprocessing; relying on the bidirectional LSTM model of the full-innovation platform, the heat load prediction is optimized in combination with the dynamic weight of user behavior parameters to accurately capture the association between time dependency and user heating mode; dynamic scheduling instructions are generated based on the prediction results and energy consumption optimization strategy to realize intelligent regulation of heat source output, pipeline flow and user terminal temperature; when the equipment is abnormal, the standby or heating path is automatically switched to another one or adjusted to ensure continuous operation of the system; at the same time, the national secret algorithm is used to encrypt transmission data, authority control and anomaly detection to ensure that the entire process is autonomous, controllable, safe and reliable, thereby achieving a dynamic matching degree between the heat supply and actual demand, and realizing unmanned and precise management of the heating system from data collection, prediction to regulation.

[0064] Figure 1 The figure is a flow chart of a heat supply trustworthy intelligent control method based on full trust creation in one embodiment. It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows; unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps may be executed in other orders; and Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0065] Based on the same technical concept, Figure 2 The embodiment of the present application also provides a heating trustworthy intelligent control device based on full trust creation, which adopts the following technical solution. The device includes:

[0066] The acquisition module is used to collect heating parameters in real time through IoT devices deployed at the heat source, pipe network, and user end, and pre-process the heating parameters through the edge computing node to obtain pre-processed data;

[0067] The prediction module is used to predict the heat load of pre-processed data using a bidirectional LSTM neural network model based on the Quanxinchuang platform, and output the heat load prediction curve for the future time period;

[0068] The optimization module is used to generate dynamic scheduling instructions for heat source output, pipe network flow, and user-end water supply temperature based on the heat load prediction curve, combined with real-time equipment status and energy consumption optimization strategy;

[0069] The scheduling module is used to send dynamic scheduling instructions to edge computing nodes, automatically adjust the operating parameters of the heat source, pipeline network and user end, and automatically switch to backup equipment or adjust the heating path when equipment abnormalities are detected, realizing unmanned control of the heating system.

[0070] In some embodiments, the acquisition module is specifically used to deploy temperature sensors, pressure sensors, and combustion efficiency monitoring equipment at the heat source end, deploy flow sensors, pressure transmitters, and valve status monitoring equipment at the pipe network end, and deploy room temperature sensors and heat dissipation equipment operation status acquisition devices at the user end, respectively collecting parameters such as heat source outlet temperature, pipe network flow, user indoor temperature, and user behavior, to generate multi-dimensional heating parameters;

[0071] The collected heating parameters are cleaned to remove abnormal jump values, time aligned, and normalized, and missing data are filled with linear interpolation or historical mean to form pre-processed data with continuous time and consistent dimensions.

[0072] In some embodiments, the prediction module is specifically configured to organize the pre-processed multi-dimensional time series data into time windows of fixed length in chronological order, the pre-processed multi-dimensional time series data including: historical heating parameters, environmental parameters of the synchronization segment, and historical load data, to obtain initial input data;

[0073] Calculate the similarity between the user behavior parameters at each time point and the historical load data at the corresponding time point, and normalize the similarity to obtain the attention coefficient of the user behavior parameters at each time point;

[0074] Multiply the user behavior parameter by the corresponding attention coefficient to obtain the dynamically adjusted user behavior characteristics;

[0075] The dynamically adjusted user behavior characteristics are spliced ​​with heating parameters and historical load data in chronological order to form the final input data that integrates the dynamic weights of user behavior;

[0076] Based on the final input data, the forward LSTM layer in the bidirectional LSTM neural network model analyzes the data in chronological order, and the reverse LSTM layer analyzes the data in reverse chronological order. This captures the temporal dependencies between the time series and merges the hidden states extracted from the two temporal directions to obtain temporal features that integrate the temporal dependencies with the dynamic weights of user behavior.

[0077] Based on the time series characteristics, the heat load forecast value in the future time period is output through linear transformation to form a continuous heat load forecast curve.

[0078] In some embodiments, the scheduling module is specifically used to construct a multi-objective optimization model based on the marginal cost and heat load prediction curve of each heat source when there are multiple heat sources;

[0079] Minimizing the total operating cost is used as the objective function, and the constraints include the upper limit of the pipe network pressure and the maximum output of each heat source;

[0080] The optimal heat source output allocation scheme is solved using the particle swarm optimization algorithm.

[0081] In some embodiments, the scheduling module is further used to construct user behavior patterns by analyzing historical user usage data;

[0082] When the predicted heat load matches the user behavior pattern, the user-side water supply temperature is automatically adjusted.

[0083] In some embodiments, the scheduling module is further configured to prioritize the use of a backup unit with the same heat source if an abnormality in the heat source equipment is detected;

[0084] If an abnormality is detected in key node equipment of the pipe network, the abnormal section valve will be closed and the bypass pipeline will be activated to maintain normal heating in other areas;

[0085] If an abnormality is detected in the user-end equipment, the temperature loss will be compensated by increasing the heating parameters of adjacent users.

[0086] In some embodiments, the scheduling module is further configured to perform signature verification and hash verification on data transmitted between the cloud and the edge computing node using a national secret algorithm;

[0087] Through role-based permission control, users' operating rights to key equipment are restricted. Based on predictive residual analysis and isolation forest algorithm, pipeline leaks, equipment failures or illegal heat use behaviors are identified, and alarm information is generated and pushed to the operation and maintenance terminal.

[0088] The embodiment of the present application also discloses a control device.

[0089] Specifically, the control device includes a memory and a processor, and the memory stores a computer program that can be loaded and executed by the processor for the above-mentioned trusted intelligent control method for heating based on full-innovation.

[0090] The embodiment of the present application also discloses a computer-readable storage medium.

[0091] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed such as the above-mentioned trusted intelligent control method for heating based on full trust creation. The computer-readable storage medium includes, for example: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0092] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A reliable intelligent control method for heating based on full trust creation, characterized in that: include: The IoT devices deployed at the heat source, pipe network, and user end collect heating parameters in real time, and pre-process the heating parameters through the edge computing node to obtain pre-processed data; Based on the all-innovation platform, the heat load of the pre-processed data is predicted by a bidirectional LSTM neural network model, and a heat load prediction curve for the future time period is output; Based on the heat load prediction curve, combined with real-time equipment status and energy consumption optimization strategy, dynamic scheduling instructions for heat source output, pipe network flow and user-end water supply temperature are generated; The dynamic scheduling instructions are sent to the edge computing node to automatically adjust the operating parameters of the heat source, pipeline network and user end. When equipment abnormalities are detected, the backup equipment is automatically switched or the heating path is adjusted to realize unmanned control of the heating system.

2. A heat supply trustworthy intelligent control method based on full trust creation according to claim 1, characterized in that: The IoT devices deployed at the heat source, pipe network, and user end collect heating parameters in real time, and pre-process the heating parameters through the edge computing node to obtain pre-processed data, including: Temperature sensors, pressure sensors, and combustion efficiency monitoring equipment are deployed at the heat source end; flow sensors, pressure transmitters, and valve status monitoring equipment are deployed at the pipe network end; and room temperature sensors and heat dissipation equipment operation status collection devices are deployed at the user end. These sensors collect parameters such as heat source outlet temperature, pipe network flow, user indoor temperature, and user behavior, and generate multi-dimensional heating parameters. The collected heating parameters are cleaned to remove abnormal jump values, time aligned, and normalized, and missing data are filled with linear interpolation or historical mean to form pre-processed data with continuous time and consistent dimensions.

3. A heat supply trustworthy intelligent control method based on full trust creation according to claim 2, characterized in that: Based on the all-innovation platform, the pre-processed data is subjected to heat load prediction through a bidirectional LSTM neural network model, and a heat load prediction curve for a future time period is output, including: The pre-processed multi-dimensional time series data is organized into time windows of fixed length in chronological order, wherein the pre-processed multi-dimensional time series data includes: historical heating parameters, environmental parameters of the synchronization section, and historical load data, to obtain initial input data; Calculating the similarity between the user behavior parameter at each time point and the historical load data at the corresponding time point, and normalizing the similarity to obtain the attention coefficient of the user behavior parameter at each time point; Multiplying the user behavior parameter by the corresponding attention coefficient to obtain a dynamically adjusted user behavior feature; The dynamically adjusted user behavior characteristics are combined with the heating parameters and the historical load data in chronological order to form final input data integrating the dynamic weights of user behavior; Based on the final input data, the forward LSTM layer in the bidirectional LSTM neural network model analyzes the data in chronological order, and the reverse LSTM layer analyzes the data in reverse chronological order, capturing the front-to-back dependency of the time series and merging the hidden states extracted in the two time series directions to obtain a time series feature that integrates the front-to-back dependency and the dynamic weight of user behavior; Based on the time series characteristics, a heat load prediction value in a future time period is output through linear transformation to form a continuous heat load prediction curve.

4. A heat supply trustworthy intelligent control method based on full trust creation according to claim 2, characterized in that: After performing heat load prediction on the pre-processed data using a bidirectional LSTM neural network model based on the all-innovation platform and outputting a heat load prediction curve for a future time period, the method further includes: When there are multiple heat sources, a multi-objective optimization model is constructed based on the marginal cost of each heat source and the heat load prediction curve; Minimizing the total operating cost is used as the objective function, and the constraints include the upper limit of the pipe network pressure and the maximum output of each heat source; The optimal heat source output allocation scheme is solved using the particle swarm optimization algorithm.

5. The heat supply trustworthy intelligent control method based on full trust creation according to claim 2 is characterized in that: After sending the dynamic scheduling instruction to the edge computing node and automatically adjusting the operating parameters of the heat source, the pipe network, and the user end, the method further includes: Build user behavior patterns by analyzing user historical usage data; When the predicted heat load matches the user behavior pattern, the user-side water supply temperature is automatically adjusted.

6. A heat supply trustworthy intelligent control method based on full trust creation according to claim 2, characterized in that: When an equipment anomaly is detected, the system automatically switches to a backup device or adjusts the heating path to achieve unmanned control of the heating system, including: If an abnormality is detected in the heat source equipment, the backup unit with the same heat source will be activated first; If an abnormality is detected in key node equipment of the pipe network, the abnormal section valve will be closed and the bypass pipeline will be activated to maintain normal heating in other areas; If an abnormality is detected in the user-end equipment, the temperature loss will be compensated by increasing the heating parameters of adjacent users.

7. The heat supply trustworthy intelligent control method based on full trust creation according to claim 1 is characterized in that: The method further comprises: Use national secret algorithms to perform signature verification and hash verification on data transmitted between the cloud and edge computing nodes; Through role-based permission control, users' operating rights to key equipment are restricted. Based on predictive residual analysis and isolation forest algorithm, pipeline leaks, equipment failures or illegal heat use behaviors are identified, and alarm information is generated and pushed to the operation and maintenance terminal.

8. A reliable intelligent heating control device based on full trust innovation, characterized in that: The device comprises: The acquisition module is used to collect heating parameters in real time through IoT devices deployed at the heat source, pipe network, and user end, and pre-process the heating parameters through the edge computing node to obtain pre-processed data; A prediction module is used to perform heat load prediction on the preprocessed data through a bidirectional LSTM neural network model based on the all-innovation platform, and output a heat load prediction curve for a future time period; An optimization module is used to generate dynamic scheduling instructions for heat source output, pipe network flow and user-end water supply temperature based on the heat load prediction curve, combined with real-time equipment status and energy consumption optimization strategy; The scheduling module is used to send the dynamic scheduling instructions to the edge computing node, automatically adjust the operating parameters of the heat source, pipeline network and user end, and automatically switch to backup equipment or adjust the heating path when equipment abnormality is detected, to achieve unmanned control of the heating system.

9. A control device, characterized in that: The device comprises: The method comprises a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that A computer program is stored which can be loaded by a processor and execute the method according to any one of claims 1 to 7.

Images