Low-temperature coupling heat pump remote monitoring and data acquisition system based on Internet of Things

Abstract

The invention provides a low-temperature coupling heat pump remote monitoring and data acquisition system based on the Internet of Things, and the system is characterized in that a data acquisition assembly collects key parameters such as water temperature, flow, compressor state and environment temperature in real time through a sensor, and carries out the preliminary processing of the collected key parameters, and obtains the key parameters after the preliminary processing; the data transmission component transmits the key parameters subjected to primary processing to the cloud server in real time through the Internet of Things; the cloud server processes the key parameters by using a machine learning algorithm and optimizes the operation parameters of the compressor; the feedback control assembly automatically adjusts the working state of the low-temperature coupling heat pump according to the analysis result of the cloud server; meanwhile, key parameters are checked in real time through the intelligent terminal, and a state alarm of the low-temperature coupling heat pump is received. According to the invention, continuous monitoring, intelligent analysis and quick response are realized. The system realizes efficient energy use and optimal management, reduces energy consumption, improves the reliability of equipment and prolongs the service life of the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal energy conversion equipment, and in particular to a low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things. Background Art

[0002] A low temperature coupled heat pump (LTHP) is a device for heat transfer, specially designed to extract heat from low temperature sources (such as air, water or underground) and transfer it to a higher temperature environment. It works on the principle of reverse Carnot cycle, and achieves heat transfer by compressing and expanding the refrigerant. A low temperature coupled heat pump is an efficient and environmentally friendly heat transfer device suitable for various heating and cooling scenarios. Its outstanding energy efficiency performance and environmentally friendly characteristics make it an increasingly popular energy solution in modern buildings and industries, and promote the widespread application of renewable energy. However, the operating efficiency and operating capacity of existing low temperature coupled heat pumps need to be further improved.

[0003] Prior art 1, application number: CN 202410470412.5 discloses a simulation visual system for coupling heating of air source heat pump and gas boiler, including: an operation terminal, the operation terminal is used to operate and control the system, the output terminal signal of the operation terminal is connected to a remote computing module for calculating the coupled heating data of air source heat pump and gas boiler, and the output terminal signal of the operation terminal is connected to a data storage database for storing the coupled heating data of air source heat pump and gas boiler. Although the heating mode of coupling heating of air source heat pump and gas boiler is adopted, and the shortcomings of these two heating systems when heating separately are analyzed, and the simulation model of the coupled heating system of air source heat pump and gas boiler is established, the basic heating mode of air source heat pump and gas boiler is determined by combining the heating characteristics of the two heating modes, and a heating system scheme with good economic benefits, low energy consumption and small impact on the environment is obtained. However, the monitoring means of key parameters such as water temperature and flow rate are limited, and the effect of remote monitoring and control cannot be effectively improved, resulting in poor working efficiency of low-temperature coupled heat pump.

[0004] Prior art 2, application number: CN 202310368505.2 discloses a method for fine-grained evaluation of the long-term operating performance of a ground-source heat pump system. It uses a design parameter data group for simulation calculation to establish a database of the ground-based heat exchanger-heat pump-building coupled heat transfer model; establishes a mapping between design parameters and operating results to form a neural network model for fitting the long-term operating performance of the ground-source heat pump system; inputs the design parameters, uses the neural network model to fit and predict the energy efficiency ratio, energy conservation and emission reduction, and ground temperature in the buried pipe area, and finely evaluates the applicability, economy, and environmental protection of the long-term operation of the system. Although it is used to evaluate the energy efficiency ratio, energy conservation and emission reduction, and ground temperature in the buried pipe area of ​​the ground-source heat pump system during its life cycle, it has the advantages of fast calculation speed, high calculation accuracy, and a comprehensive evaluation system. However, the lack of remote control of the heat pump has resulted in the need to improve the level of intelligent control of the equipment.

[0005] Prior art three, application number: CN 202311504768.8 discloses a medium-deep casing buried pipe thermal seepage coupling numerical heat transfer model and its solution method, establishes the medium-deep casing buried pipe and rock and soil thermal seepage coupling control equation; establishes the equation of the fluid in the outer annular cavity of the medium-deep casing buried pipe; establishes the equation of the fluid in the inner pipe of the medium-deep casing buried pipe; obtains the formula for the convective heat transfer coefficient and the formula for the Nusselt number; and obtains the heat transfer model based on the above equations and formulas. Although the numerical heat transfer model that considers complex factors such as rock and soil stratification, geothermal gradient and groundwater seepage can be used to reveal the influence of factors such as groundwater seepage on the heat extraction characteristics of the buried pipe; under the premise of ensuring the calculation accuracy, a method for quickly solving the model is proposed, which greatly improves the calculation efficiency of the numerical heat transfer model and can be used to guide engineering practice. However, the improvement of the operating efficiency and response capability of the heat pump is limited, and there is a lack of effective monitoring of the operating parameters.

[0006] At present, the existing technologies 1, 2 and 3 lack remote real-time monitoring and feedback adjustment functions, and have poor operating efficiency and responsiveness. Therefore, the present invention provides a remote monitoring and data acquisition system for low-temperature coupled heat pumps based on the Internet of Things, which collects and transmits key parameter data such as water temperature, flow rate and compressor status in real time through the Internet of Things technology, and provides remote real-time monitoring and feedback adjustment functions based on data analysis, which effectively improves the system's operating efficiency and responsiveness through real-time monitoring and data analysis optimization. Summary of the invention

[0007] In order to solve the above technical problems, the present invention provides a low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things, comprising:

[0008] The data acquisition component is responsible for real-time acquisition of key parameters, and preliminarily processes the acquired key parameters to obtain the preliminarily processed key parameters;

[0009] The data transmission component is responsible for transmitting the key parameters after preliminary processing to the cloud server in real time through the Internet of Things; the cloud server processes the key parameters and optimizes the operating parameters of the compressor;

[0010] The feedback control component is responsible for automatically adjusting the working state of the low-temperature coupled heat pump according to the analysis results of the cloud server.

[0011] Optional data collection components, including:

[0012] The position screening module is responsible for obtaining the target data of the area to be detected of the low-temperature coupled heat pump, obtaining multiple target points in the area to be detected, the target points correspond to the measurement points of the key data, classifying the multiple measurement points using the classification strategy, and obtaining at least two measurement points of the same target data based on the corresponding relationship;

[0013] The equipment deployment module is responsible for obtaining the key parameter response time of the measurement point according to the historical key parameter change law, and selecting the measurement point with the shortest response time to the target parameter from at least two measurement points with the same target data, i.e., the deployment point;

[0014] The data fusion module is responsible for synchronizing the data of different types of sensors according to timestamps, comprehensively calculating the confidence and characteristics of different types of sensor data, and generating the key parameters for final monitoring.

[0015] Optional, data fusion module, including:

[0016] The time synchronization submodule is responsible for data collection of each sensor within the set time interval, and synchronizes the data of all sensors according to the timestamp to ensure the same time point;

[0017] The feature extraction submodule is responsible for assigning a confidence value to the data of each sensor when it is collected, and deriving the data confidence of each sensor at the current moment based on the standard deviation and mean of historical data; and performing unified feature extraction for each data type;

[0018] The weighted fusion submodule is responsible for setting the weight of each sensor feature, dynamically adjusting the weight according to the response time and confidence, and merging the features of each sensor with the corresponding confidence; it integrates the data results of all sensors and fuses them through weighted sum to obtain the final key parameters.

[0019] Optional data transmission components, including:

[0020] The signal processing module is responsible for realizing the key parameters after the initial processing of the IoT reception. During the IoT transmission, the IoT signal is interfered by the algorithm and the key parameters after the interference suppression processing are transmitted to the cloud server.

[0021] The model training module is responsible for preprocessing the key parameters after interference suppression processing, obtaining the preprocessed key parameters, and dividing them into training sets and test sets according to the proportion;

[0022] The model evaluation module is responsible for using the test set data to evaluate the support vector machine, evaluating the performance of the support vector machine through mean square error and accuracy, and adjusting the support vector machine parameters.

[0023] Optional, signal processing module, including:

[0024] The signal normalization submodule is responsible for defining the normalization scale function for the IoT signal that needs to be processed;

[0025] The interference judgment submodule is responsible for defining the time-frequency positioning strategy when the IoT signal is interfered with, and judging whether there is interference in the IoT signal to be processed;

[0026] The suppression execution submodule is responsible for suppressing the interference component of the IoT signal with interference and outputting the suppressed IoT signal.

[0027] Optional, model training module, including:

[0028] The dataset construction submodule is responsible for labeling the key parameters after preprocessing according to the data source;

[0029] The data set processing submodule is responsible for dividing the data set into training set and test set;

[0030] The training execution submodule is responsible for inputting the training set for training and updating the support vector machine model parameters through the optimization algorithm.

[0031] Optionally, the dataset construction submodule constructs the preprocessed key parameters and annotation information into a dataset.

[0032] Optionally, the data set processing submodule uses center cropping to downsample the preprocessed key parameters in the data set, and inverts or rotates the downsampled preprocessed key parameters according to preset probabilities to expand the preprocessed key parameters; the preprocessed key parameters are expanded to obtain a new data set.

[0033] Optionally, the training execution submodule selects a core function of the support vector machine and sets a penalty parameter C and a kernel parameter.

[0034] Optional, feedback control components, including:

[0035] The instruction sending module is responsible for generating control instructions based on the analysis results and decision making, and sending them to the feedback control component;

[0036] The state adjustment module is responsible for adjusting the working state of the low-temperature coupled heat pump and regulating the compressor state and circulation flow rate after receiving instructions from the cloud;

[0037] The alarm issuing module is responsible for viewing key parameters in real time through the intelligent terminal and monitoring the operating status of the low-temperature coupled heat pump; if the low-temperature coupled heat pump has any abnormality, an alarm notification will be sent in time.

[0038] The data acquisition component of the present invention collects key parameters such as water temperature, flow rate, compressor status and ambient temperature in real time through various sensors (such as temperature sensors, flow sensors, pressure sensors, etc.); performs preliminary processing (such as filtering, correction, etc.) on the collected data to ensure the accuracy and effectiveness of the data; organizes the key parameters after preliminary processing into a format suitable for subsequent transmission to reduce the network burden. Significance achieved: Through real-time and accurate parameter collection, the effectiveness of system monitoring is improved, which helps to detect potential anomalies early; provides accurate and reliable basic data for subsequent data analysis and feedback control, which helps to optimize management and operation strategies. The data transmission component uses Internet of Things technology (such as LoRa, NB-IoT, Wi-Fi, etc.) to achieve safe and rapid transmission of key parameters after preliminary processing to the cloud server; ensure that data can be uploaded in real time to facilitate rapid response. Significance achieved: Through real-time data transmission, the monitoring system can quickly adapt to changes and respond in a timely manner; data can be seamlessly connected between decentralized IoT edge devices and the cloud, enhancing the flexibility and accessibility of the system; the cloud server centrally stores all received key parameter data and provides an effective database management system; analyzes historical data and uses machine learning algorithms to optimize compressor operating parameters to improve energy efficiency and reduce wear; generates real-time reports to identify trends and abnormal patterns and support the decision-making process. Significance achieved: Decisions based on data analysis are more scientific and can effectively improve the operating efficiency and stability of the system; through continuous algorithm learning and optimization, the system can adapt to different operating conditions and environments and improve overall performance. The feedback control component automatically adjusts the working status of the low-temperature coupled heat pump (such as compressor speed, working time, etc.) according to the analysis results of the cloud server; provides an intelligent terminal view, where users can view key parameters at any time and receive alarm notifications about the status of the equipment. Significance achieved: Allowing the system to automatically adjust according to real-time data improves the responsiveness to environmental changes and load changes; through real-time monitoring and instant reminders, the user experience is enhanced, allowing users to be transparent about the status of the equipment at any time, so that maintenance actions can be taken in a timely manner.

[0039] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0040] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0042] Figure 1 This is a block diagram of a remote monitoring and data acquisition system for a low-temperature coupled heat pump based on the Internet of Things in Example 1 of the present invention;

[0043] Figure 2 This is a block diagram of a data acquisition component in Embodiment 2 of the present invention;

[0044] Figure 3 This is a block diagram of a data fusion module in Embodiment 3 of the present invention;

[0045] Figure 4 This is a block diagram of a data transmission component in Embodiment 4 of the present invention;

[0046] Figure 5 This is a block diagram of a signal processing module in Embodiment 5 of the present invention;

[0047] Figure 6 This is a block diagram of a model training module in Embodiment 6 of the present invention;

[0048] Figure 7 This is a block diagram of the training execution submodule in Example 7 of the present invention;

[0049] Figure 8 This is a block diagram of a training set input unit in Embodiment 8 of the present invention;

[0050] Fig. 9 This is a block diagram of a model evaluation module in Embodiment 9 of the present invention;

[0051] Fig.10 This is a block diagram of the feedback control component in Example 10 of the present invention. DETAILED DESCRIPTION

[0052] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0053] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the embodiments of the present application. The singular forms of "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

[0054] When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims. In the description of the present application, it should be understood that the terms "first", "second", "third", etc. are only used to distinguish similar objects, and do not have to be used to describe a specific order or sequence, nor can they be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0055] Example 1: Figure 1 As shown, an embodiment of the present invention provides a low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things, comprising:

[0056] The data acquisition component is responsible for collecting key parameters such as water temperature, flow, compressor status and ambient temperature in real time through sensors, and performing preliminary processing on the collected key parameters to obtain the key parameters after preliminary processing; among which, the preliminary processing includes filtering, verification and abnormal value processing, etc.;

[0057] The data transmission component is responsible for transmitting the key parameters after preliminary processing to the cloud server in real time through the Internet of Things; the cloud server uses machine learning algorithms to process the key parameters and optimize the operating parameters of the compressor;

[0058] The feedback control component is responsible for automatically adjusting the working status of the low-temperature coupled heat pump according to the analysis results of the cloud server; at the same time, it can view key parameters in real time through the smart terminal and receive status alerts of the low-temperature coupled heat pump.

[0059] The working principle and beneficial effects of the above technical solution are as follows: the data acquisition component of this embodiment collects key parameters such as water temperature, flow rate, compressor status and ambient temperature in real time through sensors, and preliminarily processes the collected key parameters to obtain the preliminarily processed key parameters; the data transmission component transmits the preliminarily processed key parameters to the cloud server in real time through the Internet of Things; the cloud server uses machine learning algorithms to process key parameters and optimize the operating parameters of the compressor; the feedback control component automatically adjusts the working state of the low-temperature coupled heat pump according to the analysis results of the cloud server; at the same time, the key parameters are viewed in real time through the smart terminal, and the status alarm of the low-temperature coupled heat pump is received. The data acquisition component of the above scheme collects key parameters such as water temperature, flow rate, compressor status and ambient temperature in real time through various sensors (such as temperature sensors, flow sensors, pressure sensors, etc.); preliminarily processes the collected data (such as filtering, correction, etc.) to ensure the accuracy and effectiveness of the data; and organizes the preliminarily processed key parameters into a format suitable for subsequent transmission to reduce the network burden. The significance achieved: through real-time and accurate parameter acquisition, the effectiveness of system monitoring is improved, which helps to detect potential abnormalities early; it provides accurate and reliable basic data for subsequent data analysis and feedback control, which helps to optimize management and operation strategies. The data transmission component uses IoT technologies (such as LoRa, NB-IoT, Wi-Fi, etc.) to achieve safe and fast transmission of key parameters after preliminary processing to the cloud server; ensure that data can be uploaded in real time to facilitate rapid response. Significance achieved: Through instant data transmission, the monitoring system can quickly adapt to changes and respond in a timely manner; data can be seamlessly connected between decentralized IoT edge devices and the cloud, enhancing the flexibility and accessibility of the system; the cloud server centrally stores all received key parameter data and provides an effective database management system; analyzes historical data and uses machine learning algorithms to optimize compressor operating parameters to improve energy efficiency and reduce wear; generates real-time reports to identify trends and abnormal patterns and support the decision-making process. Significance achieved: Decisions based on data analysis are more scientific and can effectively improve the operating efficiency and stability of the system; through continuous algorithm learning and optimization, the system can adapt to different operating conditions and environments and improve overall performance. The feedback control component automatically adjusts the working status of the low-temperature coupled heat pump (such as compressor speed, working time, etc.) according to the analysis results of the cloud server; provides an intelligent terminal view, where users can view key parameters at any time and receive alarm notifications about device status. Significance achieved: Allows the system to automatically adjust according to real-time data, improving the ability to respond to environmental and load changes; enhances user experience through real-time monitoring and instant reminders, allowing users to maintain transparency of equipment status at any time, so that maintenance actions can be taken in a timely manner.

[0060] In summary, the entire low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things in this embodiment realizes continuous monitoring, intelligent analysis and rapid response through the collaborative work of the above-mentioned components. The system realizes efficient energy use and optimized management, reduces energy consumption, improves the reliability and life of the equipment, and gives users higher control and convenience. This not only improves the overall performance of the equipment, but also promotes the goal of sustainable development and helps smart buildings and green energy utilization.

[0061] Example 2: Figure 2 As shown, based on Example 1, the data acquisition component provided by the embodiment of the present invention includes:

[0062] The position screening module is responsible for obtaining the target data of the area to be detected of the low-temperature coupled heat pump, obtaining multiple target points in the area to be detected, the target points correspond to the measurement points of the key data, and classifying the multiple measurement points using the classification strategy to obtain the corresponding relationship between the target number and the measurement point, and obtaining at least two measurement points of the same target data based on the corresponding relationship;

[0063] The equipment deployment module is responsible for obtaining the key parameter response time of the measurement point according to the historical key parameter change law, and selecting the measurement point with the shortest response time to the target parameter from at least two measurement points of the same target data, that is, the deployment point of the temperature sensor, flow sensor, pressure sensor or ambient temperature sensor of the target data; multiple deployment points form a key parameter monitoring network;

[0064] The data fusion module is responsible for synchronizing the data of different types of sensors according to timestamps, comprehensively calculating the confidence and characteristics of different types of sensor data, and generating the key parameters for final monitoring.

[0065] The working principle and beneficial effects of the above technical solution are as follows: the position screening module of this embodiment obtains the target data of the area to be detected of the low-temperature coupled heat pump, obtains multiple target points in the area to be detected, the target points correspond to the measurement points of the key data, and uses the classification strategy to classify and process the multiple measurement points to obtain the corresponding relationship between the target number and the measurement point, and obtains at least two measurement points of the same target data based on the corresponding relationship; the equipment layout module obtains the key parameter reaction time of the measurement point according to the historical key parameter change law, and screens out the measurement point with the shortest reaction time to the target parameter from at least two measurement points of the same target data, that is, the layout point of the temperature sensor, flow sensor, pressure sensor or ambient temperature sensor, etc. as the target data; multiple layout points are constructed into a key parameter monitoring network; the data fusion module synchronizes the data of different types of sensors according to timestamps, comprehensively calculates the confidence and characteristics of different types of sensor data, and generates the key parameters for final monitoring. The location screening module of the above scheme can effectively identify the key equipment and related measurement points in the area to be detected, including the monitoring locations of important parameters such as temperature, flow, pressure and environment; use classification strategies to effectively classify the measurement points, associate the measurement points with the target data, and provide basic data support for subsequent data analysis; establish a clear correspondence between each target point and its measurement point to ensure that the data collection has a clear direction and target. Implementation significance: Through accurate identification and classification, redundant measurement point selection and unnecessary data collection can be avoided, thereby improving the efficiency of monitoring work; clear correspondence can provide a reliable basis for subsequent data processing and analysis, ensuring the validity and accuracy of the data; laying the foundation for future automatic adjustment and optimization strategies, and realizing intelligent monitoring and management. The equipment layout module analyzes historical data to determine the reaction time of each measurement point, so as to select the most suitable sensor location for deployment; according to the principle of the shortest reaction parameter time, the optimal measurement point is selected for sensor layout to improve the response speed and real-time data; the selected multiple layout points are integrated into a key parameter monitoring network to achieve comprehensive monitoring of the entire heat pump system. Implementation significance: Optimizing the layout of measurement points makes the key parameters respond faster, thereby improving the agility and responsiveness of the overall system; timely obtaining feedback on changes in key parameters helps to quickly identify potential faults and reduce the operating risks of equipment; fast data updates and monitoring ensure that managers can grasp the system status in real time and support rapid decision-making. The data fusion module synchronizes the measurement data of different sensors according to timestamps to ensure the consistency of data in the same time frame; by calculating the confidence and characteristics of different types of sensor data, comprehensive monitoring data is formed to generate the final key parameters; redundant data and noise are effectively suppressed, improving the accuracy and reliability of monitoring data, and providing strong support for subsequent decision-making.Significance of implementation: High-quality fusion data can support more accurate intelligent analysis and decision-making, help optimize the operating efficiency of heat pumps; integrate data from different sources to achieve comprehensive monitoring, laying the foundation for system-level monitoring and analysis; through accurate data analysis and feedback, improve the overall efficiency of low-temperature coupled heat pumps, reduce energy waste, and support sustainable development.

[0066] In summary, each module of this embodiment plays a vital role in the low-temperature coupled heat pump data acquisition system. It not only realizes the scientific and accurate data acquisition, but also improves the system performance, reduces the operation risk, and supports the intelligent decision-making management function, which has far-reaching significance for the effective operation of the entire heat pump system. The multi-level and multi-module design concept provides strong support and guarantee for the future intelligent and digital energy management system.

[0067] Example 3: Figure 3 As shown, based on Example 2, the data fusion module specifically includes:

[0068] The time synchronization submodule is responsible for data collection of each sensor within the set time interval, and synchronizes the data of all sensors according to the timestamp to ensure the same time point;

[0069] The feature extraction submodule is responsible for assigning a confidence value to the data of each sensor when it is collected. Based on the standard deviation and mean of historical data, the data confidence of each sensor at the current moment is obtained; uniform feature extraction is performed on each data type, and the features include: mean, standard deviation, maximum value, minimum value and rate of change;

[0070] The weighted fusion submodule is responsible for setting the weight of each sensor feature, dynamically adjusting the weight according to the response time and confidence, and merging the features of each sensor with the corresponding confidence; it integrates the data results of all sensors and fuses them through weighted sum to obtain the final key parameters.

[0071] The working principle and beneficial effects of the above technical scheme are as follows: each sensor of the time synchronization submodule of this embodiment collects data within a set time interval, synchronizes the data of all sensors according to the timestamp, and ensures the same time point; the data of each sensor in the feature extraction submodule is assigned a confidence value when it is collected, and the data confidence of each sensor at the current moment is obtained based on the standard deviation and mean of historical data; a unified feature extraction is performed on each data type, and the features include: mean, standard deviation, maximum value, minimum value and rate of change; the weighted fusion submodule sets the weight of each sensor feature, dynamically adjusts the weight according to the reaction time and confidence, and merges the features of each sensor with the corresponding confidence; the data results of all sensors are integrated, and the final key parameters are obtained by weighted sum. The time synchronization submodule of the above scheme ensures that all sensors collect synchronous data within a set time interval (such as once per second); the data of each sensor is aligned by timestamp, which can effectively handle the possible delay of different sensors during data collection, and ensure that the data at the same time point can be compared and fused. Implementation significance: Accurate time synchronization ensures the timeliness and consistency of data, so that subsequent analysis can be based on the accurate time dimension; ensure that the relationship between the monitoring parameters is accurately reflected in real-time data monitoring, so that the processing results can also reflect the system status in real time; lay a solid foundation for subsequent data analysis and feature extraction. The feature extraction submodule calculates the confidence for each sensor data, and evaluates the reliability of the current data by statistically analyzing the standard deviation and mean of historical data; extracts multiple key features from the sensor data, such as mean, standard deviation, rate of change, etc., for subsequent analysis. Implementation significance: By calculating the confidence, unreliable data can be identified and eliminated, increasing the credibility of subsequent calculation results; the unified feature extraction strategy allows managers to systematically compare and analyze the data of different types of sensors, and then dig out deeper operating rules; through rich feature data, in-depth understanding of the system operation status, support intelligent decision-making and optimization adjustment. The weighted fusion submodule dynamically adjusts the weight of the features according to the reaction time and confidence of each sensor to ensure that the fusion algorithm can adapt to changes in input data; by combining the features of each sensor with its corresponding confidence, the final key parameters are comprehensively generated to ensure that a complete monitoring parameter is effectively output. Implementation significance: Through weight adjustment, actual reliable data is given priority to improve the quality of data output; data from different sensors are effectively fused through weighted averaging to form a global perspective and help the status assessment of the overall system; the key parameters of the final output provide more reliable data support for operational decisions and help technicians make more effective adjustments and optimizations.

[0072] In summary, through the organic cooperation of these three submodules, the data fusion module of this embodiment can realize real-time, accurate and efficient data processing, integrate the data of various sensors, and improve the monitoring ability of the low-temperature coupled heat pump system. It not only improves the reliability and efficiency of system operation, but also provides a rich data foundation for subsequent intelligent management and decision support, and lays the foundation for optimizing energy use and fault warning mechanism. The synergy between the various submodules can enhance the intelligent analysis and control capabilities of the entire system, and contribute to the realization of sustainable energy management goals.

[0073] Example 4: Figure 4 As shown, based on Example 1, the data transmission component provided in this embodiment of the present invention includes:

[0074] The signal processing module is responsible for realizing the key parameters after the initial processing of the IoT reception. During the IoT transmission, the IoT signal is interfered by the algorithm to obtain the IoT signal after the interference suppression processing, and then the key parameters after the interference suppression processing are transmitted to the cloud server;

[0075] The model training module is responsible for preprocessing the key parameters after interference suppression processing, obtaining the preprocessed key parameters, and dividing them into training set and test set according to the proportion; using the training set to train the support vector machine, obtain the strategy for optimizing the compressor operating parameters, and learn the relationship between the input features (such as water temperature, flow rate, etc.) and the target variables (such as the compressor's energy efficiency ratio, power consumption, etc.);

[0076] The model evaluation module is responsible for evaluating the support vector machine using the test set data, evaluating the performance of the support vector machine through mean square error and accuracy, and adjusting the support vector machine parameters.

[0077] The working principle and beneficial effects of the above technical scheme are as follows: the signal processing module of this embodiment realizes the key parameters after the initial processing of the Internet of Things, the interference suppression of the Internet of Things signal is performed by the algorithm during the Internet of Things transmission, the Internet of Things signal after the interference suppression processing is obtained, and then the key parameters after the interference suppression processing are transmitted to the cloud server; the model training module preprocesses the key parameters after the interference suppression processing, obtains the preprocessed key parameters, and divides them into training set and test set according to the proportion; the support vector machine is trained using the training set to obtain the strategy of optimizing the compressor operation parameters, and the relationship between the input features (such as water temperature, flow rate, etc.) and the target variables (such as the energy efficiency ratio of the compressor, power consumption, etc.) is learned; the model evaluation module uses the test set data to evaluate the support vector machine, evaluates the performance of the support vector machine by the mean square error and accuracy, and adjusts the support vector machine parameters. The signal processing module of the above scheme effectively suppresses the interference of the received Internet of Things signal through the signal processing algorithm (such as filtering algorithm or error detection) to ensure the signal quality of data transmission; by eliminating noise and errors, the accuracy and reliability of sensor data are improved, making it suitable for subsequent data analysis and processing. Significance achieved: Ensure reliable and accurate data is transmitted to subsequent processing links, thereby improving the reliability and accuracy of the entire system; reduce failures and errors caused by data interference, improve the stability of the entire Internet of Things system, and ensure effective operation in complex environments. The model training module further cleans and extracts the key parameters after interference suppression to generate data suitable for training machine learning models; use the training set data to train the support vector machine (SVM) to obtain the relationship between input features (such as water temperature, flow, etc.) and target variables (such as compressor energy efficiency ratio, power consumption, etc.). Significance achieved: Through the training model, the system can predict the operating behavior of the compressor based on historical data and key parameters, so as to achieve optimized operation under different conditions; generate optimization strategies for compressor operation, improve the efficiency of the system under different working conditions, reduce energy consumption, and extend the service life of the equipment. The model evaluation module uses the test set data to evaluate the trained support vector machine, calculates indicators such as mean square error and accuracy, and reflects the prediction performance of the model; according to the evaluation results, adjust the model's hyperparameters and optimize the model performance so that it can better adapt to actual data. Significance achieved: Ensure that the trained model has good generalization ability in practical applications and can make accurate predictions for unknown data; through model evaluation and optimization, the system can continuously learn and adapt to new environments, improve self-regulation and decision-making capabilities, and thus improve the overall performance of the system.

[0078] In summary, through the synergy of the signal processing module, model training module and model evaluation module, the data transmission component of this embodiment realizes the effective reception and processing of IoT data, model training and evaluation, thereby providing intelligent support for the operation optimization of the low-temperature coupled heat pump system. Improve data quality and reliability, laying the foundation for subsequent analysis; enhance the system's prediction ability and intelligent decision-making ability, so that it can optimize operating parameters and improve energy efficiency in different environments and different working conditions; realize intelligent management and maintenance through continuous feedback and optimization mechanism, so that the system can adapt to changing usage needs.

[0079] Example 5: Figure 5 As shown, based on Example 4, the signal processing module provided in this embodiment of the present invention includes:

[0080] The signal normalization submodule is responsible for defining the normalization scale function for the IoT signal that needs to be processed;

[0081] The scaling function expression is:

[0082]

[0083] In the formula, R represents the slope of the IoT signal to be processed, R min and R max The minimum and maximum values ​​of the slope of the IoT signal to be processed;

[0084] The interference judgment submodule is responsible for defining the time-frequency positioning strategy when the IoT signal is interfered with, and judging whether there is interference in the IoT signal to be processed;

[0085] The expression is:

[0086]

[0087] In the formula, Mask(n,k) represents the interference mask Mask that marks a specific signal, and its value is 1 or 0. A value of 1 indicates that the signal is not interfered with, and a value of 0 indicates that the IoT signal is interfered with. n represents the index of the nth sample or observation point in the signal sequence, k represents the index of the kth feature or dimension in the IoT signal sequence, and R(n,k); τ-c represents the signal value after filtering, τ represents the time delay, c represents the offset, and R represents the threshold value, which is used to determine whether the signal is interfered with. If the absolute value of the IoT signal after transformation is greater than this threshold, it is considered that the IoT signal is interfered with.

[0088] The suppression execution submodule is responsible for suppressing the interference component of the IoT signal with interference and outputting the suppressed IoT signal;

[0089] The expression is:

[0090]

[0091] In the formula, It represents the output suppressed IoT signal, which is used to represent the IoT signal after interference suppression processing. R(n,k) represents the original value of the IoT signal to be processed, represents the signal data under sample n and feature k, ⊙ represents the point-by-point multiplication operation, S(n,k) represents the useful part or effective signal component of the IoT signal, I(n,k) represents the interference component, that is, the signal part that needs to be suppressed, and Ω(n,k) represents the additional component in the IoT signal, which may be noise or other non-interference signal components.

[0092] The working principle and beneficial effects of the above technical solution are as follows: the signal normalization submodule of this embodiment defines a normalized scale function for the IoT signal to be processed; the interference judgment submodule defines the time-frequency positioning strategy when the IoT signal is interfered with, and judges whether there is interference in the IoT signal to be processed; the suppression execution submodule suppresses the interference component of the IoT signal with interference, and outputs the suppressed IoT signal. The signal normalization submodule of the above scheme normalizes the slope of the signal to be processed by a standardized formula, which will ensure that the value of the signal is within a unified range, usually [0,1]. The significance achieved: After normalization, the magnitude difference between different signal features is eliminated to prevent numerical instability during training or processing; in subsequent model training or signal analysis, the learning effect of the algorithm on each feature is balanced, and the training effect of the entire model will not be affected by the large values ​​of some features; after scaling the signal value to a unified interval, the user can intuitively understand the relative situation of different signals. The interference judgment submodule uses a threshold discrimination method to mark whether the signal is interfered with by a mask. In the formula, the interference signal is automatically identified by comparing the filtered signal value with the set threshold. Significance achieved: The signal quality can be judged quickly to ensure that the signal will not cause erroneous data analysis or decision-making due to interference during use; effectively identifying and processing interfered signals helps maintain the stability and reliability of the system. Only when the signal is considered to be undisturbed can it be included in subsequent data processing; by dynamically monitoring whether the signal is interfered, the system can adaptively adjust the processing strategy, thereby improving the flexibility of signal processing. The suppression execution submodule outputs the suppressed signal through point-by-point multiplication operations, mainly by suppressing the signal components marked as interference, thereby restoring the effective components of the signal. Significance achieved: By suppressing the interference components, the quality of the final output signal is improved, which can better reflect the real IoT parameters and play an important role in subsequent analysis and decision-making; after effective interference processing and signal recovery, the system can provide more accurate data, thereby supporting data-driven decision-making and prediction; after suppressing interference, the system can use data more efficiently for learning and reasoning, avoiding misleading results and waste of resources.

[0093] In summary, through the organic combination of the signal normalization submodule, the interference judgment submodule and the suppression execution submodule, the entire signal processing module of this embodiment achieves the following key technical effects and system significance: ensuring the quality of the input signal, so that subsequent data analysis and processing can proceed smoothly; real-time monitoring and interference suppression help the system dynamically adjust its operating state to adapt to the ever-changing environment; providing accurate and high-quality data for subsequent analysis and model training, and ultimately promoting the intelligence and efficiency of the Internet of Things system. This series of functions not only enhances the performance of the system, but also improves its feasibility and effectiveness in practical applications, so that end users can better obtain valuable information and insights.

[0094] Example 6: Figure 6 As shown, based on Example 4, the model training module provided in this embodiment of the present invention includes:

[0095] The data set construction submodule is responsible for labeling the preprocessed key parameters according to the data source, and constructing the preprocessed key parameters and labeling information into a data set. The labeling information is the type index and layout point location of each sensor in each preprocessed key parameter.

[0096] The data set processing submodule is responsible for downsampling the preprocessed key parameters in the data set by center cropping, and performing inversion or rotation operations on the preprocessed key parameters after downsampling according to the preset probability to expand the preprocessed key parameters and obtain a new data set; the data set is divided into a training set and a test set in an 8:2 ratio; among them, inversion (such as upside down flipping) for sensors involves negative inversion of the data to indicate reverse behavior; rotation is to transform the time series signal in the data; for example, for periodic signals, rolling or moving can be performed in the time dimension;

[0097] The training execution submodule is responsible for selecting the core function of the support vector machine and setting the penalty parameter C and kernel parameters; inputting the training set for training and updating the support vector machine model parameters through the optimization algorithm.

[0098] The working principle and beneficial effects of the above technical solution are as follows: the data set construction submodule of this embodiment performs labeling processing on the preprocessed key parameters according to the data source, and constructs the preprocessed key parameters and the labeling information into a data set, wherein the labeling information is the type index and the layout point location of each sensor in each preprocessed key parameter; the data set processing submodule adopts the central cropping method to downsample the preprocessed key parameters in the data set, and performs inversion or rotation operations on the downsampled preprocessed key parameters according to the preset probability to expand the preprocessed key parameters and obtain a new data set; the data set is divided into a training set and a test set according to a ratio of 8:2; wherein, inversion (such as upside down flipping) for sensors involves negative reversal of the data to indicate reverse behavior; rotation is the conversion of the time series signal in the data; for example, for periodic signals, it can be scrolled or moved in the time dimension; the training execution submodule selects the core function of the support vector machine, and sets the penalty parameter C and the kernel parameter; the training set is input for training, and the support vector machine model parameters are updated through the optimization algorithm. The annotation information of the dataset construction submodule of the above scheme includes the type index and the location of the deployment point of each sensor, which provides rich context information for subsequent model training. Significance achieved: By constructing a dataset with annotated information, background knowledge can be better utilized to improve the learning efficiency of the model; accurate annotation provides a basis for subsequent data analysis, so that the model can better understand the characteristics and environmental factors of the sensor when processing actual data, thereby improving the prediction accuracy. The dataset processing submodule adopts the key parameters after center cropping downsampling preprocessing to reduce the amount of data while retaining important features; the dataset is expanded by inversion and rotation operations to provide diversified samples for model training to improve the robustness of the model; the dataset is divided into training set and test set in a ratio of 8:2 to ensure effective verification of model performance. Significance achieved: Downsampling and data expansion not only provide richer training data, which helps prevent overfitting, but also improves the model's adaptability to different situations and signal anomalies; dividing the training set and test set ensures the rigor of model evaluation and helps ensure that the model can perform well on unknown data. The training execution submodule selects and sets the core function, penalty parameter C and kernel parameters of the support vector machine (SVM) to adapt to the characteristics of the data set; inputs the training set for model training, updates the model parameters of the support vector machine through optimization algorithms (such as gradient descent, SMO, etc.), and evaluates the test set. Significance achieved: By adjusting the parameters of the SVM for a specific data set, the learning effect of the model can be improved, thereby achieving better performance.

[0099] In summary, the three submodules of this embodiment perform their respective duties in the entire model training process, and their collaborative work ensures efficient data processing, model training, and performance evaluation. By constructing a data set with rich annotations, implementing effective data processing and expansion, and adopting optimized training methods, this module system can significantly improve the performance and reliability of the model in practical applications. This approach provides a systematic method for the analysis and processing of various sensor data, which can effectively deal with real-world problems.

[0100] Example 7: Figure 7 As shown, based on Example 6, the training execution submodule provided in this embodiment of the present invention includes:

[0101] The kernel function confirmation unit is responsible for setting multiple kernel functions, including linear kernel, polynomial kernel and radial basis kernel. For each kernel function, the support vector machine model is constructed using the k subsets divided by the data set. The support vector machine model is trained using k-1 subsets each time, and the remaining subset is used for verification. The average verification accuracy of each kernel function is recorded. The verification results of all kernel functions are compared, and the kernel function with the highest verification accuracy is selected as the kernel function used in the end.

[0102] The hyperparameter confirmation unit is responsible for writing an objective function to evaluate the effects of different hyperparameter combinations; determining the range of hyperparameter values; initially randomly sampling parameter combinations in the hyperparameter space, training the model, and calculating the effects of the combinations. Based on the initial sampling results, a probability model is constructed to capture the relationship between hyperparameters and model performance; on the probability model, the point with the greatest uncertainty is selected for sampling (i.e., the hyperparameter combination with the highest potential for improving model performance) and the performance results of the point are updated; when a given number of iterations or performance threshold is reached, the best performing hyperparameter combination is selected and provided to the support vector machine model for training;

[0103] The training set input unit is responsible for inputting the training set into the selected support vector machine model, training the support vector machine model, and obtaining the trained support vector machine model.

[0104] The working principle and beneficial effects of the above technical solution are as follows: the kernel function confirmation unit of this embodiment sets multiple kernel functions, including linear kernel, polynomial kernel and radial basis kernel. For each kernel function, a support vector machine model is constructed using k subsets divided by the data set. The support vector machine model is trained using k-1 subsets each time, and the remaining subset is used for verification, and the average verification accuracy of each kernel function is recorded; the verification results of all kernel functions are compared, and the kernel function with the highest verification accuracy is selected as the kernel function finally used; the hyperparameter confirmation unit writes an objective function for evaluating the effects of different hyperparameter combinations; the value range of the hyperparameter is determined; the initial When the hyperparameter combination is randomly sampled in the hyperparameter space, the model is trained, and the effect of the combination is calculated. Based on the initial sampling results, a probability model is constructed to capture the relationship between the hyperparameter and the model performance; on the probability model, the point with the greatest uncertainty is selected for sampling (i.e., the hyperparameter combination with the highest potential for improving the model performance), and the performance result of the point is updated; when the given number of iterations or performance threshold is reached, the best performing hyperparameter combination is selected and provided to the support vector machine model for training; the training set input unit passes the training set into the selected support vector machine model to train the support vector machine model and obtain the trained support vector machine model. The kernel function confirmation unit of the above scheme sets multiple kernel functions (such as linear kernel, polynomial kernel and radial basis kernel), which ensures that the model can process data with various distribution characteristics; each kernel function is trained and verified by k-fold cross validation to ensure that the performance of each kernel function is fully evaluated. Each training uses k-1 subsets and verifies the remaining subset to reduce the risk of overfitting; the average verification accuracy of each kernel function is recorded and compared to select the kernel function with the best performance. Significance: Through the setting of multiple kernel functions, this unit enables the model to adapt to linear and nonlinear features and improve classification capabilities; by systematically evaluating the performance of different kernel functions, it ensures that the kernel function finally selected is the best choice that has been verified, thereby improving model stability and prediction accuracy. The hyperparameter confirmation unit constructs a clear objective function for evaluating different hyperparameter combinations, which can accurately measure the performance of the model under specific hyperparameter settings, facilitating subsequent optimization; through initial random sampling, it explores the hyperparameter space and evaluates model performance, laying the foundation for subsequent efficient optimization; using the initial sampled data, a probabilistic model between hyperparameters and model performance is established, so that hyperparameters can be systematically analyzed and selected. Significance: Through scientific and systematic exploration and evaluation of hyperparameters, the hyperparameter confirmation unit makes model training no longer rely on empirical selection, improving the efficiency and effectiveness of parameter tuning; by optimizing hyperparameters, the classification performance of the model can be significantly improved, making it more adaptable to the data distribution in actual application scenarios and improving the generalization ability of the model.The training set input unit passes the selected training set data into the optimized SVM model for training, and updates the model parameters through the training process; after training, the trained support vector machine model is output and is ready for prediction and application on new samples. Significance: This unit represents the final step of the entire training process, implementing the work results of the first two units to form a model that can be used for practical applications; by using the training set for model training, it can ensure that the model fully learns the characteristics of the data, especially after the kernel function and hyperparameter optimization, the model can more effectively identify the characteristics and improve the prediction accuracy on new data.

[0105] In summary, the training execution submodule of this embodiment forms a complete support vector machine training system by hierarchically setting the functions of different units. The kernel function confirmation unit ensures that the model can flexibly respond to different data features, the hyperparameter confirmation unit improves the efficiency and accuracy of parameter selection, and the training set input unit realizes the final construction and application of the model. The modular design speeds up the model training and optimizes the model performance, so that the final model not only has excellent classification capabilities, but also can maintain good stability in different application scenarios.

[0106] Example 8: Figure 8 As shown, based on Example 7, the training set input unit provided in this embodiment of the present invention includes:

[0107] The data batch processing subunit is responsible for obtaining the real-time stream of input data and processing it by splitting the data into small batches; initializing the support vector machine model parameters, including the weight vector and bias term;

[0108] The iterative training subunit is responsible for calculating the current support vector machine model output for each batch of data, comparing it with the true value, and obtaining the loss; based on the loss of the current sample, the stochastic gradient descent method is used to update the support vector machine model parameters, weights, and biases;

[0109] The online learning mechanism subunit is responsible for randomly choosing whether to accept new data samples after each iteration and adjusting the support vector machine model according to the characteristics of the new samples; setting a threshold, and updating the support vector machine model if the score of the new sample is significantly different from the current model output.

[0110] The working principle and beneficial effects of the above technical solution are as follows: the data batch processing subunit of this embodiment obtains the real-time stream of input data, and processes the data by splitting it into small batches; initializes the support vector machine model parameters, including weight vectors and bias items; the iterative training subunit calculates the current support vector machine model output for each batch of data, and compares it with the true value to obtain the loss; according to the loss of the current sample, the support vector machine model parameters are updated using the stochastic gradient descent method, and the weights and biases are updated; the online learning mechanism subunit randomly selects whether to accept new data samples after each iteration, and adjusts the support vector machine model according to the characteristics of the new sample; sets a threshold, and if the score of the new sample is significantly different from the current model output, the support vector machine model is updated. The data batch processing subunit of the above scheme can receive and process new data instantly and adapt to the dynamically changing environment by obtaining the real-time stream of input data; splitting the data into small batches helps to reduce memory usage and speed up data processing; only the current small batch of data needs to be processed in each iteration, which can improve training efficiency; prepare initial parameters for the training of the support vector machine model, including weight vectors and bias items. Reasonable initialization of these parameters can help the model converge faster. Significance: Small batch processing allows the model to remain efficient when facing large-scale or real-time data. The highly adaptable design can better handle streaming data and changing inputs; through effective parameter initialization, it lays the foundation for effective learning of the model, making the training process more efficient and reducing convergence time. The iterative training subunit infers each batch of data and calculates the output value of the current support vector machine model. This process enables the model to evaluate its classification performance based on the existing parameters; by comparing with the true value, the loss between the current output and the actual label is calculated to evaluate the performance of the model. This is a key step in the training process and is essential for guiding parameter updates; using stochastic gradient descent, the model parameters (weights and biases) are updated according to the loss calculation results; the model is adjusted through gradient information to gradually reduce the loss. Significance: Through feedback on each batch of data, the model parameters are continuously optimized to ensure that the model continues to improve during the training process and converges towards higher accuracy; stochastic gradient descent can effectively process large data sets, especially when processing small batches of data, it can reduce the computational overhead of each iteration, accelerate the training process, and ensure that new samples can be quickly integrated into the model. After each iteration, the online learning mechanism subunit randomly chooses whether to accept new data samples, flexibly responding to dynamic changes in data streams; it updates the model online based on the characteristics of the new samples. This allows the model to quickly adapt to changes when new data is input without the need for complete retraining; it sets a threshold to determine the degree of difference between the new sample and the current model output, and if the difference is significant, updates the model parameters to ensure the adaptability and accuracy of the model.Significance: The online learning mechanism enables the model to adjust its decision boundary in a timely manner while continuously receiving new data, thereby improving the model's responsiveness to the real environment; by dynamically updating the model, it can effectively respond to data drift and changes, reduce the risk of model obsolescence, and thus ensure continued prediction accuracy and reliability.

[0111] In summary, the synergy of these subunits in this embodiment ensures the effectiveness, flexibility and adaptability of support vector machine model training. The data batch processing subunit enables the model to efficiently process real-time streaming data, the iterative training subunit ensures that the model can be continuously optimized in each iteration, and the online learning mechanism subunit enables the model to adapt to changes in new data and maintain long-term performance. This design not only improves the training efficiency, but also enhances the performance of the model in practical applications, making it have stronger practical application value.

[0112] Example 9: Fig. 9 As shown, based on Example 4, the model evaluation module provided by the embodiment of the present invention includes:

[0113] The result prediction submodule is responsible for inputting the test set data into the trained support vector machine model to obtain the predicted results of the support vector machine model;

[0114] The effect classification submodule is responsible for calculating the performance index of the mean square error machine accuracy; constructing a confusion matrix to count the relationship between each category predicted by the support vector machine model and the true category; presenting the confusion matrix in a graphical form to intuitively display the classification effect of the support vector machine model on each category;

[0115] The result feedback submodule is responsible for parameter tuning and support vector machine model improvement based on the feedback of confusion matrix and performance indicators.

[0116] The working principle and beneficial effects of the above technical solution are as follows: the result prediction submodule of this embodiment inputs the test set data into the trained support vector machine model to obtain the predicted results of the support vector machine model; the effect classification submodule calculates the performance index of the mean square error machine accuracy; constructs a confusion matrix, and statistically calculates the relationship between each category predicted by the support vector machine model and the true category; the confusion matrix is ​​presented in a graphical form to intuitively display the classification effect of the support vector machine model on each category; the result feedback submodule performs parameter tuning and support vector machine model improvement based on the feedback of the confusion matrix and performance indicators. The result prediction submodule of the above scheme inputs the prepared test set data into the trained support vector machine model to generate the prediction results; it is the basis of model evaluation, ensuring that we can view the classification of unseen data from the perspective of the model; through the prediction operation, the model's predicted label for each sample is obtained, and these results will be used for subsequent performance evaluation. Significance: Provide direct feedback on the model's classification ability for new and unseen data, laying the foundation for evaluating the performance of the model; ensure the comparison between the predicted results and the true labels, and form the basic data for subsequent effect classification. The effect classification submodule quantitatively measures the prediction performance of the model on the test set by calculating the mean square error (MSE) and accuracy; counts the number of true positives, false positives, false negatives, and true negatives to form a confusion matrix, thereby showing the specific performance of the model in each category; and displays the confusion matrix in a graphical form, such as a heat map, to help intuitively understand the classification effect of each category. Significance: Through multiple indicators (mean square error and accuracy) and confusion matrix, the performance of the model is comprehensively evaluated to identify the classification accuracy of the model in each category; the graphical display of the confusion matrix helps non-technical personnel to more easily understand the classification of the model, thereby improving the scientific nature of decision-making. The result feedback submodule analyzes the strengths and weaknesses of the model based on the results of the confusion matrix and performance indicators (such as mean square error and accuracy); performs parameter tuning for misclassifications found in the confusion matrix (such as a large number of false positives or false negatives in certain categories), and adjusts the hyperparameters of the support vector machine model, such as the penalty parameter C, the kernel function type, etc.; optimization can be performed based on feedback, such as data resampling (such as SMOTE), increasing the sample size of the training set, or modifying the feature selection method, etc. Significance: It provides a basis for the continuous optimization and improvement of the model, ensuring that the model can adapt to the new data environment and improve the long-term stability and accuracy of the model; through data feedback, targeted modifications can be made to reduce the error rate and enhance the performance of the model under specific conditions.

[0117] In summary, the synergy of these three submodules of the present embodiment ensures a comprehensive evaluation of the support vector machine model. The result prediction submodule can provide basic prediction output, and the effect classification submodule comprehensively reflects the model performance through multiple evaluation indicators and confusion matrix, and the result feedback submodule is dynamically adjusted and optimized on this basis. Through such a design, the model evaluation module not only promotes the performance improvement of the model, but also ensures a deep understanding of the model performance, and provides solid data support and feedback loop for subsequent applications.

[0118] Example 10: Fig.10 As shown, based on Example 1, the feedback control component provided by the embodiment of the present invention includes:

[0119] The command sending module is responsible for generating control commands based on the analysis results and decision-making. The cloud server sends the control commands to the feedback control component, including adjusting the operating speed of the compressor, switching the working mode (such as cooling or heating), adjusting the flow rate or changing the water temperature setting value, etc.;

[0120] The state adjustment module is responsible for adjusting the working state of the low-temperature coupled heat pump after receiving instructions from the cloud, adjusting the compressor state and circulation flow. Adjusting the compressor state includes changing the compressor operating speed or stopping / starting the compressor; adjusting the circulation flow includes adjusting the pump output or valve setting;

[0121] The alarm issuing module is responsible for viewing key parameters in real time through the intelligent terminal and monitoring the operating status of the low-temperature coupled heat pump; if the low-temperature coupled heat pump has an abnormality (such as overheating, compressor failure, etc.), an alarm notification will be sent in time.

[0122] The working principle and beneficial effects of the above technical solution are as follows: the instruction sending module of this embodiment is based on the analysis results and decision-making, and the cloud server generates control instructions and sends them to the feedback control component, including adjusting the operating speed of the compressor, switching the working mode (such as cooling or heating), adjusting the flow rate or changing the water temperature setting value, etc.; after receiving the instructions from the cloud, the state adjustment module adjusts the working state of the low-temperature coupled heat pump, adjusts the compressor state and the circulation flow rate, and adjusting the compressor state includes changing the operating speed of the compressor or stopping / starting the compressor; adjusting the circulation flow rate includes adjusting the output of the pump or the valve setting; the alarm issuing module checks the key parameters in real time through the intelligent terminal and monitors the operating status of the low-temperature coupled heat pump; if the low-temperature coupled heat pump is abnormal (such as overheating, compressor failure, etc.), an alarm notification is sent in time. The instruction sending module of the above scheme receives the analysis results and decisions of the cloud server, and generates accurate control instructions based on this information to ensure that the operating state of the low-temperature coupled heat pump can be adjusted in time; the instruction content covers the adjustment of the compressor operating speed, the switching of the working mode (such as cooling / heating), the adjustment of the flow rate or the change of the water temperature setting value, ensuring that the system adapts to different working conditions. Significance: Realize the automatic management and control of heat pumps, reduce manual intervention, and improve the intelligence level of the system; through timely instruction sending, the system can quickly respond to environmental changes or changes in load demand, and improve the overall operating efficiency and energy saving effect. The state adjustment module is responsible for the actual execution of the adjustment of the compressor and flow according to the instructions of the instruction sending module; by changing the operating speed of the compressor or performing stop / start operations, ensure that the compressor operates in the best working state; by adjusting the output of the pump or the valve setting, reasonably control the fluid flow in the system and optimize the heat exchange effect. Significance: Through precise state adjustment, ensure that the heat pump still operates efficiently when the load demand changes, avoid energy waste; reasonable operation parameter settings can reduce equipment wear and damage, extend the service life of the equipment, and reduce maintenance costs. The alarm issuance module continuously monitors the key parameters of the low-temperature coupled heat pump (such as temperature, flow, compressor status, etc.) through the intelligent terminal to ensure timely detection of abnormal conditions; when an abnormal operation (such as overheating, compressor failure, etc.) is detected, an alarm notification is immediately sent to prompt the operation and maintenance personnel to take immediate measures. Significance: Through real-time monitoring and abnormal alarm mechanisms, potential risks can be identified early to prevent equipment damage or personal injury, and improve system stability and safety. Through the alarm function of smart terminals, operation and maintenance personnel can obtain system status information in a timely manner, respond quickly and take correct measures to reduce the possibility of accidents and improve operation and maintenance efficiency.

[0123] In summary, the three modules of this embodiment each play a unique role in the feedback control component, and form a complete feedback control system through collaborative work. The instruction sending module is responsible for the implementation of intelligent decision-making, the state adjustment module performs specific operations, and the alarm issuing module ensures the safe operation of the system and responds to abnormalities in a timely manner. The feedback control component not only improves the automation and intelligence level of the low-temperature coupled heat pump, but also creates conditions for optimizing operating efficiency and equipment safety, and ultimately promotes the continuous improvement and optimization of the system.

[0124] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A remote monitoring and data acquisition system for low-temperature coupled heat pumps based on the Internet of Things, characterized in that: Include: The data acquisition component is responsible for real-time acquisition of key parameters, and preliminarily processes the acquired key parameters to obtain the preliminarily processed key parameters; The data transmission component is responsible for transmitting the key parameters after preliminary processing to the cloud server in real time through the Internet of Things; The cloud server processes key parameters and optimizes the operating parameters of the compressor; The feedback control component is responsible for automatically adjusting the working state of the low-temperature coupled heat pump according to the analysis results of the cloud server.

2. The low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things according to claim 1 is characterized in that: Data collection components, including: The position screening module is responsible for obtaining the target data of the area to be detected of the low-temperature coupled heat pump, obtaining multiple target points in the area to be detected, the target points correspond to the measurement points of the key data, classifying the multiple measurement points using the classification strategy, and obtaining at least two measurement points of the same target data based on the corresponding relationship; The equipment deployment module is responsible for obtaining the key parameter response time of the measurement point according to the historical key parameter change law, and selecting the measurement point with the shortest response time to the target parameter from at least two measurement points with the same target data, i.e., the deployment point; The data fusion module is responsible for synchronizing the data of different types of sensors according to timestamps, comprehensively calculating the confidence and characteristics of different types of sensor data, and generating the key parameters for final monitoring.

3. The low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things as claimed in claim 2 is characterized in that: Data fusion module, including: The time synchronization submodule is responsible for data collection of each sensor within the set time interval, and synchronizes the data of all sensors according to the timestamp to ensure the same time point; The feature extraction submodule is responsible for assigning a confidence value to the data of each sensor when it is collected, and deriving the data confidence of each sensor at the current moment based on the standard deviation and mean of historical data; and performing unified feature extraction for each data type; The weighted fusion submodule is responsible for setting the weight of each sensor feature, dynamically adjusting the weight according to the response time and confidence, and merging the features of each sensor with the corresponding confidence; it integrates the data results of all sensors and fuses them through weighted sum to obtain the final key parameters.

4. The low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things according to claim 1, characterized in that: Data transmission components, including: The signal processing module is responsible for realizing the key parameters after the initial processing of the IoT reception. During the IoT transmission, the IoT signal is interfered by the algorithm and the key parameters after the interference suppression processing are transmitted to the cloud server. The model training module is responsible for preprocessing the key parameters after interference suppression processing, obtaining the preprocessed key parameters, and dividing them into training sets and test sets according to the proportion; The model evaluation module is responsible for evaluating the support vector machine using the test set data, evaluating the performance of the support vector machine through mean square error and accuracy, and adjusting the support vector machine parameters.

5. The low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things according to claim 4, characterized in that: Signal processing module, including: The signal normalization submodule is responsible for defining the normalization scale function for the IoT signal that needs to be processed; The interference judgment submodule is responsible for defining the time-frequency positioning strategy when the IoT signal is interfered with, and judging whether there is interference in the IoT signal to be processed; The suppression execution submodule is responsible for suppressing the interference component of the IoT signal with interference and outputting the suppressed IoT signal.

6. The remote monitoring and data acquisition system for low-temperature coupled heat pumps based on the Internet of Things according to claim 4, characterized in that: Model training module, including: The dataset construction submodule is responsible for labeling the key parameters after preprocessing according to the data source; The data set processing submodule is responsible for dividing the data set into training set and test set; The training execution submodule is responsible for inputting the training set for training and updating the support vector machine model parameters through the optimization algorithm.

7. The low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things according to claim 6, characterized in that: The dataset construction submodule constructs the preprocessed key parameters and annotation information into a dataset.

8. The remote monitoring and data acquisition system for low-temperature coupled heat pumps based on the Internet of Things according to claim 6, characterized in that: The data set processing submodule uses a central cropping method to downsample the preprocessed key parameters in the data set, and performs an inversion or rotation operation on the downsampled preprocessed key parameters according to a preset probability to expand the preprocessed key parameters; Expand the key parameters after preprocessing and obtain new data sets.

9. The low-temperature coupled heat pump remote monitoring and data acquisition system based on the Internet of Things according to claim 6, characterized in that: The training execution submodule selects the core function of the support vector machine and sets the penalty parameter C and the kernel parameter.

10. The remote monitoring and data acquisition system for low-temperature coupled heat pumps based on the Internet of Things according to claim 1, characterized in that: Feedback control components, including: The instruction sending module is responsible for generating control instructions based on the analysis results and decision making, and sending them to the feedback control component; The state adjustment module is responsible for adjusting the working state of the low-temperature coupled heat pump and regulating the compressor state and circulation flow rate after receiving instructions from the cloud; The alarm issuing module is responsible for viewing key parameters in real time through the smart terminal and monitoring the operating status of the low-temperature coupled heat pump; If the low-temperature coupled heat pump has any abnormality, an alarm notification will be sent in time.

Images