An intelligent HVAC terminal with collaborative cloud data storage function

By using deep learning technology in HVAC systems for temperature data processing and analysis, the problem of insufficient response speed in existing systems during environmental changes is solved, and more accurate and intelligent temperature control is achieved.

CN118836556BActive Publication Date: 2024-12-06HANGZHOU SULI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411314110.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-20
Publication Date
2024-12-06
Estimated Expiration
2044-09-20

AI Technical Summary

Technical Problem

When the environmental conditions of the existing HVAC cloud edge collaborative system changes rapidly, the preset artificial intelligence regulation model cannot promptly reflect the latest changes, resulting in insufficient response speed and control flexibility, making it difficult to achieve highly accurate temperature control.

Method used

Using deep learning-based data processing technology, the inlet pipe, return pipe and indoor temperature data transmitted by edge servers are obtained through cloud servers, and local timing characteristics are correlated and principal components are significantly integrated to generate HVAC temperature adjustment results to achieve environmental changes. Quick response and precise control.

Benefits of technology

It improves the response speed to instant changes, captures temperature fluctuations in a short time, and achieves more accurate and intelligent temperature control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118836556B_ABST
    Figure CN118836556B_ABST
Patent Text Reader

Abstract

The present application relates to the field of HVAC equipment, and specifically to an intelligent HVAC terminal with collaborative cloud data storage function. Among them, the acquisition terminal includes an inlet water temperature sensor, a return water temperature sensor and a room temperature sensor respectively deployed in the inlet water pipe and return water pipe of the HVAC equipment and used to collect the inlet water pipe temperature data and return water pipe temperature data of the HVAC equipment; the room temperature sensor is used to collect indoor temperature data; the edge server is used to upload the inlet water pipe temperature data, the return water pipe temperature data and the indoor temperature data to the cloud server for storage; the cloud server is used to analyze the inlet water pipe temperature data, the return water pipe temperature data and the indoor temperature data to obtain the HVAC equipment temperature adjustment result, and transmit the HVAC equipment temperature adjustment result to the control terminal; the control terminal is used to control the power of the HVAC equipment according to the HVAC equipment temperature adjustment result.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of HVAC equipment, and in particular to an intelligent HVAC terminal with a collaborative cloud data storage function. Background Art

[0002] With the rapid progress of HVAC technology, it is increasingly integrated into people's daily life, and at the same time, the problem of energy consumption is becoming increasingly serious. Therefore, it is inevitable that HVAC equipment will develop in the direction of energy conservation and emission reduction. Under the current background of advocating green and environmental protection, the market demand for energy conservation and environmental protection is becoming increasingly urgent. In this fast-growing and highly competitive market environment, modern refrigeration technology has achieved significant breakthroughs and opened up new possibilities for the development of the industry. Today, products with energy conservation and environmental protection as the core selling point are rapidly rising and gradually gaining market recognition.

[0003] Chinese patent CN113535233B proposes an artificial intelligence system for HVAC cloud-edge collaboration, which is responsible for collecting equipment operation data through the acquisition terminal, transmitting it to the edge server for preliminary analysis, selecting control algorithms and strategies, and uploading the results to the cloud server. The cloud server further optimizes the control parameters, and the edge server generates control instructions based on this, which are implemented through the control terminal to adjust the operation of HVAC equipment and improve energy efficiency and performance.

[0004] In the above patent, the process of data analysis by the cloud server mainly relies on the preset artificial intelligence control model, and generates the optimal energy-saving parameter adjustment plan through simulation execution. However, since these models are usually trained based on historical data and pre-set scenarios, they may not be able to immediately reflect the latest changes in the environment. That is, when environmental conditions evolve rapidly, the prediction and control strategies of the model may lag behind and cannot be adjusted in time, thereby reducing the system's response speed to immediate changes and the flexibility of control. In addition, the method in the above patent focuses on the overall optimization strategy and lacks the correlation of refined timing feature analysis of each parameter. This means that the system may not be able to capture the details of temperature changes in a short period of time, making it difficult to achieve highly accurate temperature control. Especially in cases where fine-tuning is required, insufficient control accuracy may cause the HVAC equipment to fail to achieve the ideal operating state, affecting the overall energy-saving effect and user experience.

[0005] Therefore, an optimized intelligent HVAC terminal is desired. Summary of the invention

[0006] This application is made in view of the above problems. One object of this application is to provide an intelligent HVAC terminal with collaborative cloud data storage function.

[0007] An embodiment of the present application provides an intelligent HVAC terminal with a collaborative cloud data storage function, which includes: a cloud server, an edge server, a collection terminal, a control terminal and HVAC equipment; wherein the collection terminal includes an inlet water temperature sensor, a return water temperature sensor and a room temperature sensor, and the inlet water temperature sensor and the return water temperature sensor are respectively deployed in the inlet pipe and the return pipe of the HVAC equipment, and are used to collect the inlet pipe temperature data and the return pipe temperature data of the HVAC equipment; the room temperature sensor is used to collect indoor temperature data; the edge server is used to upload the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data to the cloud server for storage; the cloud server is used to analyze the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data to obtain the temperature adjustment result of the HVAC equipment, and transmit the temperature adjustment result of the HVAC equipment to the control terminal; the control terminal is used to control the power of the HVAC equipment according to the temperature adjustment result of the HVAC equipment.

[0008] For example, according to an embodiment of the present application, an intelligent HVAC terminal with a collaborative cloud data storage function, wherein the cloud server includes: a data acquisition module, used to acquire the time series of the inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data transmitted by the edge server; a timing encoding module, used to arrange the time series of the inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data into vectors respectively and then pass them through a 1D-CNN-based timing encoder to obtain the inlet pipe temperature local time series correlation feature vector, the return pipe temperature local time series correlation feature vector and the indoor temperature local time series correlation feature vector; an inlet pipe-return pipe temperature significant fusion module, used to combine the inlet pipe temperature local time series correlation feature vector and the indoor temperature local time series correlation feature vector. The local time series associated feature vector of the return pipe temperature is obtained through a significant fusion network based on feature principal component query matching to obtain a sparse significant fusion representation vector of the inlet-return pipe temperature; a HVAC equipment state-room temperature state local time series interaction module is used to obtain a HVAC equipment state-room temperature state fine-grained local time series semantic interaction response feature vector by combining the inlet-return pipe temperature sparse significant fusion representation vector and the indoor temperature local time series associated feature vector through a local fine-grained sequence aggregation responder based on feature decoupling; a HVAC equipment temperature adjustment result generation module is used to obtain the HVAC equipment temperature adjustment result based on the HVAC equipment state-room temperature state fine-grained local time series semantic interaction response feature vector, and the HVAC equipment temperature adjustment result is used to indicate whether the temperature of the HVAC equipment should increase, decrease or remain unchanged.

[0009] For example, according to the intelligent HVAC terminal with collaborative cloud data storage function of the embodiment of the present application, the inlet-return pipe temperature significant fusion module includes: an inlet-return pipe temperature local time series feature standardization unit, which is used to standardize the inlet pipe temperature local time series correlation feature vector and the return pipe temperature local time series correlation feature vector to obtain a standardized inlet pipe temperature local time series correlation feature vector and a standardized return pipe temperature local time series correlation feature vector; an inlet-return pipe temperature sample covariance matrix calculation unit, which is used to calculate the standardized inlet pipe temperature local time series correlation feature vector and the standardized return pipe temperature local time series correlation feature vector. A sample covariance matrix of the associated eigenvectors is used to obtain the sample covariance matrix of the inlet pipe temperature and the sample covariance matrix of the return pipe temperature; a matrix decomposition unit is used to extract the eigenvectors based on the matrix decomposition of the inlet pipe temperature sample covariance matrix and the return pipe temperature sample covariance matrix to obtain the set of the inlet pipe temperature principal component eigenvectors and the set of the return pipe temperature principal component eigenvectors; an inlet pipe-return pipe temperature maximum approximation query unit is used to input the set of the inlet pipe temperature principal component eigenvectors and the set of the return pipe temperature principal component eigenvectors into the maximum approximation query matching network to obtain the inlet pipe temperature principal component eigenvectors and the return pipe temperature principal component eigenvectors. a set of best matching pairs of component feature vectors; an inlet-return pipe temperature principal component fusion unit, used to input the best matching pairs of each inlet pipe temperature principal component feature vector and return pipe temperature principal component feature vector in the set of best matching pairs of the inlet pipe temperature principal component feature vector and the return pipe temperature principal component feature vector into a semantic fine-grained gated joint module to obtain a set of inlet-return pipe temperature principal component fusion feature vectors; a vector cascade unit, used to cascade the set of inlet-return pipe temperature principal component fusion feature vectors to obtain the inlet-return pipe temperature sparse significant fusion representation vector; wherein the inlet-return pipe temperature is the best matching pair of the inlet-return pipe temperature principal component feature vector. The large approximate query unit is used to: extract a predetermined water inlet pipe temperature principal component feature vector from the set of water inlet pipe temperature principal component feature vectors; calculate the cosine similarity between the predetermined water inlet pipe temperature principal component feature vector and each return pipe temperature principal component feature vector in the set of return pipe temperature principal component feature vectors to obtain a set of matching query similarities; and use the return pipe temperature principal component feature vector corresponding to the largest matching query similarity in the set of matching query similarities and the predetermined water inlet pipe temperature principal component feature vector as the best matching pair of the predetermined water inlet pipe temperature principal component feature vector and the return pipe temperature principal component feature vector.

[0010] For example, according to the embodiment of the present application, in an intelligent HVAC terminal with a collaborative cloud data storage function, the inlet and return pipe temperature local time series feature standardization unit is used to: respectively calculate the mean and standard deviation of the inlet pipe temperature local time series correlation feature vector to obtain the inlet pipe temperature local time series correlation feature mean and the inlet pipe temperature local time series correlation feature standard deviation; after subtracting the inlet pipe temperature local time series correlation feature vector from the inlet pipe temperature local time series correlation feature mean by position, the calculated inlet pipe temperature local time series correlation offset vector and the inlet pipe temperature local time series correlation feature standard deviation are obtained. The method comprises the following steps: performing positional division to obtain the standardized inlet pipe temperature local time series association feature vector; calculating the mean and standard deviation of the return pipe temperature local time series association feature vector respectively to obtain the mean of the return pipe temperature local time series association feature and the standard deviation of the return pipe temperature local time series association feature; subtracting the return pipe temperature local time series association feature vector from the mean of the return pipe temperature local time series association feature by position, and then performing positional division of the calculated return pipe temperature local time series association offset vector and the standard deviation of the return pipe temperature local time series association feature to obtain the standardized return pipe temperature local time series association feature vector.

[0011] For example, according to an embodiment of the present application, an intelligent HVAC terminal with a collaborative cloud data storage function, wherein the inlet and return pipe temperature sample covariance matrix calculation unit is used to: multiply the transposed vector of the standardized inlet pipe temperature local time series correlation feature vector by the standardized inlet pipe temperature local time series correlation feature vector, and then divide the obtained standardized inlet pipe temperature correlation matrix by the value obtained by subtracting one from the length of the standardized inlet pipe temperature local time series correlation feature vector to obtain the inlet pipe temperature sample covariance matrix; multiply the transposed vector of the standardized return pipe temperature local time series correlation feature vector by the standardized return pipe temperature local time series correlation feature vector, and then divide the obtained standardized return pipe temperature correlation matrix by the value obtained by subtracting one from the length of the standardized return pipe temperature local time series correlation feature vector to obtain the return pipe temperature sample covariance matrix.

[0012] For example, according to the embodiment of the present application, an intelligent HVAC terminal with a collaborative cloud data storage function, wherein the inlet-return pipe temperature principal component fusion unit is used to: respectively calculate the position difference, position dot multiplication and position addition between the best matching pairs of the inlet pipe temperature principal component feature vector and the return pipe temperature principal component feature vector to obtain the inlet pipe-return pipe temperature principal component pattern difference vector, the inlet pipe-return pipe temperature principal component pattern dot product vector and the inlet pipe-return pipe temperature principal component pattern sum vector; cascade the inlet pipe-return pipe temperature principal component pattern difference vector, the inlet pipe-return pipe temperature principal component pattern dot product vector and the inlet pipe-return pipe temperature principal component pattern sum vector, and perform one-dimensional convolution encoding to obtain the inlet pipe-return pipe temperature principal component pattern multi-dimensional fusion vector; perform local window-based maximum pooling processing on the inlet pipe-return pipe temperature principal component pattern multi-dimensional fusion vector to obtain the inlet pipe-return pipe temperature principal component fusion feature vector.

[0013] For example, according to an intelligent HVAC terminal with a collaborative cloud data storage function according to an embodiment of the present application, the HVAC equipment state-room temperature state local time series interaction module includes: a feature decoupling unit, which is used to perform feature decoupling on the inlet-return pipe temperature sparse significant fusion representation vector and the indoor temperature local time series associated feature vector to obtain a sequence of inlet-return pipe temperature local sparse significant fusion representation vectors and a sequence of indoor temperature local time series decoupling associated feature vectors; an inlet-return pipe-indoor temperature local time series semantic association interaction unit, which is used to calculate the local semantic association interaction vector between each group of corresponding inlet-return pipe temperature local sparse significant fusion representation vectors and indoor temperature local time series decoupling associated feature vectors in the sequence of the inlet-return pipe temperature local sparse significant fusion representation vectors and the sequence of the indoor temperature local time series decoupling associated feature vectors to obtain a sequence of inlet-return pipe-indoor temperature local time series semantic association interaction vectors; The water inlet pipe return pipe-indoor temperature local temporal series fine-grained attention weight calculation unit is used to calculate the fine-grained attention weights of each water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction vector in the sequence of the water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction vector to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal series fine-grained attention weights; the deactivation processing unit is used to input the sequence of the water inlet pipe return pipe-indoor temperature local temporal series fine-grained attention weights into a random deactivation module to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal series fine-grained attention weights after deactivation processing; the weighted sum calculation unit is used to use the sequence of the water inlet pipe return pipe-indoor temperature local temporal series fine-grained attention weights after the deactivation processing as the weight distribution, and calculate the weighted sum of the sequence of the water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction vectors to obtain the HVAC equipment state-room temperature state fine-grained local temporal series semantic interaction response feature vector.

[0014] For example, according to an intelligent HVAC terminal with collaborative cloud data storage function according to an embodiment of the present application, the water inlet return pipe-indoor temperature local temporal semantic association interaction unit is used to: respectively calculate the position difference, position dot multiplication and position addition between the local sparse significant fusion representation vector of the water inlet-return pipe temperature and the local temporal decoupling association feature vector of the indoor temperature to obtain the local temporal difference vector of the water inlet return pipe-indoor temperature, the local temporal dot product vector of the water inlet return pipe-indoor temperature and the water inlet return pipe-indoor temperature. The local time series sum vector of the water inlet pipe return pipe-indoor temperature is obtained by cascading the local time series difference vector of the water inlet pipe return pipe-indoor temperature, the local time series point product vector of the water inlet pipe return pipe-indoor temperature and the local time series sum vector of the water inlet pipe return pipe-indoor temperature, and performing one-dimensional convolution encoding to obtain a local time series multi-dimensional fusion vector of the water inlet pipe return pipe-indoor temperature; the local window-based maximum pooling processing is performed on the local time series multi-dimensional fusion vector of the water inlet pipe return pipe-indoor temperature to obtain the local time series semantic association interaction vector of the water inlet pipe return pipe-indoor temperature.

[0015] For example, according to an embodiment of the present application, an intelligent HVAC terminal with a collaborative cloud data storage function, wherein the water inlet pipe return pipe-indoor temperature local temporal sequence fine-grained attention weight calculation unit is used to: multiply each water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction vector in the sequence of the water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction vectors by a weight matrix, and then add the obtained vector to the bias vector by position to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction modulation vectors; input the sequence of water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction modulation vectors into a tanh function for activation processing to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction activation vectors; multiply each water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction activation vector in the sequence of the water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction activation vectors by a weight coefficient vector to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal sequence fine-grained attention weights.

[0016] For example, according to an embodiment of the present application, an intelligent HVAC terminal with a collaborative cloud data storage function, wherein the HVAC equipment temperature adjustment result generation module is used to: pass the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector through a classifier-based HVAC equipment temperature adjuster to obtain the HVAC equipment temperature adjustment result, and the HVAC equipment temperature adjustment result is used to indicate whether the temperature of the HVAC equipment should increase, decrease, or remain unchanged.

[0017] According to the intelligent HVAC terminal with collaborative cloud data storage function of the embodiment of the present application, it obtains the time series of the inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data transmitted by the edge server, and adopts the data processing and analysis technology based on deep learning to associate the local time series features of the inlet pipe temperature, the return pipe temperature and the indoor temperature data, and adaptively adjusts the temperature of the HVAC equipment based on the main component significant fusion features of the inlet pipe temperature and the return pipe temperature and the local time series semantic interaction response features between the indoor temperature time series features. In this way, environmental changes can be quickly identified and corresponding adjustments can be made, which improves the response speed to immediate changes. At the same time, temperature fluctuations in a short period of time can be captured, thereby achieving more accurate and intelligent temperature control. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0019] Figure 1 A structural schematic diagram of an intelligent HVAC terminal with collaborative cloud data storage function in an embodiment of the present application is shown.

[0020] Figure 2 A schematic diagram of the structure of a cloud server of an intelligent HVAC terminal with a collaborative cloud data storage function in an embodiment of the present application is shown.

[0021] Figure 3 A flow chart of an intelligent HVAC terminal control method with collaborative cloud data storage function in an embodiment of the present application is shown.

[0022] Figure 4 An application scenario diagram of an intelligent HVAC terminal with collaborative cloud data storage function in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0024] The terms used in this specification are those common terms currently widely used in the art in consideration of the functions of the present application, but these terms may vary according to the intentions of those skilled in the art, precedents, or new technologies in the art. In addition, specific terms may be selected, and in this case, their detailed meanings will be described in the detailed description of the present application. Therefore, the terms used in the specification should not be understood as simple names, but rather as a general description based on the meaning of the terms and the present application.

[0025] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0026] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0027] In addition, the application architecture diagram in the embodiment of the present application is intended to more clearly illustrate the technical solution in the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present application is also applicable to similar problems.

[0028] The following is a non-restrictive description of the smart HVAC terminal with collaborative cloud data storage function provided according to at least one embodiment of the present application through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present application.

[0029] It should be understood that cloud data usually refers to a large amount of data stored in the cloud, which can include structured and unstructured data, such as text, pictures, videos, etc. Cloud data storage refers to storing data on cloud servers instead of traditional local hard drives or private servers. This storage method allows data to be accessed anytime and anywhere through the Internet, improving the accessibility and flexibility of data. Collaborative cloud data storage function refers to the synchronization and collaboration of data between multiple cloud servers or data centers. This function can ensure high availability, disaster recovery capabilities and load balancing of data. In the application of HVAC terminals, collaborative cloud data storage can make the data synchronization and storage between the various components of the HVAC system, such as acquisition terminals, edge servers and cloud servers, more efficient and reliable.

[0030] Based on this, the present application proposes an intelligent HVAC terminal 800 with collaborative cloud data storage function, such as Figure 1 As shown, it includes: a cloud server 810, an edge server 820, a collection terminal 830, a control terminal 840 and a HVAC device 850; wherein the collection terminal 830 includes an inlet water temperature sensor, a return water temperature sensor and a room temperature sensor, the inlet water temperature sensor and the return water temperature sensor are respectively deployed in the inlet pipe and the return pipe of the HVAC device 850, and are used to collect the inlet pipe temperature data and the return pipe temperature data of the HVAC device 850; the room temperature sensor is used to collect indoor temperature data; the edge server 820 is used to upload the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data to the cloud server 810 for storage; the cloud server 810 is used to analyze the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data to obtain the HVAC device temperature adjustment result, and transmit the HVAC device temperature adjustment result to the control terminal 840; the control terminal 840 is used to control the power of the HVAC device 850 according to the HVAC device temperature adjustment result.

[0031] It should be understood that each component has its specific role and function, and they work together to achieve efficient and automated HVAC control. As the core data processing and analysis center of the system, the cloud server 810 is responsible for storing a large amount of data. It performs complex analysis on the collected data to optimize the performance and efficiency of the HVAC equipment. It can be remotely accessed and monitored to facilitate system maintenance and upgrades. It usually has high reliability and data redundancy to ensure data security and stable system operation. As the initial processing and caching point of data, the edge server 820 is close to the data source to reduce the delay in data transmission to the cloud. It pre-processes the collected data, such as cleaning, filtering and aggregation, to reduce the amount of data transmitted to the cloud. In this way, it can quickly respond to local control needs and improve the real-time and reliability of the system. When the cloud server is unavailable, the edge server can process data independently to ensure the basic operation of the system. The acquisition terminal 830 is in direct contact with the HVAC equipment and the environment, and is responsible for real-time monitoring and data collection. It includes an inlet water temperature sensor, a return water temperature sensor and a room temperature sensor, which are used to collect the working status of the HVAC equipment and indoor environmental parameters, and further send the collected data to the edge server or directly to the cloud server. The control terminal 840 receives instructions from the cloud server and precisely controls the HVAC equipment. It adjusts the operating parameters of the HVAC equipment, such as temperature setting, power output, etc., according to the data analysis results. It can also be remotely controlled or automatically controlled according to preset logic to achieve optimal indoor comfort and energy efficiency. The HVAC equipment 850 is the execution unit of the system, responsible for actual heating, ventilation, and air conditioning functions. It adjusts its operating status according to the instructions of the control terminal to meet the needs of the indoor environment. Furthermore, its operating status and performance parameters are monitored by the collection terminal for data analysis and optimization of the system. Through cloud data storage and collaborative work, the entire system realizes efficient data processing and intelligent control of HVAC equipment, improves energy efficiency and user comfort, and reduces operating costs.

[0032] Correspondingly, in the above-mentioned cloud server, the technical concept of the present application is to obtain the time series of the inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data transmitted by the edge server, and adopt the data processing and analysis technology based on deep learning to associate the local time series features of the inlet pipe temperature, the return pipe temperature and the indoor temperature data, and adaptively adjust the temperature of the HVAC equipment based on the local time series semantic interaction response features between the main component significant fusion features of the inlet pipe temperature and the return pipe temperature and the indoor temperature time series features. In this way, environmental changes can be quickly identified and corresponding adjustments can be made, which improves the response speed to immediate changes. At the same time, temperature fluctuations in a short period of time can be captured, thereby achieving more accurate and intelligent temperature control.

[0033] Accordingly, if Figure 2 As shown, the cloud server 810 includes: a data acquisition module 811, which is used to acquire the time series of the inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data transmitted by the edge server 820; a time series encoding module 812, which is used to arrange the time series of the inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data into vectors and then pass them through a time series encoder based on 1D-CNN to obtain the inlet pipe temperature local time series correlation feature vector, the return pipe temperature local time series correlation feature vector and the indoor temperature local time series correlation feature vector; an inlet pipe-return pipe temperature significant fusion module 813, which is used to combine the inlet pipe temperature local time series correlation feature vector and the return pipe temperature local time series correlation feature vector into a vector. The feature vector is matched through a significant fusion network based on feature principal component query to obtain a sparse significant fusion representation vector of the inlet-return pipe temperature; the HVAC equipment state-room temperature state local temporal interaction module 814 is used to obtain a HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector by combining the sparse significant fusion representation vector of the inlet-return pipe temperature and the indoor temperature local temporal association feature vector through a local fine-grained sequence aggregation responder based on feature decoupling; the HVAC equipment temperature adjustment result generation module 815 is used to obtain the HVAC equipment temperature adjustment result based on the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector, and the HVAC equipment temperature adjustment result is used to indicate whether the temperature of the HVAC equipment should increase, decrease or remain unchanged.

[0034] Specifically, in the technical solution of the present application, first, the time series of the inlet pipe temperature data, the time series of the return pipe temperature data, and the time series of the indoor temperature data transmitted by the edge server are obtained. Next, considering that the inlet pipe temperature data, the return pipe temperature data, and the indoor temperature data are constantly changing with the passage of time, they all have characteristics in the time dimension, and there are different time series change trends and fluctuations in each local time segment. Based on this, in the technical solution of the present application, the time series of the inlet pipe temperature data, the time series of the return pipe temperature data, and the time series of the indoor temperature data are respectively vectorized and then used through a 1D-CNN-based time encoder to capture and refine the time series features in each local time segment, and obtain the local time series correlation feature vector of the inlet pipe temperature, the local time series correlation feature vector of the return pipe temperature, and the local time series correlation feature vector of the indoor temperature.

[0035] Accordingly, considering that the local time series correlation feature vector of the inlet pipe temperature and the local time series correlation feature vector of the return pipe temperature both have their own key main component feature information, but also contain a lot of noise and unwanted feature information, and there is also a relationship and interaction between the two. In order to effectively remove redundant information and highlight their respective important features and interactive correlation features, in the technical solution of the present application, the local time series correlation feature vector of the inlet pipe temperature and the local time series correlation feature vector of the return pipe temperature are passed through a significant fusion network based on feature principal component query matching to obtain a sparse significant fusion representation vector of the inlet pipe-return pipe temperature. In particular, the significant fusion network based on feature principal component query matching is a technology that integrates feature sparsification, principal component analysis (PCA) and feature significance fusion, aiming to create a sparse and prominent correlation representation between feature vectors. In detail, first, the influence of different eigenvectors due to different dimensions is eliminated by standardizing the local time series correlation feature vector of the inlet pipe temperature and the local time series correlation feature vector of the return pipe temperature to obtain the standardized local time series correlation feature vector of the inlet pipe temperature and the standardized local time series correlation feature vector of the return pipe temperature. Then, the sample covariance matrix of the standardized features is calculated to provide a basis and basis for the subsequent principal component analysis, and the sample covariance matrix is ​​subjected to matrix decomposition-based feature vector extraction to obtain a set of principal component feature vectors of the inlet pipe temperature and a set of principal component feature vectors of the return pipe temperature. It should be understood that principal component analysis (PCA) is a commonly used dimensionality reduction technique that can extract the most important features, namely the principal components, from the original data. By matrix decomposition, the dimension of the data can be reduced while retaining the most important temperature information of the inlet pipe and the return pipe. Afterwards, the maximum approximate query matching based on the similarity between the feature vectors is performed on the sets of the two sets of feature vectors. This process involves calculating the similarity measure between each pair of feature vectors in order to perform the most relevant semantic matching, thereby identifying the best matching pair. Subsequently, the best matching pairs selected are semantically integrated based on fine-grainedness through a gating mechanism to enhance and highlight the semantic association and importance of the inlet and return water temperatures in each pair of matching features, thereby constructing a set of principal component fusion feature vectors of the inlet-return temperature. Finally, the set of fused feature vectors is cascaded to obtain the sparse significant fusion representation vector of the inlet-return temperature. In other words, sparse representation is an effective feature representation method that can remove redundant information in the inlet and return water temperatures and highlight important features. In particular, in HVAC systems, temperature data may contain a lot of noise and irrelevant information, and sparse representation can improve the model's ability to represent key features.

[0036] Correspondingly, the inlet-return pipe temperature significant fusion module 813 includes: an inlet-return pipe temperature local time series feature standardization unit, which is used to standardize the inlet pipe temperature local time series associated feature vector and the return pipe temperature local time series associated feature vector to obtain a standardized inlet pipe temperature local time series associated feature vector and a standardized return pipe temperature local time series associated feature vector; an inlet-return pipe temperature sample covariance matrix calculation unit, which is used to calculate the sample covariance matrix of the standardized inlet pipe temperature local time series associated feature vector and the standardized return pipe temperature local time series associated feature vector to obtain an inlet pipe temperature sample covariance matrix and a return pipe temperature sample covariance matrix; a matrix decomposition unit, which is used to extract feature vectors based on matrix decomposition from the inlet pipe temperature sample covariance matrix and the return pipe temperature sample covariance matrix to obtain a set of inlet pipe temperature principal component feature vectors and a return pipe temperature principal component feature vector. A set of pipe temperature principal component feature vectors; an inlet-return pipe temperature maximum approximate query unit, used to input the set of inlet pipe temperature principal component feature vectors and the set of return pipe temperature principal component feature vectors into a maximum approximate query matching network to obtain a set of best matching pairs of inlet pipe temperature principal component feature vectors and return pipe temperature principal component feature vectors; an inlet-return pipe temperature principal component fusion unit, used to input the best matching pairs of each inlet pipe temperature principal component feature vector and return pipe temperature principal component feature vector in the set of best matching pairs of inlet pipe temperature principal component feature vectors and return pipe temperature principal component feature vectors into a semantic fine-grained gated joint module to obtain a set of inlet-return pipe temperature principal component fusion feature vectors; a vector cascade unit, used to cascade the set of inlet-return pipe temperature principal component fusion feature vectors to obtain the inlet-return pipe temperature sparse significant fusion representation vector.

[0037] Among them, the water inlet pipe-return pipe temperature maximum approximate query unit is used to: extract a predetermined water inlet pipe temperature principal component feature vector from the set of water inlet pipe temperature principal component feature vectors; calculate the cosine similarity between the predetermined water inlet pipe temperature principal component feature vector and each return pipe temperature principal component feature vector in the set of return pipe temperature principal component feature vectors to obtain a set of matching query similarities; and use the return pipe temperature principal component feature vector corresponding to the largest matching query similarity in the set of matching query similarities and the predetermined water inlet pipe temperature principal component feature vector as the best matching pair of the predetermined water inlet pipe temperature principal component feature vector and the return pipe temperature principal component feature vector.

[0038] Among them, the local time series characteristic standardization unit of the inlet pipe and return pipe temperature is used to: respectively calculate the mean and standard deviation of the local time series correlation characteristic vector of the inlet pipe temperature to obtain the mean value of the local time series correlation characteristic of the inlet pipe temperature and the standard deviation of the local time series correlation characteristic of the inlet pipe temperature; after subtracting the local time series correlation characteristic vector of the inlet pipe temperature from the mean value of the local time series correlation characteristic of the inlet pipe temperature by position, divide the calculated local time series correlation offset vector of the inlet pipe temperature by the standard deviation of the local time series correlation characteristic of the inlet pipe temperature by position to obtain the standardized inlet pipe temperature. pipe temperature local time series associated feature vector; respectively calculate the mean and standard deviation of the return pipe temperature local time series associated feature vector to obtain the return pipe temperature local time series associated feature mean and the return pipe temperature local time series associated feature standard deviation; after subtracting the return pipe temperature local time series associated feature vector from the return pipe temperature local time series associated feature mean by position, the calculated return pipe temperature local time series associated offset vector and the return pipe temperature local time series associated feature standard deviation are divided by position to obtain the standardized return pipe temperature local time series associated feature vector.

[0039] Among them, the inlet pipe return pipe temperature sample covariance matrix calculation unit is used to: after multiplying the transposed vector of the standardized inlet pipe temperature local time series correlation feature vector with the standardized inlet pipe temperature local time series correlation feature vector, the obtained standardized inlet pipe temperature correlation matrix is ​​divided by the value obtained by subtracting one from the length of the standardized inlet pipe temperature local time series correlation feature vector to obtain the inlet pipe temperature sample covariance matrix; after multiplying the transposed vector of the standardized return pipe temperature local time series correlation feature vector with the standardized return pipe temperature local time series correlation feature vector, the obtained standardized return pipe temperature correlation matrix is ​​divided by the value obtained by subtracting one from the length of the standardized return pipe temperature local time series correlation feature vector to obtain the return pipe temperature sample covariance matrix.

[0040] Among them, the inlet-return pipe temperature principal component fusion unit is used to: respectively calculate the position difference, position dot multiplication and position addition between the best matching pairs of the inlet pipe temperature principal component eigenvector and the return pipe temperature principal component eigenvector to obtain the inlet pipe-return pipe temperature principal component pattern difference vector, the inlet pipe-return pipe temperature principal component pattern dot product vector and the inlet pipe-return pipe temperature principal component pattern sum vector; cascade the inlet pipe-return pipe temperature principal component pattern difference vector, the inlet pipe-return pipe temperature principal component pattern dot product vector and the inlet pipe-return pipe temperature principal component pattern sum vector, and then perform one-dimensional convolution encoding to obtain the inlet pipe-return pipe temperature principal component pattern multi-dimensional fusion vector; perform local window-based maximum pooling processing on the inlet pipe-return pipe temperature principal component pattern multi-dimensional fusion vector to obtain the inlet pipe-return pipe temperature principal component fusion feature vector.

[0041] In one example, the water inlet-return pipe temperature significant fusion module 813 is used to: process the water inlet pipe temperature local time series correlation feature vector and the water return pipe temperature local time series correlation feature vector through a significant fusion network based on feature principal component query matching, using the following significant fusion formula to obtain the water inlet-return pipe temperature sparse significant fusion representation vector; wherein the significant fusion formula is: ;in, and represent the local time series correlation feature vector of the inlet pipe temperature and the local time series correlation feature vector of the return pipe temperature respectively, and are the means of the local time series correlation feature vector of the inlet pipe temperature and the local time series correlation feature vector of the return pipe temperature, respectively. and are the standard deviations of the local time series correlation feature vector of the inlet pipe temperature and the local time series correlation feature vector of the return pipe temperature, and are respectively the local time series correlation feature vector of the standardized water inlet pipe temperature and the local time series correlation feature vector of the standardized water return pipe temperature, and They are and The transposed vector of and are the number of eigenvalues ​​in the standardized local time series correlation feature vector of the inlet pipe temperature and the standardized local time series correlation feature vector of the return pipe temperature, respectively, and are the water inlet pipe temperature sample covariance matrix and the water return pipe temperature sample covariance matrix, and They are the orthogonal matrix of the principal component of the inlet pipe temperature and the orthogonal matrix of the principal component of the return pipe temperature, and are the inlet pipe temperature diagonal matrix and the return pipe temperature diagonal matrix respectively, The diagonal elements of the matrix are The diagonal matrix of the inlet pipe temperature, are the weight values ​​of the principal component eigenvectors of each inlet pipe temperature, The diagonal elements of the matrix are The diagonal matrix of return pipe temperature, are the weight values ​​of the principal component eigenvectors of the return pipe temperatures, and They are and The transposed matrix of is each water inlet pipe temperature principal component eigenvector in the set of water inlet pipe temperature principal component eigenvectors, is each water return pipe temperature principal component eigenvector in the set of water return pipe temperature principal component eigenvectors, To calculate the and The vector inner product between To calculate the one-norm of a vector, To return the maximum value value, is the maximum approximate matching value, , and They are position difference, position dot multiplication and position addition respectively. For cascade processing, is a one-dimensional convolutional coding operation, is the maximum pooling operation, is the first in the set of principal component fusion feature vectors of the inlet pipe-return pipe temperature The principal component fusion feature vector of the inlet-return pipe temperature, is the number of eigenvectors in the set of the principal component fusion eigenvectors of the inlet-return pipe temperature, is the sparse and significant fusion representation vector of the inlet-return pipe temperature.

[0042] Furthermore, considering that the sparsely significant fused representation vector of the inlet-return pipe temperature and the local temporal correlation feature vector of the indoor temperature have different degrees of correlation and interactive influence in the local time period, in order to be able to capture the slight temperature changes and response details in the data, so as to more accurately understand the subtle changes in the HVAC equipment status and indoor temperature. In the technical solution of the present application, the sparsely significant fused representation vector of the inlet-return pipe temperature and the local temporal correlation feature vector of the indoor temperature are aggregated through a feature-decoupled local fine-grained sequence responder to obtain a HVAC equipment status-room temperature state fine-grained local temporal semantic interactive response feature vector. That is, the feature-decoupled local fine-grained sequence aggregate responder is good at identifying and understanding subtle semantic details and correlations between features. Specifically, first, the sparse significant fusion representation vector of the inlet-return pipe temperature and the local temporal association feature vector of the indoor temperature are feature decoupled to obtain a sequence of the local sparse significant fusion representation vector of the inlet-return pipe temperature and a sequence of the local temporal decoupling association feature vector of the indoor temperature, so as to divide the entire vector into multiple local time segments, so as to capture and extract more local details and patterns. Then, the local semantic interaction vector between the sequences of the two local vectors is calculated to obtain the sequence of the local temporal semantic association interaction vector of the inlet-return pipe-indoor temperature, and the fine-grained attention weight of each local semantic interaction vector is calculated to obtain the sequence of the local temporal fine-grained attention weight of the inlet-return pipe-indoor temperature, so as to highlight the important key information in the interaction vector. That is, by calculating the local semantic interaction vector, the complex relationship and dynamic interaction between the two inlet and return water temperature sequences can be captured, which is crucial for understanding the deep-level features and patterns of the data. The fine-grained attention weight allows the model to weight each part of the interaction vector, so as to highlight the most important features and information and improve the sensitivity and accuracy of the model. Then, the sequence of fine-grained attention weights is input into the random dropout module to obtain the sequence of fine-grained attention weights of the local time series of the water inlet and return pipes-indoor temperature after the dropout treatment. It should be understood that random dropout is a form of regularization, which can reduce the overfitting of the model on the training data. By randomly "closing" or "deactivating" some neurons in the network in each training iteration (that is, temporarily not using these neurons for calculation), the model is prevented from being overly dependent on specific neurons or features, so that those key interactive feature information can be more prominent. Finally, using the sequence of fine-grained attention weights after the dropout treatment as the weight distribution, the sequence of the local time series semantic association interaction vectors of the water inlet and return pipes-indoor temperature is weighted and summed to obtain the fine-grained local time series semantic interaction response feature vector of the HVAC equipment state-room temperature state.

[0043] Correspondingly, the HVAC equipment state-room temperature state local time series interaction module 814 includes: a feature decoupling unit, which is used to perform feature decoupling on the inlet-return pipe temperature sparse significant fusion representation vector and the indoor temperature local time series associated feature vector to obtain a sequence of inlet-return pipe temperature local sparse significant fusion representation vectors and a sequence of indoor temperature local time series decoupling associated feature vectors; an inlet-return pipe-indoor temperature local time series semantic association interaction unit, which is used to calculate the local semantic association interaction vector between each group of corresponding inlet-return pipe temperature local sparse significant fusion representation vectors and indoor temperature local time series decoupling associated feature vectors in the sequence of inlet-return pipe temperature local sparse significant fusion representation vectors and the sequence of indoor temperature local time series decoupling associated feature vectors to obtain a sequence of inlet-return pipe-indoor temperature local time series semantic association interaction vectors; A sequence fine-grained attention weight calculation unit is used to calculate the fine-grained attention weights of each water inlet return pipe-indoor temperature local temporal sequence semantic association interaction vector in the sequence of the water inlet return pipe-indoor temperature local temporal sequence semantic association interaction vector to obtain a sequence of water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights; a deactivation processing unit is used to input the sequence of the water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights into a random deactivation module to obtain a sequence of the water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights after deactivation processing; a weighted sum calculation unit is used to use the sequence of the water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights after the deactivation processing as the weight distribution, and calculate the weighted sum of the sequence of the water inlet return pipe-indoor temperature local temporal sequence semantic association interaction vectors to obtain the HVAC equipment state-room temperature state fine-grained local temporal sequence semantic interaction response feature vector.

[0044] Among them, the water inlet pipe return pipe-indoor temperature local temporal semantic association interaction unit is used to: respectively calculate the position difference, position point product and position addition between the water inlet pipe-return pipe temperature local sparse significant fusion representation vector and the indoor temperature local temporal decoupling association feature vector to obtain the water inlet pipe return pipe-indoor temperature local temporal difference vector, the water inlet pipe return pipe-indoor temperature local temporal point product vector and the water inlet pipe return pipe-indoor temperature local temporal sum vector; cascade the water inlet pipe return pipe-indoor temperature local temporal difference vector, the water inlet pipe return pipe-indoor temperature local temporal point product vector and the water inlet pipe return pipe-indoor temperature local temporal sum vector, and perform one-dimensional convolution encoding to obtain the water inlet pipe return pipe-indoor temperature local temporal multi-dimensional fusion vector; perform local window-based maximum pooling processing on the water inlet pipe return pipe-indoor temperature local temporal multi-dimensional fusion vector to obtain the water inlet pipe return pipe-indoor temperature local temporal semantic association interaction vector.

[0045] Among them, the water inlet pipe return pipe-indoor temperature local temporal series fine-grained attention weight calculation unit is used to: multiply each water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction vector in the sequence of the water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction vector by the weight matrix, and then add the obtained vector and the bias vector by position to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction modulation vectors; input the sequence of water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction modulation vectors into the tanh function for activation processing to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction activation vectors; multiply each water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction activation vector in the sequence of the water inlet pipe return pipe-indoor temperature local temporal series semantic association interaction activation vectors by the weight coefficient vector to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal series fine-grained attention weights.

[0046] In one example, the HVAC equipment state-room temperature state local time series interaction module 814 is used to: process the sparse significant fusion representation vector of the inlet-return pipe temperature and the indoor temperature local time series association feature vector through a feature decoupled local fine-grained sequence aggregation responder, and obtain the HVAC equipment state-room temperature state fine-grained local time series semantic interaction response feature vector by the following fine-grained aggregation; wherein the fine-grained aggregation formula is: ;in, and are the sparse significant fusion representation vector of the inlet-return pipe temperature and the local time series correlation feature vector of the indoor temperature, respectively. For feature decoupling operations, and They are the sequence of the local sparse significant fusion representation vector of the inlet-return pipe temperature and the sequence of the local temporal decoupling correlation feature vector of the indoor temperature. The local sparse and significant fusion representation vector of the inlet-return pipe temperature and the The local temporal decoupling correlation feature vector of indoor temperature, The number of feature vectors in the sequence of the local sparse significant fusion representation vectors of the inlet-return pipe temperature and the sequence of the local temporal decoupling correlation feature vectors of the indoor temperature, and They are The local sparse significant fusion representation vector of the inlet-return pipe temperature and the local temporal decoupling correlation feature vector of the indoor temperature, , and They are position difference, position dot multiplication and position addition respectively. For cascade processing, is a one-dimensional convolutional coding operation, is the maximum pooling operation, is the first in the sequence of the local temporal semantic association interaction vectors of the water inlet pipe, water return pipe and indoor temperature. The local time series semantic association interaction vector of water inlet pipe, water return pipe and indoor temperature, is the weight matrix, is the bias vector, yes function, is the weight coefficient vector, is the first in the sequence of the local time series fine-grained attention weights of the water inlet pipe, water return pipe and indoor temperature The inlet and return pipes-indoor temperature local time series fine-grained attention weight, Indicates deactivation processing, is the first in the sequence of the local time series fine-grained attention weights of the inlet pipe, return pipe and indoor temperature after the deactivation process. After deactivation, the inlet pipe, return pipe and indoor temperature local time series fine-grained attention weights, It is the fine-grained local temporal semantic interaction response feature vector of the HVAC equipment state-room temperature state.

[0047] Then, the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector is passed through a classifier-based HVAC equipment temperature adjuster to obtain the HVAC equipment temperature adjustment result, wherein the HVAC equipment temperature adjustment result is used to indicate whether the temperature of the HVAC equipment should increase, decrease, or remain unchanged. That is, the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector obtained by the local fine-grained interactive response using the inlet-return pipe temperature sparse significant fusion representation vector and the indoor temperature local temporal association feature vector is classified and processed to adaptively adjust the temperature of the HVAC equipment. In this way, environmental changes can be quickly identified and corresponding adjustments can be made, thereby improving the response speed to immediate changes. At the same time, temperature fluctuations in a short period of time can be captured, thereby achieving more accurate and intelligent temperature control.

[0048] Correspondingly, the HVAC equipment temperature adjustment result generation module 815 is used to: pass the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector through a classifier-based HVAC equipment temperature adjuster to obtain the HVAC equipment temperature adjustment result, and the HVAC equipment temperature adjustment result is used to indicate whether the temperature of the HVAC equipment should increase, decrease or remain unchanged.

[0049] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression, SVM, etc. are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification, but this is prone to errors and inefficient. Commonly used multi-classification methods include the Softmax classification function.

[0050] In a preferred example, the HVAC equipment temperature adjustment result generation module 815 includes: an optimization unit, used to optimize the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector to obtain an optimized HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector; a classification unit, used to pass the optimized HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector through the classifier-based HVAC equipment temperature adjuster to obtain the HVAC equipment temperature adjustment result.

[0051] Wherein, the optimization unit is used to: determine the maximum eigenvalue of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response and the minimum eigenvalue of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector, and calculate the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response mean and HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response standard deviation of the feature set of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector; calculate the quotient of the mean of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response and the standard deviation of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response to obtain the statistical normalization value of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response; calculate the inverse of each eigenvalue of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector, and compare it with the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response standard deviation. After multiplying the difference between the maximum eigenvalue of the fine-grained local temporal semantic interaction response of the room temperature state and the minimum eigenvalue of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state, the result is point-subtracted from the statistical normalized value of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state to obtain the distribution approximation vector of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state; an exponential function with a natural constant as the base, in which each eigenvalue of the distribution approximation vector of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state is exponentially calculated, and the result is point-subtracted from the statistical normalized value of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state to obtain the probability approximation vector of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state; the logarithm with base two of the absolute value of each eigenvalue of the probability approximation vector of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state is calculated to obtain the optimized characteristic vector of the fine-grained local temporal semantic interaction response of the HVAC equipment state-room temperature state.

[0052] The optimization unit is used to optimize the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector using the following optimization formula to obtain the optimized HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector; wherein the optimization formula is expressed as: ;in, is the fine-grained local temporal semantic interaction response feature vector of the HVAC equipment state-room temperature state, , , and They are the feature sets of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vectors The mean, standard deviation, minimum and maximum values ​​of Indicates point reduction, represents dot product, Indicates point addition, To calculate the inverse of each eigenvalue of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response eigenvector, is an exponential operation of a value, wherein the exponential operation of the value represents the calculation of the value of a natural exponential function with the value as a power, and is the logarithmic function operation with base 2, It is the optimized HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector.

[0053] Here, in the preferred example, considering that the inlet-return pipe temperature sparse significant fusion representation vector and the indoor temperature local temporal association feature vector respectively represent the temporal principal component significant fusion features of the one-dimensional local temporal association features of the inlet-return pipe temperature and the one-dimensional local temporal association features of the indoor temperature, when performing local fine-grained sequence aggregation response based on feature decoupling, the local fine-grained differences in the temporal sequence aggregation response caused by the distribution differences under different feature temporal association dimensions will make the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector have a diversified set expression distribution of temporal aggregation response features. Therefore, it is expected to improve the balance between the regression mapping accuracy and completeness of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector when performing class regression through a classifier-based HVAC equipment temperature adjuster, thereby improving the accuracy of the HVAC equipment temperature adjustment result obtained.

[0054] Therefore, by performing random statistical normalization on the diversified feature set of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector, the confidence space constructed based on the overall feature value of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector is approximated by the standardized continuous probability density distribution relative to the response hypothesis test of each feature value of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector, thereby establishing the target reachability of the diversified feature distribution of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector to the unified regression target, so as to achieve the balanced executableness between mapping accuracy and mapping completeness in the class regression process based on the diversified feature distribution of the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector, and improve the accuracy of the HVAC equipment temperature adjustment result obtained by the HVAC equipment temperature adjuster based on the classifier of the HVAC equipment state-room temperature fine-grained local temporal semantic interaction response feature vector. In this way, environmental changes can be quickly identified and corresponding adjustments can be made, which improves the response speed to immediate changes. At the same time, it can capture temperature fluctuations in a short period of time, thereby achieving more precise and intelligent temperature control.

[0055] Based on the above embodiments, see Figure 3 As shown, it is a flow chart of a smart HVAC terminal control method with collaborative cloud data storage function in an embodiment of the present application. Figure 3As shown, according to the embodiment of the present application, the intelligent HVAC terminal control method with collaborative cloud data storage function includes the following steps: S510, obtaining the time series of water inlet pipe temperature data, the time series of return pipe temperature data and the time series of indoor temperature data transmitted by the edge server; S520, arranging the time series of the water inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data into vectors respectively and then passing through a 1D-CNN-based time series encoder to obtain a local time series correlation feature vector of the water inlet pipe temperature, a local time series correlation feature vector of the return pipe temperature and a local time series correlation feature vector of the indoor temperature; S530, arranging the local time series correlation feature vector of the water inlet pipe temperature and the time series of the return pipe temperature into vectors respectively and then passing through a 1D-CNN-based time series encoder to obtain a local time series correlation feature vector of the water inlet pipe temperature, a local time series correlation feature vector of the return pipe temperature and a local time series correlation feature vector of the indoor temperature; S530, The local time series associated feature vector of the return pipe temperature is obtained through a significant fusion network based on feature principal component query matching to obtain a sparse significant fusion representation vector of the inlet-return pipe temperature; S540, the sparse significant fusion representation vector of the inlet-return pipe temperature and the indoor temperature local time series associated feature vector are obtained through a local fine-grained sequence aggregation responder based on feature decoupling to obtain a HVAC equipment state-room temperature state fine-grained local time series semantic interaction response feature vector; S550, based on the HVAC equipment state-room temperature state fine-grained local time series semantic interaction response feature vector, the HVAC equipment temperature adjustment result is obtained, and the HVAC equipment temperature adjustment result is used to indicate that the temperature of the HVAC equipment should increase, decrease or remain unchanged.

[0056] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned intelligent HVAC terminal control method with collaborative cloud data storage function have been referred to above. Figure 1 to Figure 2 The description of the smart HVAC terminal 800 with collaborative cloud data storage function has been introduced in detail, and therefore, its repeated description will be omitted.

[0057] Figure 4 FIG. 1 is an application scenario diagram of an intelligent HVAC terminal with collaborative cloud data storage function according to an embodiment of the present application. Figure 4 As shown, in this application scenario, first, the time series of the inlet pipe temperature data, the return pipe temperature data, and the indoor temperature data transmitted by the edge server are obtained (for example, Figure 4 D), then, the time series of the water inlet pipe temperature data, the time series of the return pipe temperature data, and the time series of the indoor temperature data are input to a server (for example, Figure 4S) as shown in the figure, wherein the server is able to use the intelligent HVAC terminal control algorithm with collaborative cloud data storage function to process the time series of the water inlet pipe temperature data, the time series of the return pipe temperature data and the time series of the indoor temperature data to obtain a HVAC equipment temperature adjustment result indicating that the temperature of the HVAC equipment should increase, decrease or remain unchanged.

[0058] Based on the above embodiments, another exemplary embodiment of an electronic device is also provided in the embodiments of the present application. In some possible implementations, the electronic device in the embodiments of the present application may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of the intelligent HVAC terminal control method with collaborative cloud data storage function in the above embodiments when executing the program.

[0059] The embodiment of the present application also provides a computer-readable storage medium, on which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the intelligent HVAC terminal control method with collaborative cloud data storage function according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc.

[0060] The embodiment of the present application also provides a computer program product or a computer program, which includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. The processor of the computer device reads the computer executable instructions from the computer readable storage medium, and the processor executes the computer executable instructions, so that the computer device executes the intelligent HVAC terminal control method with collaborative cloud data storage function according to the embodiment of the present application.

[0061] Those skilled in the art will appreciate that the contents disclosed in this application may be subject to various variations and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.

[0062] In addition, although the present application makes various references to certain units in the system according to embodiments of the present application, any number of different units can be used and run on the client and / or server. The units are only illustrative, and different aspects of the system and method can use different units.

[0063] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of software function modules. The present application is not limited to any particular form of combination of hardware and software.

[0064] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined as such herein.

[0065] The above is an explanation of the present application and should not be considered as limiting thereof. Although several exemplary embodiments of the present application have been described, those skilled in the art will readily appreciate that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present application.

Claims

1. An intelligent HVAC terminal with collaborative cloud data storage function, comprising: A cloud server, an edge server, a collection terminal, a control terminal and a HVAC device; characterized in that the collection terminal comprises an inlet water temperature sensor, a return water temperature sensor and a room temperature sensor, the inlet water temperature sensor and the return water temperature sensor are respectively deployed in the inlet pipe and the return pipe of the HVAC device, and are used to collect the inlet pipe temperature data and the return pipe temperature data of the HVAC device; the room temperature sensor is used to collect indoor temperature data; the edge server is used to upload the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data to the cloud server for storage; the cloud server is used to analyze the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data to obtain the HVAC device temperature adjustment result, and transmit the HVAC device temperature adjustment result to the control terminal; the control terminal is used to control the power of the HVAC device according to the HVAC device temperature adjustment result; The cloud server includes: a data acquisition module, which is used to acquire the time series of the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data transmitted by the edge server; a time series encoding module, which is used to arrange the time series of the inlet pipe temperature data, the return pipe temperature data and the indoor temperature data into vectors respectively and then pass them through a time series encoder based on 1D-CNN to obtain the inlet pipe temperature local time series correlation feature vector, the return pipe temperature local time series correlation feature vector and the indoor temperature local time series correlation feature vector; an inlet pipe-return pipe temperature significant fusion module, which is used to pass the inlet pipe temperature local time series correlation feature vector and the return pipe temperature local time series correlation feature vector into a vector. A sparse significant fusion representation vector of the inlet-return pipe temperature is obtained through a significant fusion network based on feature principal component query matching; a HVAC equipment state-room temperature state local temporal interaction module is used to obtain a HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector by combining the inlet-return pipe temperature sparse significant fusion representation vector and the indoor temperature local temporal association feature vector through a feature decoupled local fine-grained sequence aggregation responder; a HVAC equipment temperature adjustment result generation module is used to obtain the HVAC equipment temperature adjustment result based on the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector, and the HVAC equipment temperature adjustment result is used to indicate whether the temperature of the HVAC equipment should increase, decrease or remain unchanged.

2. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 1 is characterized in that: The water inlet-return pipe temperature significant fusion module includes: a water inlet-return pipe temperature local time series feature normalization unit, which is used to normalize the water inlet temperature local time series associated feature vector and the return pipe temperature local time series associated feature vector to obtain a standardized water inlet temperature local time series associated feature vector and a standardized return pipe temperature local time series associated feature vector; a water inlet-return pipe temperature sample covariance matrix calculation unit, which is used to calculate the sample covariance matrix of the standardized water inlet temperature local time series associated feature vector and the standardized return pipe temperature local time series associated feature vector to obtain the water inlet temperature sample covariance matrix. a matrix and a covariance matrix of the return pipe temperature samples; a matrix decomposition unit, used for performing eigenvector extraction based on matrix decomposition on the inlet pipe temperature sample covariance matrix and the return pipe temperature sample covariance matrix to obtain a set of inlet pipe temperature principal component eigenvectors and a set of return pipe temperature principal component eigenvectors; an inlet pipe-return pipe temperature maximum approximation query unit, used for inputting the set of inlet pipe temperature principal component eigenvectors and the set of return pipe temperature principal component eigenvectors into a maximum approximation query matching network to obtain a set of best matching pairs of inlet pipe temperature principal component eigenvectors and return pipe temperature principal component eigenvectors; The water pipe-return pipe temperature principal component fusion unit is used to input the best matching pairs of the water pipe temperature principal component feature vector and the return pipe temperature principal component feature vector in the set of the best matching pairs of the water pipe temperature principal component feature vector and the return pipe temperature principal component feature vector into the semantic fine-grained gated joint module to obtain the set of water pipe-return pipe temperature principal component fusion feature vectors; the vector cascade unit is used to cascade the set of the water pipe-return pipe temperature principal component fusion feature vectors to obtain the water pipe-return pipe temperature sparse significant fusion representation vector; wherein the water pipe-return pipe temperature maximum approximate query unit is used The invention relates to: extracting a predetermined water inlet pipe temperature principal component feature vector from the set of water inlet pipe temperature principal component feature vectors; calculating the cosine similarity between the predetermined water inlet pipe temperature principal component feature vector and each return water pipe temperature principal component feature vector in the set of return water pipe temperature principal component feature vectors to obtain a set of matching query similarities; and taking the return water pipe temperature principal component feature vector corresponding to the largest matching query similarity in the set of matching query similarities and the predetermined water inlet pipe temperature principal component feature vector as the best matching pair of the predetermined water inlet pipe temperature principal component feature vector and the return water pipe temperature principal component feature vector.

3. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 2 is characterized in that: The water inlet and return pipe temperature local time series feature standardization unit is used to: respectively calculate the mean and standard deviation of the water inlet temperature local time series associated feature vector to obtain the water inlet temperature local time series associated feature mean and the water inlet temperature local time series associated feature standard deviation; after subtracting the water inlet temperature local time series associated feature vector from the water inlet temperature local time series associated feature mean by position, the calculated water inlet temperature local time series associated offset vector and the water inlet temperature local time series associated feature standard deviation are divided by position to obtain the standardized water inlet temperature local time series associated offset vector. Temperature local time series associated feature vector; respectively calculate the mean and standard deviation of the return pipe temperature local time series associated feature vector to obtain the return pipe temperature local time series associated feature mean and the return pipe temperature local time series associated feature standard deviation; after subtracting the return pipe temperature local time series associated feature vector from the return pipe temperature local time series associated feature mean by position, the calculated return pipe temperature local time series associated offset vector and the return pipe temperature local time series associated feature standard deviation are divided by position to obtain the standardized return pipe temperature local time series associated feature vector.

4. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 3 is characterized in that: The inlet and return pipe temperature sample covariance matrix calculation unit is used to: multiply the transposed vector of the standardized inlet pipe temperature local time series associated feature vector by the standardized inlet pipe temperature local time series associated feature vector, and then divide the obtained standardized inlet pipe temperature association matrix by the value obtained by subtracting one from the length of the standardized inlet pipe temperature local time series associated feature vector to obtain the inlet pipe temperature sample covariance matrix; multiply the transposed vector of the standardized return pipe temperature local time series associated feature vector by the standardized return pipe temperature local time series associated feature vector, and then divide the obtained standardized return pipe temperature association matrix by the value obtained by subtracting one from the length of the standardized return pipe temperature local time series associated feature vector to obtain the return pipe temperature sample covariance matrix.

5. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 4 is characterized in that: The inlet-return pipe temperature principal component fusion unit is used to: respectively calculate the position difference, position dot product and position addition between the best matching pairs of the inlet pipe temperature principal component feature vector and the return pipe temperature principal component feature vector to obtain the inlet pipe-return pipe temperature principal component pattern difference vector, the inlet pipe-return pipe temperature principal component pattern dot product vector and the inlet pipe-return pipe temperature principal component pattern sum vector; cascade the inlet pipe-return pipe temperature principal component pattern difference vector, the inlet pipe-return pipe temperature principal component pattern dot product vector and the inlet pipe-return pipe temperature principal component pattern sum vector, and then perform one-dimensional convolution encoding to obtain the inlet pipe-return pipe temperature principal component pattern multi-dimensional fusion vector; perform local window-based maximum pooling processing on the inlet pipe-return pipe temperature principal component pattern multi-dimensional fusion vector to obtain the inlet pipe-return pipe temperature principal component fusion feature vector.

6. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 5 is characterized in that: The HVAC equipment state-room temperature state local time series interaction module includes: a feature decoupling unit, which is used to perform feature decoupling on the inlet-return pipe temperature sparse significant fusion representation vector and the indoor temperature local time series associated feature vector to obtain a sequence of inlet-return pipe temperature local sparse significant fusion representation vectors and a sequence of indoor temperature local time series decoupling associated feature vectors; an inlet-return pipe-indoor temperature local time series semantic association interaction unit, which is used to calculate the local semantic association interaction vector between each group of corresponding inlet-return pipe temperature local sparse significant fusion representation vectors and indoor temperature local time series decoupling associated feature vectors in the sequence of inlet-return pipe temperature local sparse significant fusion representation vectors and the sequence of indoor temperature local time series decoupling associated feature vectors to obtain a sequence of inlet-return pipe-indoor temperature local time series semantic association interaction vectors; an inlet-return pipe-indoor temperature local time series fine-grained semantic association interaction unit. An attention weight calculation unit is used to calculate the fine-grained attention weights of each water inlet return pipe-indoor temperature local temporal sequence semantic association interaction vector in the sequence of the water inlet return pipe-indoor temperature local temporal sequence semantic association interaction vectors to obtain a sequence of the water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights; a deactivation processing unit is used to input the sequence of the water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights into a random deactivation module to obtain a sequence of the water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights after deactivation processing; a weighted sum calculation unit is used to use the sequence of the water inlet return pipe-indoor temperature local temporal sequence fine-grained attention weights after the deactivation processing as the weight distribution, and calculate the weighted sum of the sequence of the water inlet return pipe-indoor temperature local temporal sequence semantic association interaction vectors to obtain the HVAC equipment state-room temperature state fine-grained local temporal sequence semantic interaction response feature vector.

7. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 6 is characterized in that: The water inlet pipe return pipe-indoor temperature local temporal semantic association interaction unit is used to: respectively calculate the position difference, position point product and position addition between the water inlet pipe-return pipe temperature local sparse significant fusion representation vector and the indoor temperature local temporal decoupling association feature vector to obtain the water inlet pipe return pipe-indoor temperature local temporal difference vector, the water inlet pipe return pipe-indoor temperature local temporal point product vector and the water inlet pipe return pipe-indoor temperature local temporal sum vector; cascade the water inlet pipe return pipe-indoor temperature local temporal difference vector, the water inlet pipe return pipe-indoor temperature local temporal point product vector and the water inlet pipe return pipe-indoor temperature local temporal sum vector, and perform one-dimensional convolution encoding to obtain the water inlet pipe return pipe-indoor temperature local temporal multi-dimensional fusion vector; perform local window-based maximum pooling processing on the water inlet pipe return pipe-indoor temperature local temporal multi-dimensional fusion vector to obtain the water inlet pipe return pipe-indoor temperature local temporal semantic association interaction vector.

8. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 7 is characterized in that: The water inlet pipe return pipe-indoor temperature local temporal sequence fine-grained attention weight calculation unit is used to: after multiplying each water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction vector in the sequence of the water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction vector with the weight matrix, the obtained vector is added with the bias vector by position to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction modulation vectors; the sequence of water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction modulation vectors is input into the tanh function for activation processing to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction activation vectors; each water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction activation vector in the sequence of the water inlet pipe return pipe-indoor temperature local temporal sequence semantic association interaction activation vectors is multiplied by the weight coefficient vector to obtain a sequence of water inlet pipe return pipe-indoor temperature local temporal sequence fine-grained attention weights.

9. The intelligent HVAC terminal with collaborative cloud data storage function according to claim 8, characterized in that: The HVAC equipment temperature adjustment result generation module is used to: pass the HVAC equipment state-room temperature state fine-grained local temporal semantic interaction response feature vector through a classifier-based HVAC equipment temperature adjuster to obtain the HVAC equipment temperature adjustment result, and the HVAC equipment temperature adjustment result is used to indicate whether the temperature of the HVAC equipment should increase, decrease or remain unchanged.