A heating energy-saving optimization management method and device driven by intelligent big data

By establishing the change relationship curve of heating data and the LSTM model, and optimizing the heating control parameters, the problem of insufficient real-time and accuracy of data processing in the prior art is solved, and the heating energy-saving optimization effect is improved.

CN117870008BActive Publication Date: 2024-11-15HUANENG JINAN HUANGTAI POWER GENERATION CO LTD
View PDF 13 Cites 0 Cited by

Patent Information

Application Number
CN202311733197.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-12-14
Publication Date
2024-11-15
Estimated Expiration
2043-12-14

AI Technical Summary

Technical Problem

In the prior art, the real-time and accuracy of data processing are low, resulting in poor heating energy-saving optimization effects.

Method used

By obtaining previous heating data, establishing a change relationship curve of different lengths, determining the target duration; obtaining heating data within the current target duration, extracting data processing requirements; selecting a stream processing framework and setting relevant parameters to process data; establishing an LSTM model, Optimize heating control parameters; adjust control parameters according to heating situation and visually display them.

Benefits of technology

It improves the real-time and accuracy of data processing, optimizes heating control parameters, and improves the optimization effect of heating energy saving.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117870008B_ABST
    Figure CN117870008B_ABST
Patent Text Reader

Abstract

The present invention discloses a heating energy-saving optimization management method and device driven by intelligent big data, which relates to the field of data processing technology, including establishing a change relationship curve of different durations based on previous heating data and evaluation indicators; extracting data processing requirements based on heating data within the current target duration; selecting a stream processing framework according to the data processing requirements, and setting stream processing related parameters to obtain processed data; establishing an LSTM model through the processed data; determining the current heating environment, and adjusting the heating control parameters according to the current heating environment; performing heating control according to the adjusted heating control parameters, establishing a regional thermal map, and performing visual display. This application improves the real-time and accuracy of real-time data processing through a stream processing framework and stream processing related parameters, divides the data, establishes an LSTM model, optimizes the heating control parameters, and adjusts the heating control parameters through the heating environment, thereby improving the heating energy-saving optimization effect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and more specifically, to an intelligent big data driven heating energy-saving optimization management method and device. Background Art

[0002] Intelligent big data-driven heating energy-saving optimization management integrates key technologies such as big data technology, real-time data processing, Internet of Things and sensors, machine learning, big data visualization, optimization algorithms and intelligent control, security and privacy protection, automation and remote monitoring. Through the synergy of these technologies, the system can monitor the status of the heating system in real time, predict demand, adjust parameters, improve energy efficiency, reduce costs, and ensure data security and privacy. This builds an intelligent, efficient and safe heating system that meets user needs and promotes energy-saving optimization.

[0003] In the prior art, the real-time and accuracy of data processing are low, resulting in poor heating energy-saving optimization effects.

[0004] Therefore, how to improve the real-time and accuracy of data processing is a technical problem that needs to be solved. Summary of the invention

[0005] The present invention provides an intelligent big data driven heating energy saving optimization management method to solve the technical problems of poor real-time performance and accuracy of data processing in the prior art. The method comprises:

[0006] Obtain previous heating data, and establish a change relationship curve of different durations based on previous heating data and evaluation indicators, so as to determine the target duration;

[0007] Obtain heating data within the current target duration, and extract data processing requirements based on the heating data within the current target duration;

[0008] Select a stream processing framework based on data processing requirements and set stream processing related parameters to obtain processed data;

[0009] The LSTM model is established through the processed data to obtain the optimized heating control parameters;

[0010] Determine the current heating environment and adjust the heating control parameters according to the current heating environment;

[0011] Perform heating control according to the adjusted heating control parameters, establish a regional thermal map, and perform visual display.

[0012] In some embodiments of the present application, a change relationship curve of different durations is established based on previous heating data and evaluation indicators to determine the target duration, including:

[0013] Evaluation indicators include energy-saving indicators and other indicators;

[0014] Determine the comprehensive index based on the energy-saving index and other indexes, take each previous heating data as the horizontal axis and the comprehensive index as the vertical axis, and construct the change relationship curve under different standard time lengths. Each standard time length corresponds to a change relationship curve;

[0015] Determine linear correlation or nonlinear correlation based on the change relationship curve, and determine the degree of correlation based on the linear correlation or nonlinear correlation;

[0016] Compare the correlations of the same type of past heating data under different standard durations, and use the standard duration with the greatest correlation as the only standard duration for the same type of past heating data;

[0017] Compare the unique standard duration of each type of previous heating data, and calculate the average standard duration based on all unique standard durations;

[0018] The average of the standard durations is used as the target duration.

[0019] In some embodiments of the present application, a comprehensive index is determined based on the energy-saving index and other indexes, including:

[0020] ;

[0021] Among them, Q is a comprehensive indicator, is the initial weight of the energy-saving index, exp represents the exponential function, n is the number of other indicators, is the weight corresponding to the i-th other indicator, is the i-th other indicator parameter, Y is a fixed threshold, k is a preset constant, It is the energy saving index parameter.

[0022] In some embodiments of the present application, data processing requirements are extracted based on the heating data within the current target duration, including:

[0023] Calculate the data generation rate of heating data within the current target duration, and evaluate the data complexity, timeliness requirements, and fault tolerance requirements.

[0024] In some embodiments of the present application, a stream processing framework is selected according to data processing requirements, and stream processing related parameters are set to obtain processed data, including:

[0025] Select the stream processing framework and initial window size based on data generation rate, data complexity, timeliness requirements, and fault tolerance requirements;

[0026] According to the stream processing framework, sensor data or other real-time data sources are connected to the stream processing system, and window operations are performed according to the initial window size to perform real-time data processing;

[0027] At the same time, monitor the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements, and assign different weights to the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements respectively;

[0028] According to the real-time traffic change, real-time data generation rate, real-time data complexity, real-time timeliness requirement and corresponding weights, the corresponding thresholds of real-time traffic change, real-time data generation rate, real-time data complexity and real-time timeliness requirement are set;

[0029] If a threshold is triggered, the window size is adjusted based on the difference between the real-time traffic change or the real-time data generation rate or the real-time data complexity or the real-time timeliness requirement and the corresponding threshold;

[0030] Complete stream processing and obtain processed data.

[0031] In some embodiments of the present application, an LSTM model is established through processed data, including:

[0032] Arrange the processed data in chronological order to obtain a data time series;

[0033] Dividing the data time series into a first data time series and a second data time series according to an initial ratio;

[0034] Obtaining a fitting interval according to a preset overfitting degree and underfitting degree, and determining a first length interval and a second length interval based on the fitting interval;

[0035] Compare the length of the first length interval with the length of the first data time series, and the length of the second length interval with the length of the second data time series;

[0036] If the length of the first data time series is not within the first length interval, the first data time series is expanded or reduced to an endpoint closest to the first length interval;

[0037] If the length of the second data time series is not within the second length interval, the second data time series is expanded or reduced to an endpoint closest to the second length interval;

[0038] If the length of the first data time series is not within the first length interval, and the length of the second data time series is not within the second length interval, the first data time series is expanded or reduced to the endpoint closest to the first length interval, and the remaining data time series is used as the new second data time series. If the length of the new second data time series is not within the second length interval, the data time series is expanded or reduced;

[0039] The first data time series and the second data time series are used as a training set and a test set respectively;

[0040] Build an LSTM model based on the training set and test set.

[0041] In some embodiments of the present application, determining the current heating environment and adjusting the heating control parameters according to the current heating environment include:

[0042] Define the elements of the heating environment, and set up multi-level heating environments according to different intervals of the elements of the heating environment;

[0043] Calculate the current heating and cooling environment elements and their corresponding intervals, thereby determining the current heating and cooling environment level;

[0044] Adjust heating control parameters according to the current heating environment level.

[0045] Correspondingly, the present application also provides an intelligent big data driven heating energy saving optimization management device, the device comprising:

[0046] The first module is used to obtain previous heating data, establish a change relationship curve of different durations based on previous heating data and evaluation indicators, and thus determine the target duration;

[0047] The second module is used to obtain the heating data within the current target duration and extract the data processing requirements based on the heating data within the current target duration;

[0048] The third module is used to select a stream processing framework according to data processing requirements, set stream processing related parameters, and obtain processed data;

[0049] The fourth module is used to establish an LSTM model through the processed data to obtain optimized heating control parameters;

[0050] The fifth module is used to determine the current heating environment and adjust the heating control parameters according to the current heating environment;

[0051] The sixth module is used to control the heating according to the adjusted heating control parameters, establish a regional thermal map, and perform visual display.

[0052] By applying the above technical solutions, previous heating data is obtained, and change relationship curves of different durations are established based on previous heating data and evaluation indicators to determine the target duration; the heating data within the current target duration is obtained, and the data processing requirements are extracted based on the heating data within the current target duration; the stream processing framework is selected according to the data processing requirements, and the stream processing related parameters are set to obtain the processed data; the LSTM model is established through the processed data to obtain the optimized heating control parameters; the current heating environment is determined, and the heating control parameters are adjusted according to the current heating environment; heating control is performed according to the adjusted heating control parameters, and a regional thermal map is established for visual display. This application improves the real-time and accuracy of real-time data processing through the stream processing framework and stream processing related parameters, divides the data, establishes the LSTM model, optimizes the heating control parameters, and adjusts the heating control parameters through the heating environment, thereby improving the heating energy-saving optimization effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0054] Figure 1 A schematic diagram of a process of a heating energy-saving optimization management method driven by intelligent big data proposed in an embodiment of the present invention is shown;

[0055] Figure 2 A structural schematic diagram of a heating energy-saving optimization management device driven by intelligent big data proposed in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0057] The present application embodiment provides a heating energy-saving optimization management method driven by intelligent big data, such as Figure 1 As shown, the method comprises the following steps:

[0058] Step S101, obtaining previous heating data, establishing a change relationship curve of different durations based on the previous heating data and evaluation indicators, and thus determining the target duration.

[0059] In this embodiment, different time lengths of data may affect the delay of data processing. Here, a reasonable time length is screened out to facilitate subsequent data processing.

[0060] In some embodiments of the present application, a change relationship curve of different durations is established based on previous heating data and evaluation indicators to determine the target duration, including:

[0061] Evaluation indicators include energy-saving indicators and other indicators;

[0062] Determine the comprehensive index based on the energy-saving index and other indexes, take each previous heating data as the horizontal axis and the comprehensive index as the vertical axis, and construct the change relationship curve under different standard time lengths. Each standard time length corresponds to a change relationship curve;

[0063] Determine linear correlation or nonlinear correlation based on the change relationship curve, and determine the degree of correlation based on the linear correlation or nonlinear correlation;

[0064] Compare the correlations of the same type of past heating data under different standard durations, and use the standard duration with the greatest correlation as the only standard duration for the same type of past heating data;

[0065] Compare the unique standard duration of each type of previous heating data, and calculate the average standard duration based on all unique standard durations;

[0066] The average of the standard durations is used as the target duration.

[0067] In this embodiment, the average value of the standard duration is calculated based on all unique standard durations. It is not directly divided by the type of heating data, but is obtained based on the different weights of different types.

[0068] In some embodiments of the present application, a comprehensive index is determined based on the energy-saving index and other indexes, including:

[0069] ;

[0070] Among them, Q is a comprehensive indicator, is the initial weight of the energy-saving index, exp represents the exponential function, n is the number of other indicators, is the weight corresponding to the i-th other indicator, is the i-th other indicator parameter, Y is a fixed threshold, k is a preset constant, It is the energy saving index parameter.

[0071] In this embodiment, It indicates the correction of the energy-saving index weight due to the excess of the total impact of other indicators.

[0072] Step S102, obtaining the heating data within the current target duration, and extracting the data processing requirements based on the heating data within the current target duration.

[0073] In some embodiments of the present application, data processing requirements are extracted based on the heating data within the current target duration, including:

[0074] Calculate the data generation rate of heating data within the current target duration, and evaluate the data complexity, timeliness requirements, and fault tolerance requirements.

[0075] In this embodiment, these factors will have an impact on the stream processing framework and parameter settings.

[0076] Step S103, select a stream processing framework according to data processing requirements, and set stream processing related parameters to obtain processed data.

[0077] In this embodiment, an appropriate stream processing framework is selected, such as Apache Flink, Apache Kafka Streams, Apache Storm, etc. These frameworks support the processing and analysis of real-time data streams.

[0078] In stream processing, window operations are a mechanism for aggregating, analyzing, or processing data streams within a limited range. The main purpose of window operations is to allow data streams to be segmented into limited time or size windows so that the processing engine can perform batch processing or real-time analysis on the data in these windows.

[0079] Stream processing technology is of great significance in the field of data processing, especially in application scenarios that require real-time and low latency. The following are some of the important significances of stream processing technology:

[0080] Real-time:

[0081] Stream processing technology can process data in real time, allowing the system to analyze and respond to data as it is generated. This is critical for applications that require fast decision-making and real-time monitoring, such as financial transactions, real-time monitoring, and smart IoT devices.

[0082] Low Latency:

[0083] Stream processing systems have low latency and can complete data processing at millisecond or sub-millisecond levels. This is critical for businesses that require fast feedback and instant decision-making, such as advertising, online games, and intelligent transportation systems.

[0084] Dynamicity:

[0085] Stream processing can adapt to dynamically changing data streams, so it is very useful for application scenarios with constantly changing data characteristics and patterns. This includes systems that need to adapt to environmental changes in real time, such as real-time risk control systems and supply chain management.

[0086] Real-time analytics and insights:

[0087] Through stream processing technology, real-time analysis can be performed in the data stream to extract real-time insights. This enables enterprises to discover trends, identify anomalies and make real-time decisions more quickly, making them more competitive.

[0088] Complex Event Processing:

[0089] Stream processing technology supports complex event processing, which can detect and process complex event patterns in data streams. This is very valuable for applications that need to monitor and respond to complex events in real time, such as network security monitoring and real-time market analysis.

[0090] Resource availability:

[0091] Stream processing systems are usually able to efficiently utilize resources to process large-scale data streams, thereby reducing costs. This is very important for large-scale application scenarios that need to process large amounts of data, such as social media analysis and IoT device data processing.

[0092] Continuous calculation:

[0093] Stream processing allows for computation and analysis of continuous data streams without storing the data on disk. This is very beneficial for real-time analysis of large-scale data, avoiding the data storage and reading overhead in traditional batch processing.

[0094] In some embodiments of the present application, a stream processing framework is selected according to data processing requirements, and stream processing related parameters are set to obtain processed data, including:

[0095] Select the stream processing framework and initial window size based on data generation rate, data complexity, timeliness requirements, and fault tolerance requirements;

[0096] According to the stream processing framework, sensor data or other real-time data sources are connected to the stream processing system, and window operations are performed according to the initial window size to perform real-time data processing;

[0097] At the same time, monitor the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements, and assign different weights to the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements respectively;

[0098] According to the real-time traffic change, real-time data generation rate, real-time data complexity, real-time timeliness requirement and corresponding weights, the corresponding thresholds of real-time traffic change, real-time data generation rate, real-time data complexity and real-time timeliness requirement are set;

[0099] If a threshold is triggered, the window size is adjusted based on the difference between the real-time traffic change or the real-time data generation rate or the real-time data complexity or the real-time timeliness requirement and the corresponding threshold;

[0100] Complete stream processing and obtain processed data.

[0101] In this embodiment, the data generation rate, data complexity, timeliness requirements and fault tolerance requirements are defined, and a stream processing framework that can meet all the conditions is selected, and different stream processing frameworks have different initial window sizes.

[0102] In this embodiment, different weights are assigned according to the different degrees of impact of real-time traffic changes, real-time data generation rate, real-time data complexity, and real-time timeliness requirements on window requirements.

[0103] In this embodiment, the window size is adjusted based on the difference between the real-time traffic change or the real-time data generation rate or the real-time data complexity or the real-time timeliness requirement and the corresponding threshold. Different differences correspond to different correction coefficients, and the adjustment is completed by correcting the coefficient * initial window size.

[0104] Step S104, establishing an LSTM model through the processed data to obtain optimized heating control parameters.

[0105] In this embodiment, the model building process is as follows:

[0106] Divide the training set and test set to train and debug the model.

[0107] Select the model type:

[0108] Objective: Select an appropriate time series forecasting model based on the characteristics of the problem.

[0109] Implementation steps:

[0110] Base model: ARIMA (Autoregressive Integrated Moving Average) is suitable for stationary time series.

[0111] Complex models: Deep learning models such as LSTM (Long Short-Term Memory) or GRU (Gated Recurrent Unit) are suitable for non-linear and complex time series.

[0112] Model training:

[0113] Goal: Train the selected model using the training set.

[0114] Implementation steps:

[0115] Use the historical data in the training set to train the selected model. For deep learning models, some hyperparameter tuning may be required.

[0116] Model Evaluation:

[0117] Objective: Evaluate the performance of the model using the test set.

[0118] Implementation steps:

[0119] Make predictions using the model on the test set.

[0120] Evaluation metrics such as root mean square error (RMSE), mean absolute error (MAE), etc. are used to evaluate the predictive performance of the model.

[0121] Hyperparameter Tuning:

[0122] Goal: Adjust the model’s hyperparameters based on the evaluation results.

[0123] Implementation steps:

[0124] Techniques such as grid search can be used to find the best combination of hyperparameters.

[0125] Predict future values:

[0126] Objective: Use the trained model to make predictions for future time points.

[0127] Implementation steps:

[0128] The model is trained using the entire dataset (both training and test sets) and then predictions are made for future values.

[0129] Model deployment:

[0130] Goal: Deploy the model for real-world use.

[0131] Implementation steps:

[0132] Embed trained models into applications to make new predictions in real time or periodically.

[0133] Continuous monitoring and updating:

[0134] Goal: Continuously monitor model performance and update the model when needed.

[0135] Implementation steps:

[0136] Regularly check the model's predictive accuracy; if model performance degrades, you may need to retrain or tune the model.

[0137] In some embodiments of the present application, an LSTM model is established through processed data, including:

[0138] Arrange the processed data in chronological order to obtain a data time series;

[0139] Dividing the data time series into a first data time series and a second data time series according to an initial ratio;

[0140] Obtaining a fitting interval according to a preset overfitting degree and underfitting degree, and determining a first length interval and a second length interval based on the fitting interval;

[0141] Compare the length of the first length interval with the length of the first data time series, and the length of the second length interval with the length of the second data time series;

[0142] If the length of the first data time series is not within the first length interval, the first data time series is expanded or reduced to an endpoint closest to the first length interval;

[0143] If the length of the second data time series is not within the second length interval, the second data time series is expanded or reduced to an endpoint closest to the second length interval;

[0144] If the length of the first data time series is not within the first length interval, and the length of the second data time series is not within the second length interval, the first data time series is expanded or reduced to the endpoint closest to the first length interval, and the remaining data time series is used as the new second data time series. If the length of the new second data time series is not within the second length interval, the data time series is expanded or reduced;

[0145] The first data time series and the second data time series are used as a training set and a test set respectively;

[0146] Build an LSTM model based on the training set and test set.

[0147] In this embodiment, in order to avoid under-fitting or over-fitting problems of the model, the data time series is divided.

[0148] Step S105, determining the current heating environment, and adjusting the heating control parameters according to the current heating environment.

[0149] In this embodiment, the heating is not only affected by the data situation, but also by some external influences. For example, the following:

[0150] Seasonal changes:

[0151] Heating demand is usually higher in the cold season and relatively lower in the warm season. This results in significant seasonal variation in heating data, and the impact of this cyclical nature needs to be considered.

[0152] Temperature influence:

[0153] The external temperature has a significant impact on the heating system. When the air temperature drops, the heating demand increases. Therefore, temperature data is a key influencing factor in heating data.

[0154] Timing:

[0155] Heating data is usually time series data, recording information such as energy consumption and temperature at different time points or time periods. This requires appropriate modeling and analysis of time series.

[0156] Sudden demand:

[0157] In some cases, the heating system may face unexpected demands, such as additional heating demand under extreme weather conditions. Such unexpected demands need to be considered in system design and optimization.

[0158] Uncertainty:

[0159] Heating data is affected by many factors, including weather, building structure, resident behavior, etc. The uncertainty of these factors makes the prediction and optimization of heating data challenging.

[0160] Equipment operating status:

[0161] Heating data also includes information about the operating status of heating equipment. The normal operation or failure of the equipment will directly affect the heating effect and energy consumption.

[0162] Energy consuming equipment:

[0163] The data may contain information about various energy-consuming devices, such as boilers, radiators, etc. The performance and usage of these devices have a significant impact on the overall efficiency of the heating system.

[0164] Multivariate relationships:

[0165] Heating data often involves relationships between multiple variables, such as temperature, humidity, time, and energy consumption. Modeling these multivariate relationships requires comprehensive consideration of various factors.

[0166] Energy consumption volatility:

[0167] Energy consumption in heating data usually shows a certain degree of volatility, which may be affected by factors such as residents' behavior, industrial activities, etc. Understanding this volatility is important for system optimization.

[0168] The above content is the contextual element.

[0169] In some embodiments of the present application, determining the current heating environment and adjusting the heating control parameters according to the current heating environment include:

[0170] Define the elements of the heating environment, and set up multi-level heating environments according to different intervals of the elements of the heating environment;

[0171] Calculate the current heating and cooling environment elements and their corresponding intervals, thereby determining the current heating and cooling environment level;

[0172] Adjust heating control parameters according to the current heating environment level.

[0173] In this embodiment, different intervals of different heating environment elements together constitute a level of heating environment, and each heating environment corresponds to an adjustment coefficient, and the heating control parameters are adjusted by the adjustment coefficient.

[0174] Step S106, performing heating control according to the adjusted heating control parameters, establishing a regional thermal map, and performing visual display.

[0175] By applying the above technical solutions, previous heating data is obtained, and change relationship curves of different durations are established based on previous heating data and evaluation indicators to determine the target duration; the heating data within the current target duration is obtained, and the data processing requirements are extracted based on the heating data within the current target duration; the stream processing framework is selected according to the data processing requirements, and the stream processing related parameters are set to obtain the processed data; the LSTM model is established through the processed data to obtain the optimized heating control parameters; the current heating environment is determined, and the heating control parameters are adjusted according to the current heating environment; heating control is performed according to the adjusted heating control parameters, and a regional thermal map is established for visual display. This application improves the real-time and accuracy of real-time data processing through the stream processing framework and stream processing related parameters, divides the data, establishes the LSTM model, optimizes the heating control parameters, and adjusts the heating control parameters through the heating environment, thereby improving the heating energy-saving optimization effect.

[0176] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0177] In order to further explain the technical idea of ​​the present invention, the technical solution of the present invention is now described in combination with specific application scenarios.

[0178] Correspondingly, the present application also provides a heating energy-saving optimization management device driven by intelligent big data, such as Figure 2 As shown, the device comprises:

[0179] The first module 201 is used to obtain previous heating data, establish a change relationship curve of different durations based on the previous heating data and evaluation indicators, and thus determine the target duration;

[0180] The second module 202 is used to obtain the heating data within the current target duration, and extract the data processing requirements according to the heating data within the current target duration;

[0181] The third module 203 is used to select a stream processing framework according to data processing requirements, and set stream processing related parameters to obtain processed data;

[0182] The fourth module 204 is used to establish an LSTM model through the processed data to obtain optimized heating control parameters;

[0183] The fifth module 205 is used to determine the current heating environment and adjust the heating control parameters according to the current heating environment;

[0184] The sixth module 206 is used to perform heating control according to the adjusted heating control parameters, establish a regional thermal map, and perform visual display.

[0185] Those skilled in the art will appreciate that the modules in the device in the implementation scenario can be distributed in the device in the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more devices different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple sub-modules.

[0186] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An intelligent big data driven heating energy saving optimization management method, It is characterized in that The method comprises: Obtain previous heating data, and establish a change relationship curve of different durations based on previous heating data and evaluation indicators, so as to determine the target duration; Obtain heating data within the current target duration, and extract data processing requirements based on the heating data within the current target duration; Select a stream processing framework based on data processing requirements and set stream processing related parameters to obtain processed data; The LSTM model is established through the processed data to obtain the optimized heating control parameters; Determine the current heating environment and adjust the heating control parameters according to the current heating environment; Perform heating control according to the adjusted heating control parameters, establish a regional thermal map, and perform visual display; According to the previous heating data and evaluation indicators, the change relationship curve of different durations is established to determine the target duration, including: Evaluation indicators include energy-saving indicators and other indicators; Determine the comprehensive index based on the energy-saving index and other indexes, take each previous heating data as the horizontal axis and the comprehensive index as the vertical axis, and construct the change relationship curve under different standard time lengths. Each standard time length corresponds to a change relationship curve; Determine linear correlation or nonlinear correlation based on the change relationship curve, and determine the degree of correlation based on the linear correlation or nonlinear correlation; Compare the correlations of the same type of past heating data under different standard durations, and use the standard duration with the greatest correlation as the only standard duration for the same type of past heating data; Compare the unique standard duration of each type of previous heating data, and calculate the average standard duration based on all unique standard durations; The average of the standard duration is used as the target duration; Determine comprehensive indicators based on energy-saving indicators and other indicators, including: ; Among them, Q is a comprehensive indicator, is the initial weight of the energy-saving index, exp represents the exponential function, n is the number of other indicators, is the weight corresponding to the i-th other indicator, is the i-th other indicator parameter, Y is a fixed threshold, k is a preset constant, It is the energy saving index parameter; Select a stream processing framework based on data processing requirements and set stream processing related parameters to obtain processed data, including: Select the stream processing framework and initial window size based on data generation rate, data complexity, timeliness requirements, and fault tolerance requirements; According to the stream processing framework, sensor data or other real-time data sources are connected to the stream processing system, and window operations are performed according to the initial window size to perform real-time data processing; At the same time, monitor the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements, and assign different weights to the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements respectively; According to the real-time traffic change, real-time data generation rate, real-time data complexity, real-time timeliness requirement and corresponding weights, the corresponding thresholds of real-time traffic change, real-time data generation rate, real-time data complexity and real-time timeliness requirement are set; If a threshold is triggered, the window size is adjusted based on the difference between the real-time traffic change or the real-time data generation rate or the real-time data complexity or the real-time timeliness requirement and the corresponding threshold; Complete stream processing and obtain processed data; The LSTM model is built using the processed data, including: Arrange the processed data in chronological order to obtain a data time series; Dividing the data time series into a first data time series and a second data time series according to an initial ratio; Obtaining a fitting interval according to a preset overfitting degree and underfitting degree, and determining a first length interval and a second length interval based on the fitting interval; Compare the length of the first length interval with the length of the first data time series, and the length of the second length interval with the length of the second data time series; If the length of the first data time series is not within the first length interval, the first data time series is expanded or reduced to an endpoint closest to the first length interval; If the length of the second data time series is not within the second length interval, the second data time series is expanded or reduced to an endpoint closest to the second length interval; If the length of the first data time series is not within the first length interval, and the length of the second data time series is not within the second length interval, the first data time series is expanded or reduced to the endpoint closest to the first length interval, and the remaining data time series is used as the new second data time series. If the length of the new second data time series is not within the second length interval, the data time series is expanded or reduced; The first data time series and the second data time series are used as a training set and a test set respectively; Build an LSTM model based on the training set and test set.

2. The intelligent big data driven heating energy saving optimization management method according to claim 1, It is characterized in that Extract data processing requirements based on heating data within the current target duration, including: Calculate the data generation rate of heating data within the current target duration, and evaluate the data complexity, timeliness requirements, and fault tolerance requirements.

3. The intelligent big data driven heating energy saving optimization management method according to claim 1, It is characterized in that Determine the current heating environment and adjust the heating control parameters according to the current heating environment, including: Define the elements of the heating environment, and set up multi-level heating environments according to the different intervals of the elements of the heating environment; Calculate the current heating and cooling environment elements and their corresponding intervals, thereby determining the current heating and cooling environment level; Adjust heating control parameters according to the current heating environment level.

4. An intelligent big data driven heating energy saving optimization management device, It is characterized in that The device comprises: The first module is used to obtain previous heating data, establish a change relationship curve of different durations based on previous heating data and evaluation indicators, and thus determine the target duration; According to the previous heating data and evaluation indicators, the change relationship curve of different durations is established to determine the target duration, including: Evaluation indicators include energy-saving indicators and other indicators; Determine the comprehensive index based on the energy-saving index and other indexes, take each previous heating data as the horizontal axis and the comprehensive index as the vertical axis, and construct the change relationship curve under different standard time lengths. Each standard time length corresponds to a change relationship curve; Determine linear correlation or nonlinear correlation based on the change relationship curve, and determine the degree of correlation based on the linear correlation or nonlinear correlation; Compare the correlations of the same type of past heating data under different standard durations, and use the standard duration with the greatest correlation as the only standard duration for the same type of past heating data; Compare the unique standard duration of each type of previous heating data, and calculate the average standard duration based on all unique standard durations; The average of the standard duration is used as the target duration; Determine comprehensive indicators based on energy-saving indicators and other indicators, including: ; Among them, Q is a comprehensive indicator, is the initial weight of the energy-saving index, exp represents the exponential function, n is the number of other indicators, is the weight corresponding to the i-th other indicator, is the i-th other indicator parameter, Y is a fixed threshold, k is a preset constant, It is the energy saving index parameter; The second module is used to obtain the heating data within the current target duration and extract the data processing requirements based on the heating data within the current target duration; The third module is used to select a stream processing framework according to data processing requirements, set stream processing related parameters, and obtain processed data; Select a stream processing framework based on data processing requirements and set stream processing related parameters to obtain processed data, including: Select the stream processing framework and initial window size based on data generation rate, data complexity, timeliness requirements, and fault tolerance requirements; According to the stream processing framework, sensor data or other real-time data sources are connected to the stream processing system, and window operations are performed according to the initial window size to perform real-time data processing; At the same time, monitor the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements, and assign different weights to the real-time traffic changes, real-time data generation rate, real-time data complexity and real-time timeliness requirements respectively; According to the real-time traffic change, real-time data generation rate, real-time data complexity, real-time timeliness requirement and corresponding weights, the corresponding thresholds of real-time traffic change, real-time data generation rate, real-time data complexity and real-time timeliness requirement are set; If a threshold is triggered, the window size is adjusted based on the difference between the real-time traffic change or the real-time data generation rate or the real-time data complexity or the real-time timeliness requirement and the corresponding threshold; Complete stream processing and obtain processed data; The fourth module is used to establish an LSTM model through the processed data to obtain optimized heating control parameters; The LSTM model is built using the processed data, including: Arrange the processed data in chronological order to obtain a data time series; Dividing the data time series into a first data time series and a second data time series according to an initial ratio; Obtaining a fitting interval according to a preset overfitting degree and underfitting degree, and determining a first length interval and a second length interval based on the fitting interval; Compare the length of the first length interval with the length of the first data time series, and the length of the second length interval with the length of the second data time series; If the length of the first data time series is not within the first length interval, the first data time series is expanded or reduced to an endpoint closest to the first length interval; If the length of the second data time series is not within the second length interval, the second data time series is expanded or reduced to an endpoint closest to the second length interval; If the length of the first data time series is not within the first length interval, and the length of the second data time series is not within the second length interval, the first data time series is expanded or reduced to the endpoint closest to the first length interval, and the remaining data time series is used as the new second data time series. If the length of the new second data time series is not within the second length interval, the data time series is expanded or reduced; The first data time series and the second data time series are used as a training set and a test set respectively; Establish LSTM model based on training set and test set; The fifth module is used to determine the current heating environment and adjust the heating control parameters according to the current heating environment; The sixth module is used to control the heating according to the adjusted heating control parameters, establish a regional thermal map, and perform visual display.

Citation Information

Patent Citations

  • Method and system of heat network energy consumption index analysis and early warning based on BP neural network

    CN110332605A

  • Non-interference geothermal direct supply type heating control system and method based on Internet of Things

    CN111062523A

  • Intelligent heat load prediction regulation and control method and system and storage medium

    CN111121150A

  • Heat supply dispatching automation implementation method, system and platform based on digital twinning

    CN113657031A

  • Regional thermal load rolling prediction method based on finite difference working domain division

    CN114117852A