Distributed temperature and humidity regulation and control method and system based on deep learning

Abstract

The invention relates to a distributed temperature and humidity regulation and control method and system based on deep learning, and the method comprises the steps: obtaining temperature and humidity data measured by a wireless sensor, and carrying out the assimilation processing of the temperature and humidity data through a deep learning set smoother, so as to construct a temperature and humidity data set. And calling a parameter estimation and data learning algorithm to analyze the assimilated temperature and humidity data so as to identify and extract environmental characteristics in the temperature and humidity data. And taking the environmental characteristics as input of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics in a future first time period, and generating a control vector based on the temperature and humidity change trend. And controlling the operation state of the HVAC equipment according to the temperature and humidity change trend in response to the control vector so as to automatically regulate and control the temperature and humidity of the environment in the current space. The temperature and humidity prediction model is obtained by training a deep learning model based on the real-time temperature and humidity data and the historical temperature and humidity data after assimilation processing.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent temperature and humidity control, and in particular to a distributed temperature and humidity control method and system based on deep learning. Background Art

[0002] In the context of the rapid development of modern automation and intelligence, temperature and humidity control systems play a vital role in many fields, especially in data centers, medical facilities, precision manufacturing, and agricultural greenhouses. The precise control of environmental conditions is directly related to operational efficiency and the quality of products or services. With the integration of Internet of Things (IoT) technology and the development of artificial intelligence and machine learning algorithms, temperature and humidity control systems are gradually moving towards a more intelligent and integrated direction. Modern systems tend to use a large amount of real-time data, through advanced data analysis and prediction models, to automatically adjust control strategies to achieve better energy efficiency and environmental comfort.

[0003] At present, existing temperature and humidity control technologies are usually based on traditional control logic and algorithms, such as PID control. These technologies perform well in simple applications, but often seem to be unable to cope with complex and dynamically changing environments. Because these systems tend to rely more on rule-based programs and operate according to preset parameters, such as set temperature and humidity thresholds to trigger the start and stop of HVAC systems. Although these systems achieve basic automated control, they lack sufficient flexibility and adaptability, especially in situations where rapid response to environmental changes is required, such as commercial or industrial environments.

[0004] In addition, a major limitation of existing intelligent temperature and humidity control technology is its weak ability to process and optimize data. As big data and Internet of Things technologies become more and more popular, traditional temperature and humidity control technology is unable to fully utilize the massive amounts of data collected continuously to optimize its control decision-making process. For example, it is difficult to effectively identify patterns or trends in temperature and humidity data, and thus it is difficult to achieve predictive maintenance or energy consumption optimization.

[0005] Therefore, the flexibility and intelligence of the existing temperature and humidity control technology need to be further improved, which to a certain extent affects its ability to obtain a more optimized temperature and humidity control strategy. Summary of the invention

[0006] Based on this, it is necessary to provide a distributed temperature and humidity control method and system based on deep learning with better flexibility and higher intelligence to address the above technical problems.

[0007] The present invention provides a distributed temperature and humidity control method based on deep learning, the method comprising:

[0008] Acquire temperature and humidity data measured by wireless sensors, and assimilate the temperature and humidity data through a deep learning ensemble smoother to construct a temperature and humidity data set;

[0009] Calling parameter estimation and data learning algorithms to analyze the temperature and humidity data after assimilation processing to identify and extract environmental features in the temperature and humidity data;

[0010] The environmental characteristics are used as inputs of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and a control vector is generated based on the temperature and humidity change trend;

[0011] In response to the control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically adjust the temperature and humidity of the environment in the current space;

[0012] Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and time information corresponding to each group of temperature and humidity data, the wireless sensor includes multiple temperature sensors and humidity sensors, the environmental characteristics include abnormal fluctuations in temperature and humidity data, and when the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

[0013] In one embodiment, the step of acquiring temperature and humidity data measured by a wireless sensor and assimilating the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set includes:

[0014] Real-time acquisition of real-time temperature and humidity data measured by wireless sensors in the current space and the time information corresponding to each set of real-time temperature and humidity data, and assimilation of the real-time temperature and humidity data through a deep learning set smoother;

[0015] Based on each set of historical temperature and humidity data after assimilation and its corresponding time information, the temperature and humidity data set is updated in combination with the real-time temperature and humidity data after assimilation;

[0016] Each group of the temperature sensor and humidity sensor corresponds to a sensor node, each sensor node corresponds to a group of temperature and humidity data in the temperature and humidity data set, and each group of temperature and humidity data includes a group of temperature sensor readings and humidity sensor readings and reading time.

[0017] In one embodiment, the step of acquiring temperature and humidity data measured by a wireless sensor and assimilating the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set further includes:

[0018] Based on the temperature and humidity data set, the temperature and humidity prediction model is used to obtain the environmental prediction state at the first moment output according to the historical temperature and humidity data, and the Kalman benefit parameter and the actual temperature and humidity data at the first moment are determined;

[0019] The environmental prediction state is mapped from the state space to the observation space through the observation matrix, and the deep learning ensemble smoother is called in combination with the Kalman benefit parameter and the actual temperature and humidity data to calculate the environmental state of the temperature and humidity data after data assimilation processing at the first moment.

[0020] In one embodiment, the calling of parameter estimation and data learning algorithms to analyze the assimilated temperature and humidity data to identify and extract environmental features in the temperature and humidity data includes:

[0021] The assimilated temperature and humidity data set is used as the input of the parameter estimation and data learning algorithm, so as to analyze the temperature and humidity data set by calling the deep learning model in combination with the model parameter set through the parameter estimation and data learning algorithm to obtain the temperature and humidity data change trend;

[0022] Based on the change trend of the temperature and humidity data, the temperature and humidity data with abnormal fluctuations are extracted from the temperature and humidity data set as key environmental features, and the key environmental features include the temperature and humidity data with abnormal fluctuations and their corresponding time information.

[0023] In one embodiment, the environmental characteristics are used as inputs of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and a control vector is generated based on the temperature and humidity change trend, including:

[0024] The deep learning model is trained based on the real-time temperature and humidity data and the historical temperature and humidity data after assimilation processing, so as to call the deep learning model to learn the change trend of the temperature and humidity data over time, and obtain a temperature and humidity prediction model for predicting the temperature and humidity change trend in the first time period in the future;

[0025] The change trend of the current temperature and humidity data relative to the historical temperature and humidity data in the temperature and humidity data set is used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the current temperature and humidity data in the first future time period.

[0026] In one embodiment, the environmental characteristics are used as inputs of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and a control vector is generated based on the temperature and humidity change trend, and further includes:

[0027] Obtaining an environmental feature vector identified and extracted by the parameter estimation and data learning algorithm at the second moment, and processing the environmental feature vector through the temperature and humidity prediction model to obtain a model output result;

[0028] The cost function is called to calculate the output result of the model to evaluate the regulation effect of different control vectors and generate the optimal control vector.

[0029] In one embodiment, the method further comprises:

[0030] Monitor the change trend of the temperature and humidity data in real time, and determine whether the value change between the current temperature and humidity data and the most recent historical temperature and humidity data exceeds the set abnormal threshold range; if so,

[0031] A response instruction is acquired, and an alarm signal is issued in response to the response instruction.

[0032] The present invention also provides a distributed temperature and humidity control system based on deep learning, the system comprising:

[0033] A data processing module is used to obtain temperature and humidity data measured by wireless sensors, and assimilate the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set;

[0034] A feature extraction module, used to call parameter estimation and data learning algorithms to analyze the temperature and humidity data after assimilation processing, so as to identify and extract environmental features in the temperature and humidity data;

[0035] An environmental prediction module, used to use the environmental characteristics as input of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and to generate a control vector based on the temperature and humidity change trend;

[0036] A temperature and humidity control module, configured to control the operating state of the HVAC equipment according to the temperature and humidity change trend in response to the control vector, so as to automatically control the temperature and humidity of the environment in the current space;

[0037] Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and time information corresponding to each group of temperature and humidity data, the wireless sensor includes multiple temperature sensors and humidity sensors, the environmental characteristics include abnormal fluctuations in temperature and humidity data, and when the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

[0038] The present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the distributed temperature and humidity control method based on deep learning as described in any one of the above.

[0039] The present invention also provides a computer storage medium storing a computer program, which, when executed by a processor, implements any of the above-described distributed temperature and humidity control methods based on deep learning.

[0040] The above-mentioned distributed temperature and humidity control method and system based on deep learning obtains the temperature and humidity data measured by the wireless sensor, and assimilates the temperature and humidity data through the deep learning set smoother to construct a temperature and humidity data set. Then, the parameter estimation and data learning algorithm are called to analyze the assimilated temperature and humidity data to identify and extract the environmental characteristics of the temperature and humidity data, including the abnormal fluctuations of the temperature and humidity data. Subsequently, the extracted environmental characteristics are used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the future period of time, and a control vector is generated based on the temperature and humidity change trend. Finally, in response to the generated control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically control the temperature and humidity of the environment in the current space. This method integrates a deep learning model and data assimilation processing. On the basis of improving the quality of temperature and humidity data, the deep learning model is continuously trained through the historical accumulated temperature and humidity data, which further improves the prediction accuracy of the temperature and humidity prediction model, so that it can respond and control abnormal changes in temperature and humidity in advance, and improves the flexibility and intelligence of temperature and humidity control. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 One of the flow charts of the distributed temperature and humidity control method based on deep learning provided by the present invention;

[0043] Figure 2 A schematic diagram of the overall process of temperature and humidity control of a distributed temperature and humidity control method based on deep learning in a specific embodiment provided by the present invention;

[0044] Figure 3 The second flow chart of the distributed temperature and humidity control method based on deep learning provided by the present invention;

[0045] Figure 4 The third flow chart of the distributed temperature and humidity control method based on deep learning provided by the present invention;

[0046] Figure 5 The fourth flowchart of the distributed temperature and humidity control method based on deep learning provided by the present invention;

[0047] Figure 6 The fifth flow chart of the distributed temperature and humidity control method based on deep learning provided by the present invention;

[0048] Figure 7 The sixth flow chart of the distributed temperature and humidity control method based on deep learning provided by the present invention;

[0049] Figure 8 Flow chart of the distributed temperature and humidity control method based on deep learning provided by the present invention No. 7;

[0050] Fig. 9 A schematic diagram of the structure of a distributed temperature and humidity control system based on deep learning provided by the present invention;

[0051] Fig.10 This is a diagram of the internal structure of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Combine the following Figures 1 to 10 The present invention describes a distributed temperature and humidity control method and system based on deep learning.

[0054] like Figure 1As shown, in one embodiment, a distributed temperature and humidity control method based on deep learning includes the following steps:

[0055] Step S110, obtaining the temperature and humidity data measured by the wireless sensor, and assimilating the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set.

[0056] The temperature and humidity data set includes multiple sets of temperature and humidity data and time information corresponding to each set of temperature and humidity data, and the wireless sensor includes multiple temperature sensors and humidity sensors.

[0057] Specifically, the temperature and humidity control system obtains temperature and humidity data measured by multiple groups of temperature sensors and humidity sensors and their corresponding reading times, and assimilates the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity dataset.

[0058] Combination Figure 2 As shown, in a specific embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention first collects the temperature and humidity data of the environment through a wireless sensor network, and uses data assimilation technology (ESDL) to process these temperature and humidity data to reduce errors and improve data quality.

[0059] The expression for collecting temperature and humidity data is: S_i(t)=[T_i(t),H_i(t)].

[0060] In the formula, T is temperature, H is humidity, t is time, and S_i(t) represents the data of the i-th sensor at time t, including temperature and humidity. The temperature and humidity control system is equipped with multiple wireless sensor network (WSN) nodes, which are distributed in the monitoring area to collect temperature and humidity data in real time. The data collected by each sensor node is represented as S_i(t), where T_i(t) and H_i(t) represent the temperature and humidity readings of the i-th sensor at time t, respectively.

[0061] In this embodiment, data assimilation is a mathematical method used to combine observational data with mathematical models to improve the accuracy of predictions. The ESDL (Ensemble Smoother with Deep Learning) algorithm is an advanced data assimilation method designed specifically for processing geoscience data with nonlinear and non-Gaussian characteristics.

[0062] The expression for temperature and humidity data assimilation processing is: X_t^a=X_t^f+K_t*(Y_t-H(X_t^f)).

[0063] Where X_t^a represents the state estimate after data assimilation at time t, X_t^f represents the prior state estimate of the prediction model at time t, K_t represents the Kalman gain at time t, which is used to adjust the deviation between the prediction and the actual observation, Y_t is the actual observation data, and H is the observation matrix, which is used to map the state space to the observation space.

[0064] It should be noted that the ESDL algorithm first predicts the state X_t^f based on the historical temperature and humidity data set and the current environmental variables. When the newly acquired actual observation data Y_t is available, the ESDL algorithm calculates the deviation between the predicted value and the actual observation value, and adjusts the state estimate through the Kalman gain parameter K_t to generate a more accurate posterior state X_t^a.

[0065] In this embodiment, data collection and assimilation are key technical links to ensure the accuracy and responsiveness of the temperature and humidity control system. The process uses wireless sensor networks (WSNs) to widely collect temperature (T) and humidity (H) data, which are continuously updated over time (t). The data collected by each sensor at each time point is represented by S_i(t) = [T_i(t), H_i(t)]. Among them, T_i(t) and H_i(t) represent the temperature and humidity readings of the i-th sensor at time t, respectively. The ensemble smoother deep learning method (ESDL) is used in the assimilation process, which is designed to process nonlinear and non-Gaussian data. The core of the algorithm is to combine the observed data with the numerical prediction model to optimize the data quality and reduce the error by updating the state estimate.

[0066] In the ESDL algorithm, the formula X_t^a=X_t^f+K_t*(Y_t-H(X_t^f)) describes the basic process of assimilation. X_t^f is the state predicted based on historical data, Y_t is the real-time observation data, H is the observation matrix that converts the state space into the observation space, and K_t is the Kalman gain, which is used to adjust the deviation between the prediction and the actual observation. Through this assimilation algorithm, the temperature and humidity control system can correct sensor errors in real time and enhance the accuracy and reliability of the data. This is crucial for dynamically adjusting the HVAC system, especially in places where temperature and humidity need to be precisely controlled to maintain environmental stability and energy-saving operation, such as hospitals, laboratories, and data centers. The assimilated data can more accurately reflect the real environment, provide solid data support for system control, and ensure the efficiency and accuracy of environmental control.

[0067] Step S120, calling parameter estimation and data learning algorithms to analyze the assimilated temperature and humidity data to identify and extract environmental features in the temperature and humidity data.

[0068] Among them, the environmental characteristics include abnormal fluctuations in temperature and humidity data. When the variation range of the temperature and humidity data exceeds a set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations.

[0069] Specifically, the temperature and humidity control system calls parameter estimation and data learning algorithms to analyze the assimilated temperature and humidity data to identify and extract environmental characteristics in the temperature and humidity data, including abnormal fluctuations in the temperature and humidity data.

[0070] Combination Figure 2 As shown, in a specific embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention, after assimilating the collected temperature and humidity data, uses a parameter estimation and data learning (PEDL, Parameter Estimation and Data Learning algorithm) algorithm to perform feature recognition on the assimilated data, and accurately captures key changes in the environment such as abnormal fluctuations.

[0071] The expression for feature recognition is: Z_t=f(S_t,Θ).

[0072] In the formula, f(·) represents a deep learning model for extracting and learning data features, Z_t represents the environmental feature vector at time t, S_t represents the sensor data set after assimilation data processing, and Θ is the parameter set of the deep learning model, including weights and biases. The PEDL algorithm uses a deep learning model f(·), parameterized as Θ, to extract the environmental feature Z_t at time t from the data set S_t after data assimilation. This process involves the model learning a large amount of historical and real-time data to identify key indicators such as abnormal fluctuations in temperature and humidity.

[0073] In this embodiment, the application of the PEDL algorithm (Parameter Estimation and Data Learning algorithm) in feature recognition mainly focuses on identifying and predicting key features in environmental data, such as abnormal changes in temperature and humidity, through deep learning technology. The algorithm optimizes the processing of data, can extract useful information from a large number of sensor inputs, and convert this useful information into actionable insights, thereby guiding the system to respond. In the deep learning framework, PEDL uses a multi-layer neural network to process and analyze time series data, and each layer of the network is designed to extract more and more abstract data features. For example, the first layer may only recognize simple temperature rises and falls, while deeper layers can recognize complex patterns such as periodic fluctuations or abnormal jumps. Once these features are identified, they can be used to predict potential failures of the temperature and humidity control system or adjust the operation of the system to adapt to the detected environmental changes. The PEDL algorithm can not only improve the existing prediction model, but also continuously optimize its model parameters (Θ) by learning historical data, so that the model can more accurately reflect the actual environmental conditions. This self-optimization capability is the key feature that distinguishes PEDL from traditional models, making the algorithm show higher flexibility and accuracy in dynamic and changing environments. Applied to actual temperature and humidity control systems, the PEDL algorithm can greatly improve the system's response speed and accuracy to environmental changes. For example, in the HVAC system of a data center, the PEDL algorithm can dynamically adjust the cooling load according to the predicted temperature trend, effectively reducing energy consumption and avoiding over-cooling or heating, thereby achieving optimal energy utilization and improving the overall reliability of the system.

[0074] In practical applications, the PEDL algorithm can be effectively deployed in large-scale industrial and residential environments to monitor and control temperature and humidity conditions. For example, by identifying abnormal fluctuations in the data, the PEDL algorithm can predict possible system failures or the impact of external environmental factors, so as to take measures in advance, such as adjusting the output of the HVAC system or sending maintenance alerts. The core of the PEDL algorithm lies in its ability to process large non-Gaussian and nonlinear data sets, and use advanced neural network architectures such as convolutional neural networks (CNN) or recurrent neural networks (RNN) to learn deep features of time series data. This algorithm not only improves the accuracy of feature extraction, but also enhances the system's adaptability and predictive ability to environmental changes. Applying the PEDL algorithm in the cooling system of the data center can predict future temperature trends based on real-time and historical data, optimize cooling load distribution, significantly reduce energy consumption and improve system response speed. Through this algorithm, the intelligence level of the temperature and humidity control system can be significantly improved, making it more accurately respond to environmental changes, while reducing energy consumption and operation and maintenance costs, and improving the overall efficiency and reliability of temperature and humidity control.

[0075] Step S130, using the environmental characteristics as input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics within a first future time period, and generating a control vector based on the temperature and humidity change trend.

[0076] Among them, the temperature and humidity prediction model is obtained by training the deep learning model based on the real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

[0077] Specifically, the temperature and humidity control system uses the extracted environmental characteristics as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the future (i.e., in the first time period), and generates a control vector based on the temperature and humidity change trend.

[0078] Combination Figure 2 As shown, in a specific embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention extracts features, and the temperature and humidity control system predicts the future environmental state based on the identified features through a prediction model, and automatically generates control instructions to adjust the operation of HVAC equipment, optimize energy consumption and maintain environmental comfort. This process uses a deep learning model to analyze assimilated real-time data and historical data, predict environmental changes in the future, and automatically adjust the operating state of the HVAC system accordingly, thereby optimizing energy consumption and maintaining the required environmental conditions.

[0079] The expression for control vector generation is: U_t=argmin_u C(P_t(Z_t,u)).

[0080] In the formula, P_t represents the prediction model at time t, which is responsible for generating predictions of future states, U_t is the control vector, which is determined by the prediction model and is used to guide the optimal operation of the HVAC system, Z_t represents the environmental feature vector identified by the feature recognition algorithm (such as PEDL) at time t, and C is the cost function, which is used to evaluate the effects of different control vectors, with the goal of minimizing energy consumption and other operating costs. The prediction model P_t is used to predict the future state based on the environmental feature Z_t, and the control vector U_t that minimizes the cost function C is found through optimization problems. The cost function usually includes factors such as energy consumption, equipment wear, and operating costs.

[0081] In this embodiment, the deep learning model is applied to analyze and predict future changes in temperature and humidity, which is crucial because it is directly related to the response efficiency and energy consumption optimization of the entire control system. The prediction model P_t generates a prediction of the future state based on the environmental features Z_t obtained from the feature recognition module at each time point t, including but not limited to the current temperature, humidity, historical trends, seasonal changes, and expected external environmental impacts. The calculation formula of the control vector U_t U_t = argmin_u C (P_t (Z_t, u)) indicates that the system selects a control strategy that minimizes the cost function C through an optimization process under the guidance of the prediction model. The choice of control strategy depends on the accuracy of the prediction model, which can learn environmental changes and adaptively adjust its parameters to match the actual situation. For example, in a large commercial building, the system can predict that the external temperature will rise in the afternoon and lower the internal temperature setting in advance to avoid excessive operation of the air conditioning system during high temperature periods, thereby saving energy and reducing equipment burden. In addition, if the prediction model detects that the humidity will rise at night, the system may adjust the operation plan of the dehumidification equipment to maintain comfort and prevent potential problems caused by excessive humidity. The implementation of this prediction and control mechanism not only enhances the economic efficiency of the system, but also improves the user's comfort experience, ensuring the sustainability and efficiency of facility management through real-time and accurate environmental management.

[0082] In practical applications, such as in large office buildings or industrial facilities, temperature and humidity settings can be adjusted in real time to respond to external environmental changes (such as weather changes) and internal load changes (such as equipment operation or personnel flow). Through the predictive capabilities of deep learning models, the system can not only respond to current temperature and humidity changes in a timely manner, but also predict trends in the next few hours or even days, and implement predictive control strategies. Deep learning models are trained to identify complex patterns and trends in temperature and humidity data, such as periodic changes, the impact of emergencies, and changes in long-term trends, enabling the control system to make optimal decisions in different operating scenarios, by adjusting the operating frequency and temperature settings of the air conditioning system to adapt to the upcoming high temperature weather, or slowing down the equipment to reduce energy consumption before the peak energy demand period.

[0083] Step S140, in response to the control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically adjust the temperature and humidity of the environment in the current space.

[0084] Specifically, the temperature and humidity control system responds to the control vector generated in step S130 and controls the operating state of the HVAC equipment according to the temperature and humidity change trend to automatically control the temperature and humidity of the environment in the current space.

[0085] The above-mentioned distributed temperature and humidity control method based on deep learning obtains the temperature and humidity data measured by the wireless sensor, and assimilates the temperature and humidity data through the deep learning set smoother to construct a temperature and humidity data set. Then, the parameter estimation and data learning algorithm are called to analyze the assimilated temperature and humidity data to identify and extract the environmental characteristics of the temperature and humidity data, including the abnormal fluctuations of the temperature and humidity data. Subsequently, the extracted environmental characteristics are used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the future period of time, and a control vector is generated based on the temperature and humidity change trend. Finally, in response to the generated control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically control the temperature and humidity of the environment in the current space. This method integrates a deep learning model and data assimilation processing. On the basis of improving the quality of temperature and humidity data, the deep learning model is continuously trained through the historical accumulated temperature and humidity data, which further improves the prediction accuracy of the temperature and humidity prediction model, so that it can respond and control abnormal changes in temperature and humidity in advance, and improves the flexibility and intelligence of temperature and humidity control.

[0086] like Figure 3 As shown, in one embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention obtains temperature and humidity data measured by wireless sensors, and assimilates the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set, specifically including the following steps:

[0087] Step S112, real-time temperature and humidity data measured by wireless sensors in the current space and time information corresponding to each set of real-time temperature and humidity data are acquired in real time, and the real-time temperature and humidity data are assimilated through a deep learning set smoother.

[0088] Step S114, based on each set of historical temperature and humidity data after assimilation and its corresponding time information, combined with the real-time temperature and humidity data after assimilation, the temperature and humidity data set is updated.

[0089] Each group of temperature sensors and humidity sensors corresponds to a sensor node, each sensor node corresponds to a group of temperature and humidity data in the temperature and humidity data set, and each group of temperature and humidity data includes a group of temperature sensor readings and humidity sensor readings and reading time.

[0090] like Figure 4 As shown, in one embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention obtains temperature and humidity data measured by wireless sensors, and assimilates the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set, and specifically includes the following steps:

[0091] Step S116, based on the temperature and humidity data set, obtain the environmental prediction state at the first moment output by the temperature and humidity prediction model according to the historical temperature and humidity data, and determine the Kalman benefit parameter and the actual temperature and humidity data at the first moment.

[0092] Step S118, mapping the environmental prediction state from the state space to the observation space through the observation matrix, and calling the deep learning ensemble smoother in combination with the Kalman benefit parameter and the actual temperature and humidity data to calculate the environmental state of the temperature and humidity data after data assimilation processing at the first moment.

[0093] The first moment is any determined time point used for calculation.

[0094] like Figure 5 As shown, in one embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention calls parameter estimation and data learning algorithms to analyze the assimilated temperature and humidity data to identify and extract environmental features in the temperature and humidity data, specifically including the following steps:

[0095] In step S122, the assimilated temperature and humidity data set is used as the input of the parameter estimation and data learning algorithm, so as to analyze the temperature and humidity data set by calling the deep learning model in combination with the model parameter set through the parameter estimation and data learning algorithm to obtain the change trend of the temperature and humidity data.

[0096] Step S124, based on the temperature and humidity data change trend, extract the temperature and humidity data with abnormal fluctuations as key environmental features from the temperature and humidity data set, where the key environmental features include the temperature and humidity data with abnormal fluctuations and their corresponding time information.

[0097] like Figure 6 As shown, in one embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention takes environmental characteristics as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the first time period in the future, and generates a control vector based on the temperature and humidity change trend, specifically including the following steps:

[0098] Step S132, training the deep learning model based on the real-time temperature and humidity data and the historical temperature and humidity data after assimilation processing, so as to call the deep learning model to learn the changing trend of the temperature and humidity data over time, and obtain a temperature and humidity prediction model for predicting the changing trend of temperature and humidity in the first time period in the future.

[0099] Step S134, taking the change trend of the current temperature and humidity data relative to the historical temperature and humidity data in the temperature and humidity data set as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the current temperature and humidity data in the first future time period.

[0100] like Figure 7As shown, in one embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention takes environmental characteristics as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the first time period in the future, and generates a control vector based on the temperature and humidity change trend, and specifically also includes the following steps:

[0101] Step S136, obtaining the environmental feature vector identified and extracted by the parameter estimation and data learning algorithm at the second moment, and processing the environmental feature vector through the temperature and humidity prediction model to obtain a model output result.

[0102] The second moment is any determined time point.

[0103] Step S138, calling the cost function to calculate the model output result to evaluate the regulation effect of different control vectors and generate the optimal control vector.

[0104] like Figure 8 As shown, in one embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention further includes the following steps:

[0105] Step S810, monitor the change trend of the temperature and humidity data in real time, and determine whether the value change between the current temperature and humidity data and the most recent historical temperature and humidity data exceeds the set abnormal threshold range.

[0106] Step S820, obtaining a response instruction, and issuing an alarm signal in response to the response instruction.

[0107] Specifically, when the judgment result in step S810 is that the numerical change between the current temperature and humidity data and the most recent historical temperature and humidity data exceeds the set abnormal threshold range, a response instruction is obtained, and an alarm signal is sent to the user in response to the response instruction.

[0108] Combination Figure 2 As shown, in a specific embodiment, the distributed temperature and humidity control method based on deep learning provided by the present invention can continuously monitor its performance and environmental conditions through a real-time monitoring and feedback mechanism, and quickly adjust the control strategy according to real-time data to ensure the efficient operation of the temperature and humidity control system and the stability of the environment. The real-time monitoring and feedback mechanism can ensure that the temperature and humidity control system can respond to environmental changes and potential equipment failures in real time. This process can detect and respond to any abnormal situation in a timely manner by continuously monitoring and evaluating the real-time data of the system.

[0109] The expression of real-time monitoring and feedback mechanism is: R_t=A_t(X_t^a,ΔX).

[0110] In the formula, A_t is the real-time monitoring algorithm at time t, R_t is the response instruction, which is an operation or alarm signal generated based on the monitoring results, X_t^a represents the assimilated state estimate at the current time t, reflecting the latest system state, and ΔX is the predetermined abnormal threshold, which is used to determine when the state change needs to trigger a response. The monitoring algorithm A_t compares the current assimilated state X_t^a with the set abnormal threshold ΔX to determine whether the temperature and humidity control system is in normal operation. If X_t^a deviates from ΔX beyond the predetermined range, a response instruction R_t is generated, including adjusting control parameters, sending alarms or other corrective measures.

[0111] In this embodiment, the application of real-time monitoring and feedback mechanism in the temperature and humidity control system is the key to ensure the efficient and reliable operation of the system. This mechanism realizes real-time evaluation of system performance and rapid response to abnormal situations by continuously monitoring the environmental state and comparing it with the set threshold. The core function of the real-time monitoring algorithm A_t is to analyze the assimilated state data X_t^a in real time and compare it with the preset abnormal threshold ΔX. The comparison process not only monitors whether the data point exceeds the normal range, but also analyzes the change trend of the data to predict possible problems in the future. For example, if the data at several consecutive time points show that the temperature continues to rise and exceeds the threshold ΔX, the algorithm will determine it as an abnormal state and generate a response instruction R_t. The formula R_t=A_t(X_t^a,ΔX) describes the generation process of the response instruction. When the current state represented by X_t^a deviates significantly from the normal state (i.e., exceeds ΔX), the monitoring system will automatically start early warning or adjustment measures, such as adjusting the temperature and humidity settings or sending maintenance notifications. In a data center or precision manufacturing environment, the real-time monitoring system can respond quickly to environmental changes to prevent equipment failures or production quality problems caused by abnormal temperature and humidity. Through continuous data collection and analysis, the system can timely adjust the operation of the air conditioning system, optimize energy use, and ensure that environmental conditions meet precise production requirements. In addition, the real-time monitoring and feedback mechanism can also be integrated with the maintenance scheduling system to ensure that any situation requiring manual intervention can be handled in a timely manner, which not only improves operational efficiency, but also greatly reduces the risks and costs caused by equipment failure or unsuitable environmental conditions.

[0112] In practical applications, such as in a complex industrial manufacturing environment, a real-time monitoring system can continuously detect the output of temperature and humidity sensors. By comparing with historical data and model predictions, the system can instantly identify any changes outside the normal range, such as abnormal temperature rise or sudden drop in ambient humidity caused by equipment failure. The real-time monitoring mechanism continuously learns and adapts to small changes in normal operating parameters through advanced data processing and machine learning algorithms. When the detected data exceeds the normal operating range, the system will immediately analyze whether it is a temporary fluctuation or a potential system failure, and then take appropriate response measures, such as adjusting the output of the HVAC system, or in more serious cases, notifying maintenance personnel to check and repair. The real-time monitoring and feedback mechanism not only enhances the system's self-healing ability, but also greatly improves operational efficiency and safety. In addition, the mechanism ensures that environmental conditions are always accurately controlled, while also significantly reducing energy waste caused by equipment failure or operational errors.

[0113] The distributed temperature and humidity control system based on deep learning provided by the present invention is described below. The distributed temperature and humidity control system based on deep learning described below and the distributed temperature and humidity control method based on deep learning described above can be referenced to each other.

[0114] like Fig. 9 As shown, in one embodiment, a distributed temperature and humidity control system based on deep learning includes a data processing module 910, a feature extraction module 920, an environment prediction module 930 and a temperature and humidity control module 940.

[0115] The data processing module 910 is used to obtain the temperature and humidity data measured by the wireless sensor, and assimilate the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set.

[0116] The feature extraction module 920 is used to call parameter estimation and data learning algorithms to analyze the temperature and humidity data after assimilation processing, so as to identify and extract environmental features in the temperature and humidity data.

[0117] The environmental prediction module 930 is used to use environmental characteristics as input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the first future time period, and generate a control vector based on the temperature and humidity change trend.

[0118] The temperature and humidity control module 940 is used to control the operating state of the HVAC equipment according to the temperature and humidity change trend in response to the control vector, so as to automatically control the temperature and humidity of the environment in the current space.

[0119] Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and the time information corresponding to each group of temperature and humidity data. The wireless sensor includes multiple temperature sensors and humidity sensors. The environmental characteristics include abnormal fluctuations in temperature and humidity data. When the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on the real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

[0120] In this embodiment, in the distributed temperature and humidity control system based on deep learning provided by the present invention, the data processing module 910 is specifically used for:

[0121] The real-time temperature and humidity data measured by wireless sensors in the current space and the time information corresponding to each set of real-time temperature and humidity data are obtained in real time, and the real-time temperature and humidity data are assimilated through a deep learning set smoother.

[0122] Based on each set of historical temperature and humidity data after assimilation and its corresponding time information, the temperature and humidity dataset is updated in combination with the real-time temperature and humidity data after assimilation.

[0123] Each group of temperature sensors and humidity sensors corresponds to a sensor node, each sensor node corresponds to a group of temperature and humidity data in the temperature and humidity data set, and each group of temperature and humidity data includes a group of temperature sensor readings and humidity sensor readings and reading time.

[0124] In this embodiment, in the distributed temperature and humidity control system based on deep learning provided by the present invention, the data processing module 910 is specifically used for:

[0125] Based on the temperature and humidity data set, the temperature and humidity prediction model is used to obtain the environmental prediction state at the first moment according to the historical temperature and humidity data output, and the Kalman benefit parameters and actual temperature and humidity data at the first moment are determined.

[0126] The environmental prediction state is mapped from the state space to the observation space through the observation matrix, and the deep learning ensemble smoother is called in combination with the Kalman benefit parameter and the actual temperature and humidity data to calculate the environmental state of the temperature and humidity data after data assimilation at the first moment.

[0127] In this embodiment, in the distributed temperature and humidity control system based on deep learning provided by the present invention, the feature extraction module 920 is specifically used for:

[0128] The assimilated temperature and humidity dataset is used as the input of the parameter estimation and data learning algorithm. The deep learning model is called by the parameter estimation and data learning algorithm in combination with the model parameter set to analyze the temperature and humidity dataset and obtain the changing trend of the temperature and humidity data.

[0129] Based on the changing trend of temperature and humidity data, temperature and humidity data with abnormal fluctuations are extracted from the temperature and humidity data set as key environmental features. The key environmental features include temperature and humidity data with abnormal fluctuations and their corresponding time information.

[0130] In this embodiment, in the distributed temperature and humidity control system based on deep learning provided by the present invention, the environment prediction module 930 is specifically used for:

[0131] The deep learning model is trained based on the real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, so as to call the deep learning model to learn the changing trend of temperature and humidity data over time, and obtain a temperature and humidity prediction model for predicting the changing trend of temperature and humidity in the first time period in the future.

[0132] The change trend of the current temperature and humidity data relative to the historical temperature and humidity data in the temperature and humidity data set is used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the current temperature and humidity data in the first time period in the future.

[0133] In this embodiment, in the distributed temperature and humidity control system based on deep learning provided by the present invention, the environment prediction module 930 is specifically used for:

[0134] The environmental feature vector identified and extracted by the parameter estimation and data learning algorithm at the second moment is obtained, and the environmental feature vector is processed by the temperature and humidity prediction model to obtain the model output result.

[0135] The cost function is called to calculate the model output results to evaluate the regulatory effects of different control vectors and generate the optimal control vector.

[0136] In this embodiment, the distributed temperature and humidity control system based on deep learning provided by the present invention further includes an abnormal alarm module for:

[0137] Monitor the changing trend of temperature and humidity data in real time, and determine whether the value change between the current temperature and humidity data and the most recent historical temperature and humidity data exceeds the set abnormal threshold range. If so,

[0138] A response instruction is obtained, and an alarm signal is issued in response to the response instruction.

[0139] Fig.10 The physical structure diagram of an electronic device is shown as an example. The electronic device may be a smart terminal, and its internal structure diagram may be as follows: Fig.10As shown. The electronic device includes a processor, an internal memory and a network interface connected via a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a distributed temperature and humidity control method based on deep learning is implemented, and the method includes:

[0140] Acquire the temperature and humidity data measured by wireless sensors, and assimilate the temperature and humidity data through a deep learning ensemble smoother to construct a temperature and humidity dataset;

[0141] Call parameter estimation and data learning algorithms to analyze the assimilated temperature and humidity data to identify and extract environmental features in the temperature and humidity data;

[0142] The environmental characteristics are used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the first time period in the future, and a control vector is generated based on the temperature and humidity change trend;

[0143] In response to the control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically adjust the temperature and humidity of the environment in the current space;

[0144] Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and the time information corresponding to each group of temperature and humidity data. The wireless sensor includes multiple temperature sensors and humidity sensors. The environmental characteristics include abnormal fluctuations in temperature and humidity data. When the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on the real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

[0145] Those skilled in the art will understand that Fig.10 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present invention, and does not constitute a limitation on the electronic device to which the scheme of the present invention is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0146] On the other hand, the present invention further provides a computer storage medium storing a computer program, which, when executed by a processor, implements a distributed temperature and humidity control method based on deep learning, the method comprising:

[0147] Acquire the temperature and humidity data measured by wireless sensors, and assimilate the temperature and humidity data through a deep learning ensemble smoother to construct a temperature and humidity dataset;

[0148] Call parameter estimation and data learning algorithms to analyze the assimilated temperature and humidity data to identify and extract environmental features in the temperature and humidity data;

[0149] The environmental characteristics are used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the first time period in the future, and a control vector is generated based on the temperature and humidity change trend;

[0150] In response to the control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically adjust the temperature and humidity of the environment in the current space;

[0151] Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and the time information corresponding to each group of temperature and humidity data. The wireless sensor includes multiple temperature sensors and humidity sensors. The environmental characteristics include abnormal fluctuations in temperature and humidity data. When the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on the real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

[0152] In another aspect, a computer program product or a computer program is provided, the computer program product or the computer program comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and when the processor executes the computer instructions, a distributed temperature and humidity control method based on deep learning is implemented, the method comprising:

[0153] Acquire the temperature and humidity data measured by wireless sensors, and assimilate the temperature and humidity data through a deep learning ensemble smoother to construct a temperature and humidity dataset;

[0154] Call parameter estimation and data learning algorithms to analyze the assimilated temperature and humidity data to identify and extract environmental features in the temperature and humidity data;

[0155] The environmental characteristics are used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the environmental characteristics in the first time period in the future, and a control vector is generated based on the temperature and humidity change trend;

[0156] In response to the control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically adjust the temperature and humidity of the environment in the current space;

[0157] Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and the time information corresponding to each group of temperature and humidity data. The wireless sensor includes multiple temperature sensors and humidity sensors. The environmental characteristics include abnormal fluctuations in temperature and humidity data. When the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on the real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

[0158] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0159] By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0160] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0161] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A distributed temperature and humidity control method based on deep learning, characterized in that: The method comprises: Acquire temperature and humidity data measured by wireless sensors, and assimilate the temperature and humidity data through a deep learning ensemble smoother to construct a temperature and humidity data set; Calling parameter estimation and data learning algorithms to analyze the temperature and humidity data after assimilation processing to identify and extract environmental features in the temperature and humidity data; The environmental characteristics are used as inputs of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and a control vector is generated based on the temperature and humidity change trend; In response to the control vector, the operating state of the HVAC equipment is controlled according to the temperature and humidity change trend to automatically adjust the temperature and humidity of the environment in the current space; Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and time information corresponding to each group of temperature and humidity data, the wireless sensor includes multiple temperature sensors and humidity sensors, the environmental characteristics include abnormal fluctuations in temperature and humidity data, and when the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

2. The distributed temperature and humidity control method based on deep learning according to claim 1 is characterized in that: The method of acquiring temperature and humidity data measured by a wireless sensor and assimilating the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set includes: Real-time acquisition of real-time temperature and humidity data measured by wireless sensors in the current space and the time information corresponding to each set of real-time temperature and humidity data, and assimilation of the real-time temperature and humidity data through a deep learning set smoother; Based on each set of historical temperature and humidity data after assimilation and its corresponding time information, the temperature and humidity data set is updated in combination with the real-time temperature and humidity data after assimilation; Each group of the temperature sensor and humidity sensor corresponds to a sensor node, each sensor node corresponds to a group of temperature and humidity data in the temperature and humidity data set, and each group of temperature and humidity data includes a group of temperature sensor readings and humidity sensor readings and reading time.

3. The distributed temperature and humidity control method based on deep learning according to claim 2 is characterized in that: The method further includes: obtaining temperature and humidity data measured by wireless sensors, and assimilating the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set; Based on the temperature and humidity data set, the temperature and humidity prediction model is used to obtain the environmental prediction state at the first moment output according to the historical temperature and humidity data, and the Kalman benefit parameter and the actual temperature and humidity data at the first moment are determined; The environmental prediction state is mapped from the state space to the observation space through the observation matrix, and the deep learning ensemble smoother is called in combination with the Kalman benefit parameter and the actual temperature and humidity data to calculate the environmental state of the temperature and humidity data after data assimilation processing at the first moment.

4. The distributed temperature and humidity control method based on deep learning according to claim 3 is characterized in that: The calling of the parameter estimation and data learning algorithm to analyze the temperature and humidity data after assimilation processing to identify and extract environmental features in the temperature and humidity data includes: The assimilated temperature and humidity data set is used as the input of the parameter estimation and data learning algorithm, so as to analyze the temperature and humidity data set by calling the deep learning model in combination with the model parameter set through the parameter estimation and data learning algorithm to obtain the temperature and humidity data change trend; Based on the change trend of the temperature and humidity data, the temperature and humidity data with abnormal fluctuations are extracted from the temperature and humidity data set as key environmental features, and the key environmental features include the temperature and humidity data with abnormal fluctuations and their corresponding time information.

5. The distributed temperature and humidity control method based on deep learning according to claim 4 is characterized in that: The step of using the environmental characteristics as an input of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and generating a control vector based on the temperature and humidity change trend, includes: The deep learning model is trained based on the real-time temperature and humidity data and the historical temperature and humidity data after assimilation processing, so as to call the deep learning model to learn the change trend of the temperature and humidity data over time, and obtain a temperature and humidity prediction model for predicting the temperature and humidity change trend in the first time period in the future; The change trend of the current temperature and humidity data relative to the historical temperature and humidity data in the temperature and humidity data set is used as the input of the temperature and humidity prediction model to output the temperature and humidity change trend of the current temperature and humidity data in the first future time period.

6. The distributed temperature and humidity control method based on deep learning according to claim 5 is characterized in that: The method of using the environmental characteristics as an input of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and generating a control vector based on the temperature and humidity change trend, further includes: Obtaining an environmental feature vector identified and extracted by the parameter estimation and data learning algorithm at the second moment, and processing the environmental feature vector through the temperature and humidity prediction model to obtain a model output result; The cost function is called to calculate the output result of the model to evaluate the regulation effect of different control vectors and generate the optimal control vector.

7. The distributed temperature and humidity control method based on deep learning according to any one of claims 1 to 6, characterized in that: The method further comprises: Monitor the change trend of the temperature and humidity data in real time, and determine whether the value change between the current temperature and humidity data and the most recent historical temperature and humidity data exceeds the set abnormal threshold range; if so, A response instruction is acquired, and an alarm signal is issued in response to the response instruction.

8. A distributed temperature and humidity control system based on deep learning, characterized in that: The system comprises: A data processing module is used to obtain temperature and humidity data measured by wireless sensors, and assimilate the temperature and humidity data through a deep learning set smoother to construct a temperature and humidity data set; A feature extraction module, used to call parameter estimation and data learning algorithms to analyze the temperature and humidity data after assimilation processing, so as to identify and extract environmental features in the temperature and humidity data; An environmental prediction module, used to use the environmental characteristics as input of a temperature and humidity prediction model to output a temperature and humidity change trend of the environmental characteristics within a first future time period, and to generate a control vector based on the temperature and humidity change trend; A temperature and humidity control module, configured to control the operating state of the HVAC equipment according to the temperature and humidity change trend in response to the control vector, so as to automatically control the temperature and humidity of the environment in the current space; Among them, the temperature and humidity data set includes multiple groups of temperature and humidity data and time information corresponding to each group of temperature and humidity data, the wireless sensor includes multiple temperature sensors and humidity sensors, the environmental characteristics include abnormal fluctuations in temperature and humidity data, and when the change range of the temperature and humidity data exceeds the set threshold range, it is determined that the temperature and humidity data have abnormal fluctuations. The temperature and humidity prediction model is a deep learning model trained based on real-time temperature and humidity data and historical temperature and humidity data after assimilation processing, and is used to predict the temperature and humidity change trend in the first time period after any environmental characteristics.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images