Split type air conditioner load management system and method thereof

Abstract

The invention belongs to the technical field of intelligent control of air conditioning systems, and discloses a split type air conditioner load management system and method. The system comprises a central controller, a health management module, an adaptive integrated control module AICA and an intelligent scene recognition module. The central controller integrates various sensor information such as temperature and humidity, air quality and user physiological data, data processing and intelligent decision making are carried out through an AICA module, and dynamic optimization management of air conditioner loads is achieved. The health management module automatically adjusts the air conditioner operation mode according to the body temperature, heart rate and other physiological indexes of the user, and a comfortable environment is provided. And the intelligent scene recognition module is used for pre-judging user requirements and adjusting the operation of the air conditioner by learning a user behavior mode. The energy efficiency and the intelligent level of air conditioner operation are improved, the method is suitable for air conditioner load management of various buildings, energy can be effectively saved, and the comfort level of users is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioning systems, and more particularly to a split-type air conditioning load management system and method. The system combines deep learning, intelligent control, and health management technologies to achieve dynamic optimization management of air conditioning loads, improving energy efficiency and user experience. Background Art

[0002] With the widespread use of air conditioning systems in modern buildings, especially in residences, office buildings, and commercial facilities, how to efficiently manage air conditioning loads has become a major challenge in the air conditioning industry. Existing air conditioning load management systems generally have the following deficiencies in practical applications:

[0003] Lack of dynamic adaptability to user behavior and environmental changes: Existing air conditioning systems are typically based on pre-set control strategies and can only adjust based on fixed parameters. They lack the ability to respond in real time to changes in user behavior and the external environment. This results in insufficient adaptability to varying environmental conditions, making it difficult to meet diverse user needs. For example, in existing technologies, when the external ambient temperature suddenly changes or user behavior changes, the air conditioning system cannot adjust its operating mode in a timely manner, resulting in increased energy consumption and decreased comfort.

[0004] Limited energy efficiency optimization methods: Traditional air conditioning load management systems rely primarily on simple sensor data and fixed logic for control, lacking the ability to deeply analyze real-time data and conduct dynamic optimization. This static approach to energy efficiency management is insufficient in complex and ever-changing environments, leading to energy waste and system inefficiency. Existing systems perform particularly poorly in extreme conditions such as high temperature and high humidity, failing to effectively balance user comfort and energy consumption.

[0005] Lack of health management and personalized services: As users' requirements for living environment comfort and health increase, existing air conditioning systems are gradually exposing their shortcomings in health management and personalized services. Most existing systems focus only on basic temperature and humidity control, failing to integrate users' physiological health data (such as heart rate and sleep quality) for intelligent adjustments and unable to provide optimized services tailored to individual needs. In addition, existing systems are also relatively lacking in providing personalized scene configurations, making it difficult to intelligently adjust according to users' usage habits and preferences.

[0006] Poor system scalability and compatibility: Current HVAC management systems often utilize a closed architecture, lacking scalability and compatibility, making them difficult to adapt to varying building sizes and diverse user needs. This design not only limits system functionality but also hinders seamless integration with other smart home devices or systems, increasing operational complexity for users across different smart devices.

[0007] Lack of proactive and intelligent maintenance management: Existing air conditioning system maintenance and management is mostly reactive, typically requiring repairs only after a fault occurs. These systems lack intelligent fault prediction and preventive maintenance capabilities. This reactive approach often leads to a decrease in overall system stability and a poor user experience. For example, when a system anomaly occurs, there's no early warning or effective preventive measures, leading to a greater impact or prolonged repair times. Summary of the Invention

[0008] In response to the above-mentioned deficiencies in the existing technology, the present invention proposes a split-type air-conditioning load management system and method thereof, which significantly improves the intelligence level and energy efficiency performance of the system by introducing an adaptive integrated control module (AICA) and a health management module.

[0009] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0010] A split-type air conditioning load management system includes multiple air conditioning units and a central controller, the system comprising:

[0011] Multiple temperature sensors, humidity sensors, energy consumption sensors, and operating status sensors installed in each air conditioning unit to monitor the temperature, humidity, energy consumption, and operating mode of the air conditioning unit in real time;

[0012] A central controller in communication with the sensor, the central controller comprising:

[0013] A data processing module receives the monitoring data and performs real-time processing and feature extraction on the data based on a convolutional neural network (CNN) and a long short-term memory (LSTM) network;

[0014] The adaptive integrated control module AICA, which combines the deep Q-network (DQN) and the Transformer model, analyzes the extracted features and generates optimized control instructions for dynamically adjusting the operating mode of each air conditioning unit;

[0015] The prediction module performs adaptive learning based on historical and real-time data to predict future load changes and adjust system operating parameters in advance;

[0016] The central controller also includes a health management module that monitors user health data and air quality in real time, automatically adjusts air conditioning operating parameters based on the monitoring results, and provides optimized air quality and temperature control.

[0017] Preferably, the central controller further comprises:

[0018] A local optimization module dynamically adjusts the air speed, cooling capacity, and operating time of a single air conditioning unit based on the unit's real-time temperature, humidity, and energy consumption data. This optimization process is supported by the Deep Q-Network (DQN) in the Adaptive Integrated Control Module (AICA).

[0019] The global optimization module, based on the overall working status of multiple air-conditioning units, combines big data analysis and machine learning algorithms. The AICA module generates personalized control strategies to optimize load distribution and reduce energy consumption.

[0020] Preferably, the central controller includes:

[0021] Intelligent self-healing module, which monitors key system parameters, automatically isolates faulty components and reconfigures system resources when a system failure is detected;

[0022] The fault prediction module uses the deep learning algorithm of the adaptive integrated control module AICA to identify possible faults in advance and generate preventive maintenance plans based on historical fault data and real-time monitoring data;

[0023] The preventive maintenance module, based on historical operating data and the prediction model generated by the AICA module, identifies possible faults in advance and prompts users to perform maintenance operations, notifying users in real time through user terminal devices to perform necessary preventive maintenance.

[0024] Preferably, the central controller is connected to a user terminal device via wireless communication, and the user terminal device includes:

[0025] The health management module automatically adjusts the operating mode of the air conditioning system by analyzing the user's health data and air quality and combining it with the optimization control strategy generated by the AICA module;

[0026] The intelligent scene recognition module automatically identifies the current usage scenario, including work mode, rest mode, and entertainment mode, by analyzing the user's daily behavior patterns, environmental changes, and historical system operation data, and adjusts the operating parameters of the air-conditioning system accordingly.

[0027] Preferably, the central controller includes:

[0028] The environmental change perception module is used to detect changes in the external environment in real time and dynamically adjust the control strategy. The AICA module processes the perception data in real time.

[0029] The user behavior prediction module collects and analyzes historical user operation data, combines time series analysis and random forest algorithms in the AICA module to predict users' future operation needs, adjusts system operating parameters based on the prediction results, and optimizes the air conditioning mode in advance;

[0030] The personalized scene configuration module allows users to customize scene configurations through user terminal devices. The system automatically adjusts the air conditioning operation mode based on these configurations, including temperature setting, fan speed adjustment and operation time selection.

[0031] Preferably, the central controller includes:

[0032] A big data analysis module, which processes massive amounts of data from multiple sensors and extracts valuable information through advanced analysis algorithms in the adaptive integrated control module (AICA);

[0033] The real-time energy consumption optimization module analyzes real-time energy consumption data through the AICA module, generates energy-saving strategies and dynamically adjusts the operating parameters of the air-conditioning system.

[0034] Preferably, the user terminal device includes:

[0035] The personalized service module generates personalized control options and suggestions, including optimal temperature settings and energy-saving mode selection, based on the user's historical usage data and preference analysis by the adaptive integrated control module AICA.

[0036] A split air conditioner load management method, the method comprising the following steps:

[0037] S1: Monitor the temperature, humidity, energy consumption and operation mode of the air conditioning unit in real time through multiple temperature, humidity, energy consumption and operation status sensors installed in each air conditioning unit;

[0038] S2: The central controller receives and integrates the monitoring data of the sensors and generates a unified environment description and optimized control instructions using the adaptive integrated control module AICA. The AICA combines contrastive learning, deep Q network DQN, Transformer model and stable entropy optimization strategy SEOP to make control decisions.

[0039] S3: The central controller dynamically adjusts the operating parameters of the air conditioning system based on the real-time monitoring data of individual air conditioning units and the overall working status through the local optimization module and the global optimization module;

[0040] S4: Based on the intelligent algorithm and user behavior prediction module in the AICA module, user needs are predicted in advance and corresponding control instructions are generated to achieve load balancing and optimized control;

[0041] S5: Through the health management module and personalized scene configuration module, it provides customized air conditioning control strategies for user health and scenes.

[0042] Preferably, the steps for implementing the adaptive integrated control module AICA include:

[0043] 1) Using multiple sensors to collect indoor and outdoor environmental parameters and the operating status data of air conditioning units, this data is integrated to form a unified environmental description. Contrastive learning is used to pre-train the multimodal data to obtain richer feature representations suitable for dynamic load management of air conditioning systems.

[0044] 2) Use the deep learning model in the AICA module to preprocess the collected time series data, including noise filtering and missing data completion, and standardize and normalize the preprocessed data;

[0045] 3) Based on the deep Q network (DQN), the air conditioning load demand under different states is calculated, and the difference between the current state and the ideal state is analyzed through the Transformer model. The Q value function is defined as follows:

[0046]

[0047] Among them, Q(s t ,a t ) is state s t Next take action a t The expected cumulative return value that can be obtained; t Represents the current environmental status, including indoor and outdoor temperature, humidity, and air conditioning operation status information; a t Represents the state s t The actions taken under the control of the user include temperature setting and wind speed adjustment; t represents the immediate reward obtained at time step t, which is related to user comfort and energy consumption; γ represents the discount factor (0≤γ≤1), which is used to balance current and future rewards; E represents the expectation operator, which is used to calculate the expected return of the action;

[0048] 4) The stable entropy optimization strategy (SEOP) is used for control decision-making. Combining the stability of the proximal strategy optimization (PPO) and the maximum entropy exploration characteristics of the soft actor-critic algorithm (SAC), a balance is achieved between stability and adaptability, and the optimal operation mode of the air-conditioning system is selected. The core loss function of SEOP includes:

[0049] The PPO loss function is:

[0050]

[0051] The SAC loss function is:

[0052] L SAC (θ)=E t [Qφ(s t ,a t )-αlogπ θ (a t |s t)]

[0053] Among them, θ represents the parameters of the policy network; r t (θ) represents the ratio of new and old strategies; represents the potential function, which is used to evaluate the quality of the current action relative to the average level; ∈ represents the clipping range to prevent the strategy from being updated too much; Qφ(s t ,a t ) represents the Q-value estimation based on SAC, which represents the expected return of action at in state St; α represents the entropy adjustment coefficient, which controls the degree of exploration; logπ θ (a t |s t ) represents the log probability of the policy network output;

[0054] 5) Generate control instructions based on the optimization decision and send them to the air conditioning unit to perform the corresponding operation;

[0055] 6) Monitor the response of the air conditioning system and dynamically optimize the model weights using the adaptive adjustment mechanism of SEOP;

[0056] 7) Continuously collect new data during system operation, use SEOP for online learning and real-time update of model parameters.

[0057] Preferably, the multi-source data fusion technology of the big data analysis module and the AICA module is used to combine historical energy consumption data and real-time operation data to predict future energy consumption trends, and generate energy-saving tips and operation suggestions based on the prediction results, which are pushed to the user terminal device to prompt the user to adjust the air conditioning usage mode during high energy consumption periods;

[0058] Based on the user behavior prediction module and personalized service module in the AICA module, it analyzes user preferences and historical comfort data, generates a personalized comfort curve, and automatically adjusts the operating parameters of the air conditioning system. At the same time, it displays the current comfort status and adjustment suggestions on the user terminal device;

[0059] When the system detects changes in external environmental conditions, including high temperature, high humidity, and air pollution, it automatically activates an environmentally friendly operating mode to reduce the energy consumption and environmental impact of the air-conditioning system, and prioritizes natural ventilation or lowers the operating load of the air-conditioning. At the same time, the energy-saving effect and environmental impact of the current mode are displayed on the user terminal device.

[0060] Compared with existing technologies, the present invention offers the following advantages: By introducing an adaptive integrated control module (AICA), combined with a deep Q-network (DQN) and a Transformer model, the system achieves real-time response to user behavior and environmental changes. The system automatically adjusts the air conditioner's operating parameters, ensuring that when the external environment or user needs change, the air conditioning system can promptly optimize its operating mode and enhance user comfort. This dynamic response capability significantly outperforms the static control strategies used in existing technologies and can better adapt to complex and changing real-world environments. Furthermore, through in-depth analysis of sensor data and real-time optimization and control, the system of the present invention excels in energy efficiency management. Particularly in extreme environmental conditions such as high temperature and high humidity, the AICA module effectively balances user comfort and energy consumption, reducing unnecessary energy waste and significantly improving the overall energy efficiency of the air conditioning system. Compared with traditional systems, the present invention can achieve up to a 15% reduction in energy consumption, saving users operating costs while also reducing carbon emissions and meeting environmental protection requirements. Furthermore, the health management module of the present invention can monitor the user's health data (such as heart rate, sleep quality, etc.) and environmental data (such as air quality) in real time, and automatically optimize the air conditioner's operating mode based on this data, providing better air quality and comfort. Furthermore, the system supports personalized scenario configuration, allowing users to customize the air conditioner's operating mode (such as work mode, rest mode, etc.) according to their habits and needs, thereby obtaining a more personalized user experience. These features significantly improve user satisfaction and health protection. The present invention also designs a system architecture with excellent scalability and compatibility, which can adapt to buildings of different sizes and diverse user needs. The system supports seamless integration with other smart home devices, allowing users to achieve coordinated control of multiple devices through a unified platform, thereby improving operational convenience and overall system coordination. This compatibility enables the system of the present invention to be easily integrated into various smart home environments, increasing the system's application scenarios and market competitiveness. Finally, the present invention implements a fault prediction function through intelligent algorithms, which can identify potential faults in the air conditioning system in advance and provide preventive maintenance recommendations. This proactive maintenance approach significantly improves the reliability and stability of the system, avoids sudden shutdowns or performance degradation caused by faults, and reduces user inconvenience and maintenance costs. In addition, the system can automatically isolate problematic components and reconfigure resources before failure occurs, thereby ensuring the normal operation of other components and extending the service life of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0062] in:

[0063] Figure 1 This is an overall architecture diagram of a split-type air-conditioning load management system according to an embodiment of the present invention;

[0064] Figure 2 This is a diagram of the internal modules of the central controller according to an embodiment of the present invention;

[0065] Figure 3 This is a flow chart of an adaptive integrated control module (AICA) according to an embodiment of the present invention;

[0066] Figure 4 This is a diagram of the health management and personalized service modules of an embodiment of the present invention;

[0067] Figure 5 A schematic diagram of load management optimization according to an embodiment of the present invention;

[0068] Figure 6 Flowchart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0069] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of the present invention.

[0070] Example 1: Overall system architecture

[0071] like Figure 1 As shown, this embodiment provides a split-type air conditioning load management system, which includes multiple air conditioning units and a central controller. Specifically:

[0072] Air conditioning units: Installed in individual rooms or areas of a building, each unit is equipped with temperature sensors, humidity sensors, energy consumption sensors, and operating status sensors. These sensors monitor the unit's temperature, humidity, energy consumption, and operating mode in real time, and transmit this data to a central controller via wireless communication.

[0073] Air conditioning units and sensor networks

[0074] Multiple sensors are installed in each independent air conditioning unit, including temperature sensors, humidity sensors, energy consumption sensors, and operating status sensors. These sensors form the basic data acquisition network of the system, which can monitor the key parameters of the air conditioning unit in real time:

[0075] Temperature sensor: used to monitor the temperature changes in a room or area to ensure that the air conditioning system can respond promptly to temperature fluctuations.

[0076] Humidity sensor: Used to detect the humidity level of the environment, especially in high humidity conditions. The data from this sensor can help the system adjust the dehumidification function of the air conditioner.

[0077] Energy consumption sensor: monitors the energy consumption of the air conditioning unit in real time, helps the system evaluate the current energy efficiency performance, and provides data support for subsequent energy efficiency optimization.

[0078] Operation status sensor: used to detect the current operating status of the air conditioning unit, such as operating mode, wind speed, compressor working condition, etc., to ensure that the system can fully understand the actual operation of the air conditioner.

[0079] These sensors transmit monitoring data to the central controller in real time through wireless communication or wired connection.

[0080] Central controller: The core control device of the system, responsible for receiving monitoring data from each air-conditioning unit, and processing and making decisions on the data through built-in functional modules. Figure 2 As shown, the central controller includes:

[0081] Data processing module: This module receives monitoring data from sensors and performs real-time processing and feature extraction using convolutional neural networks (CNNs) and long short-term memory networks (LSTMs). CNNs are primarily used to extract spatial features, while LSTMs process time series data to predict future load demand.

[0082] After receiving data from sensors such as temperature, humidity, energy consumption, and operating status, the data processing module first performs preliminary processing on this data, including noise filtering, data cleaning, and standardization. This processed data is then passed through a convolutional neural network (CNN) for spatial feature extraction. CNNs can identify correlations between different sensor data and capture complex environmental characteristics. The processed data is then input into a long short-term memory (LSTM) network, which processes time series data and predicts future load demand and environmental changes. The combination of these two deep learning models enables the system to accurately understand and predict environmental changes, providing fundamental data support for subsequent decision-making.

[0083] Adaptive Integrated Control Module (AICA): This module combines the Deep Q-Network (DQN) and Transformer model to analyze the features extracted by the data processing module, generate optimized control instructions, and dynamically adjust the operating mode of each air conditioning unit to achieve load balancing and energy efficiency optimization.

[0084] This module combines the Deep Q Network (DQN) and the Transformer model to perform in-depth analysis of the features extracted by the data processing module. The DQN is used to calculate the optimal control strategy under different environmental conditions. These strategies include adjusting the air conditioner's temperature setting and wind speed. The Transformer model further optimizes these control strategies and can handle dependencies over long time intervals, thereby generating more precise control instructions. The dynamic adaptability of the AICA module enables the air conditioning system to continuously optimize its operating mode based on changes in real-time and historical data, maintaining efficient energy management and user comfort.

[0085] Prediction module: Performs adaptive learning based on historical data and real-time data to predict future load changes and adjust system operating parameters in advance.

[0086] The prediction module uses machine learning algorithms to adaptively learn based on historical and real-time data. Its primary function is to predict future load changes and environmental fluctuations, allowing it to proactively adjust system operating parameters. For example, if the system predicts a significant temperature increase, the prediction module will preemptively activate the air conditioner's cooling mode to prevent the resulting decrease in comfort caused by temperature fluctuations. The prediction module also helps the system optimize energy usage strategies, adjusting air conditioner operation before peak load periods to avoid energy waste.

[0087] Health management module: monitors the user's health data (such as heart rate, sleep quality, etc.) and air quality in real time, and automatically adjusts the air conditioner's operating parameters based on the monitoring results to provide optimized air quality and temperature control.

[0088] The health management module is a functional module optimized for user health and air quality. This module monitors the user's health data (such as heart rate, respiratory rate, etc.) and air quality (such as PM2.5 concentration, carbon dioxide level, etc.) in real time. When the health management module detects that the user is in poor health or the air quality has deteriorated, the system will automatically adjust the operating parameters of the air conditioner. For example, when indoor air pollution is detected, the system may increase air circulation or enable the air purification function; when it detects that the user is sleeping, the system will lower the wind speed and temperature to provide a more comfortable sleeping environment. The introduction of the health management module not only improves the intelligence level of the air-conditioning system, but also significantly improves the user's comfort and health protection.

[0089] Example 2: Implementation of Adaptive Integrated Control Module (AICA)

[0090] This embodiment describes in detail the workflow of the adaptive integrated control module (AICA):

[0091] Data Collection and Preprocessing: Through a network of sensors installed in each air conditioning unit, the system collects indoor and outdoor environmental parameters and air conditioning operating status data. This data is first processed through a CNN to extract spatial features, and then processed through an LSTM to process the time series data and predict future load demand.

[0092] Feature analysis and decision generation: After data preprocessing, the AICA module combines the Deep Q Network (DQN) and Transformer model to perform decision analysis. Figure 3 As shown in Figure 2, DQN is used to calculate the optimal control strategy under different environmental states, while the Transformer model further optimizes these control instructions to adapt to the dynamically changing environment.

[0093] Real-time Adjustment and Optimization: The AICA module adjusts the operating parameters of the air conditioning units in real time based on the generated control instructions, ensuring efficient operation under varying environmental conditions. Through SEOP (Stable Entropy Optimization Strategy), the system achieves a balance between stability and adaptability.

[0094] Online learning and self-optimization: During system operation, the AICA module continuously collects new data and uses SEOP to conduct online learning and update model parameters in real time to ensure that the system is always in the best condition.

[0095] Example 3: Health Management and Personalized Services

[0096] This example describes how to improve user experience through the health management module and personalized service module:

[0097] like Figure 4 As shown, health management: The system uses the health management module to monitor the user's physiological data (such as heart rate and sleep quality) and environmental data (such as air quality) in real time. When it detects an abnormal physiological state or a decrease in air quality, the system automatically adjusts the air conditioner's operating mode. For example, when air quality decreases, the system automatically increases ventilation or activates the air purification function.

[0098] Personalized scenario configuration: Users can customize the air conditioning operating mode for different scenarios, such as work mode, rest mode, and entertainment mode, through their terminal devices. The system automatically adjusts the air conditioner's temperature setting, fan speed, and operating time based on these scenario configurations. The AICA module continuously optimizes these settings to ensure optimal user comfort in different scenarios.

[0099] Example 4: Fault prediction and preventive maintenance

[0100] This example describes in detail the working principle of the fault prediction and preventive maintenance module:

[0101] Fault Prediction: By monitoring key parameters of the air conditioning unit (such as voltage, current, and temperature) and combining them with historical operating data, the system can predict potential faults in advance. The fault prediction module analyzes this data using AICA's deep learning algorithm and generates preventive maintenance recommendations before a fault occurs.

[0102] Preventive Maintenance: When the system detects a potential failure, the Preventive Maintenance module notifies users in real time through their devices, prompting them to perform necessary maintenance. Furthermore, the system automatically isolates the faulty component and reconfigures system resources to ensure the normal operation of other components, if necessary.

[0103] Example 5: Energy Consumption Optimization and Environmentally Friendly Operation

[0104] This example describes how to improve the energy efficiency of the system through energy consumption optimization and environmentally friendly operation mode:

[0105] like Figure 5 As shown in the figure, real-time energy consumption optimization: The central controller uses the real-time energy consumption optimization module in conjunction with the AICA module to analyze real-time energy consumption data, generate energy-saving strategies, and dynamically adjust the operating parameters of the air conditioning system. For example, during periods of high energy consumption, the system automatically adjusts the air conditioning mode to reduce energy consumption.

[0106] Environmentally friendly operation mode: When the system detects changes in external environmental conditions (such as high temperature, high humidity, and air pollution), it automatically activates environmentally friendly operation mode. For example, the system can prioritize natural ventilation or reduce the operating load of air conditioners to reduce environmental impact. The user terminal device will be notified of the energy-saving effect and environmental impact of the current mode.

[0107] Example 6: Big Data Analysis and User Behavior Prediction

[0108] This example demonstrates how to improve the intelligence level of the system through the big data analysis module and the user behavior prediction module:

[0109] Big Data Analysis: Through the Big Data Analysis module, the system processes massive amounts of data from multiple sensors and extracts valuable information through advanced analytical algorithms in AICA. For example, the system can analyze historical energy consumption data, predict future energy consumption trends, and generate corresponding energy-saving tips and operational recommendations.

[0110] User behavior prediction: The system collects and analyzes historical user data, combining time series analysis with the random forest algorithm to predict future user needs. For example, the system can predict user temperature needs at different time periods and adjust the air conditioning mode in advance to optimize energy consumption and comfort.

[0111] Example 7

[0112] like Figure 6 As shown, another embodiment of the present invention provides a split air conditioner load management method, characterized in that the method includes the following steps:

[0113] S1: Monitor the temperature, humidity, energy consumption and operation mode of the air conditioning unit in real time through multiple temperature, humidity, energy consumption and operation status sensors installed in each air conditioning unit;

[0114] S2: The central controller receives and integrates the monitoring data of the sensors and generates a unified environment description and optimized control instructions using the adaptive integrated control module AICA. The AICA combines contrastive learning, deep Q network DQN, Transformer model and stable entropy optimization strategy SEOP to make control decisions.

[0115] S3: The central controller dynamically adjusts the operating parameters of the air conditioning system based on the real-time monitoring data of individual air conditioning units and the overall working status through the local optimization module and the global optimization module;

[0116] S4: Based on the intelligent algorithm and user behavior prediction module in the AICA module, user needs are predicted in advance and corresponding control instructions are generated to achieve load balancing and optimized control;

[0117] S5: Through the health management module and personalized scene configuration module, it provides customized air conditioning control strategies for user health and scenes.

[0118] Specifically, the steps for implementing the adaptive integrated control module AICA include:

[0119] 1) Using multiple sensors to collect indoor and outdoor environmental parameters and the operating status data of air conditioning units, this data is integrated to form a unified environmental description. Contrastive learning is used to pre-train the multimodal data to obtain richer feature representations suitable for dynamic load management of air conditioning systems.

[0120] 2) Use the deep learning model in the AICA module to preprocess the collected time series data, including noise filtering and missing data completion, and standardize and normalize the preprocessed data;

[0121] 3) Based on the deep Q network (DQN), the air conditioning load demand under different states is calculated, and the difference between the current state and the ideal state is analyzed through the Transformer model. The Q value function is defined as follows:

[0122]

[0123] Among them, Q(s t ,a t ) is state s t Next take action a t The expected cumulative return value that can be obtained; t Represents the current environmental status, including indoor and outdoor temperature, humidity, and air conditioning operation status information; a t Represents the state s t The actions taken under the control of the user include temperature setting and wind speed adjustment; t represents the immediate reward obtained at time step t, which is related to user comfort and energy consumption; γ represents the discount factor (0≤γ≤1), which is used to balance current and future rewards; E represents the expectation operator, which is used to calculate the expected return of the action;

[0124] 4) The stable entropy optimization strategy (SEOP) is used for control decision-making. Combining the stability of the proximal strategy optimization (PPO) and the maximum entropy exploration characteristics of the soft actor-critic algorithm (SAC), a balance is achieved between stability and adaptability, and the optimal operation mode of the air-conditioning system is selected. The core loss function of SEOP includes:

[0125] The PPO loss function is:

[0126]

[0127] The SAC loss function is:

[0128] L SAC (θ)=E t [Qφ(s t ,a t )-αlogπ θ (a t |s t )]

[0129] Among them, θ represents the parameters of the policy network; r t (θ) represents the ratio of new and old strategies; represents the potential function, which is used to evaluate the quality of the current action relative to the average level; ∈ represents the clipping range to prevent the strategy from being updated too much; Qφ(s t ,a t ) represents the Q-value estimation based on SAC, which represents the expected return of action at in state St; α represents the entropy adjustment coefficient, which controls the degree of exploration; logπθ (a t |s t ) represents the log probability of the policy network output;

[0130] 5) Generate control instructions based on the optimization decision and send them to the air conditioning unit to perform the corresponding operation;

[0131] 6) Monitor the response of the air conditioning system and dynamically optimize the model weights using the adaptive adjustment mechanism of SEOP;

[0132] 7) Continuously collect new data during system operation, use SEOP for online learning and real-time update of model parameters.

[0133] Furthermore, through the multi-source data fusion technology of the big data analysis module and the AICA module, historical energy consumption data and real-time operation data are combined to predict future energy consumption trends. Based on the prediction results, energy-saving tips and operation suggestions are generated and pushed to user terminals, prompting users to adjust their air conditioning usage patterns during high energy consumption periods.

[0134] Based on the user behavior prediction module and personalized service module in the AICA module, it analyzes user preferences and historical comfort data, generates a personalized comfort curve, and automatically adjusts the operating parameters of the air conditioning system. At the same time, it displays the current comfort status and adjustment suggestions on the user terminal device;

[0135] When the system detects changes in external environmental conditions, including high temperature, high humidity, and air pollution, it automatically activates an environmentally friendly operating mode to reduce the energy consumption and environmental impact of the air-conditioning system, and prioritizes natural ventilation or lowers the operating load of the air-conditioning. At the same time, the energy-saving effect and environmental impact of the current mode are displayed on the user terminal device.

[0136] Example 8: Experimental data and effect verification

[0137] Experimental Background: This example aims to demonstrate the effectiveness of the split-type air conditioning load management system proposed in this invention in improving energy efficiency, optimizing user comfort, and facilitating fault prediction and maintenance in practical applications. The experiment was conducted in a medium-sized office building with four floors and a total area of ​​3,000 square meters, each containing five independent rooms. The experiment lasted seven days, with 24 / 7 monitoring, covering a variety of practical application scenarios, including normal working hours, high-temperature and high-humidity environments, and low-temperature and low-humidity environments.

[0138] Experimental conditions: Location: A medium-sized office building with 20 separate rooms. Experimental duration: 7 consecutive days.

[0139] Environmental conditions: Normal working day: external temperature 25℃, humidity 50%.

[0140] High temperature and high humidity: external temperature 35°C, humidity 80%.

[0141] Low temperature and low humidity: external temperature 15℃, humidity 30%.

[0142] Comparison system: A traditional air-conditioning system (without intelligent control) is compared with the air-conditioning load management system of the present invention.

[0143] Experimental results

[0144] 1. Energy efficiency performance

[0145] The energy efficiency data demonstrates the energy-saving effect of the system of the present invention under different environmental conditions.

[0146]

[0147] 2. User comfort rating

[0148] User comfort scores and sleep quality scores reflect the comfort performance of different systems in actual applications.

[0149]

[0150] 3. Fault prediction and preventive maintenance

[0151] The performance of the fault prediction module demonstrates the intelligent maintenance capability of the system.

[0152]

[0153] Experimental data analysis

[0154] Energy efficiency improvement: Under normal working conditions, the energy efficiency of the system of the present invention is improved by 12.5%, and in high temperature and high humidity environments, the energy efficiency improvement is most significant, reaching 14.3%.

[0155] The system of the present invention significantly reduces the energy consumption of the system through real-time optimization of the adaptive integrated control module (AICA) and shows high energy efficiency under different environmental conditions.

[0156] User comfort optimization: User comfort scores show that the system of the present invention significantly improves user comfort in all test environments, especially in high temperature and high humidity environments, where the user comfort score increased by 20.6%.

[0157] Sleep quality scores also showed a similar trend, indicating that the system is significantly optimized in rest mode.

[0158] Fault prediction and maintenance efficiency: The fault prediction module successfully predicts potential system failures and provides preventative maintenance recommendations in advance, significantly reducing fault response time. Compared to traditional systems, the fault response time of the proposed system is reduced by more than 80%.

[0159] Experimental results demonstrate that the split-type air conditioning load management system of the present invention exhibits excellent energy efficiency and user comfort optimization in practical applications. Furthermore, system reliability is significantly improved through fault prediction and preventive maintenance capabilities. These data validate the technical advantages of the present invention and provide strong support for its practical application.

[0160] In summary, this invention leverages advanced intelligent control technology and deep learning algorithms to significantly enhance the intelligence, energy efficiency, and user experience of air conditioning load management systems. The system incorporates multiple functions, including dynamic response, health management, fault prediction, and preventive maintenance. This system significantly overcomes the shortcomings of existing technologies, providing a more energy-efficient, environmentally friendly, healthy, comfortable, stable, and reliable air conditioning system solution. These technological advantages make this invention highly competitive in the market and bring tangible benefits to users.

[0161] In the description of this specification, the reference terms "one embodiment," "some embodiments," "example," "specific example," or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0162] Any process or method description in the flowchart or otherwise described herein can be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process. The scope of the preferred embodiments of the present application includes additional implementations in which the functions may be performed in a different order than shown or discussed, including in a substantially simultaneous manner or in a reverse order depending on the functions involved.

[0163] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included within the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A split air conditioning load management system, comprising a plurality of air conditioning units and a central controller, characterized in that: The system comprises: Multiple temperature sensors, humidity sensors, energy consumption sensors and operating status sensors installed in each air conditioning unit to monitor the temperature, humidity, energy consumption and operating mode of the air conditioning unit in real time; A central controller in communication with the sensor, the central controller comprising: A data processing module receives the monitoring data and performs real-time processing and feature extraction on the data based on a convolutional neural network (CNN) and a long short-term memory (LSTM) network; The adaptive integrated control module AICA, combined with the deep Q network DQN and the Transformer model, analyzes the extracted features and generates optimized control instructions for dynamically adjusting the operating mode of each air conditioning unit; The prediction module performs adaptive learning based on historical data and real-time data to predict future load changes and adjust system operating parameters in advance; The central controller also includes a health management module that monitors user health data and air quality in real time, automatically adjusts air conditioning operating parameters based on the monitoring results, and provides optimized air quality and temperature control.

2. The split-type air conditioning load management system according to claim 1, characterized in that: The central controller further comprises: A local optimization module is used to dynamically adjust the wind speed, cooling capacity and operation time of a single air conditioning unit according to the real-time temperature, humidity and energy consumption data of the unit. The optimization process is supported by the deep Q network DQN in the adaptive integrated control module AICA for decision support; The global optimization module is based on the overall working status of multiple air-conditioning units, combined with big data analysis and machine learning algorithms. The AICA module generates personalized control strategies to optimize load distribution and reduce energy consumption.

3. The split-type air conditioning load management system according to claim 1, characterized in that: The central controller includes: an intelligent self-healing module that automatically isolates faulty components and reconfigures system resources when a system fault is detected by monitoring key system parameters; The fault prediction module, based on historical fault data and real-time monitoring data, uses the deep learning algorithm of the adaptive integrated control module AICA to identify possible faults in advance and generate a preventive maintenance plan; The preventive maintenance module identifies possible faults in advance based on historical operating data and the prediction model generated by the AICA module, and prompts users to perform maintenance operations, notifying users in real time through user terminal devices to perform necessary preventive maintenance.

4. The split-type air conditioning load management system according to claim 1, characterized in that: The central controller is connected to the user terminal device via wireless communication, and the user terminal device includes: The health management module automatically adjusts the operation mode of the air conditioning system by analyzing the user's health data and air quality and combining the optimization control strategy generated by the AICA module; The intelligent scene recognition module automatically identifies the current usage scenario, including work mode, rest mode, and entertainment mode, by analyzing the user's daily behavior patterns, environmental changes, and historical system operation data, and adjusts the operating parameters of the air-conditioning system accordingly.

5. The split-type air conditioning load management system according to claim 1, characterized in that: The central controller includes: an environment change perception module, which is used to detect changes in the external environment in real time and dynamically adjust the control strategy, and the AICA module processes the perception data in real time; The user behavior prediction module collects and analyzes the user's historical operation data, combines the time series analysis and random forest algorithm in the AICA module, predicts the user's future operation needs, adjusts the system operation parameters according to the prediction results, and optimizes the air conditioning mode in advance; The personalized scene configuration module allows users to customize scene configurations through user terminal devices. The system automatically adjusts the air-conditioning operation mode based on these configurations, including temperature setting, wind speed adjustment and operation time selection.

6. The split-type air conditioning load management system according to claim 1, characterized in that: The central controller includes: a big data analysis module for processing massive data from multiple sensors and extracting valuable information through advanced analysis algorithms in the adaptive integrated control module AICA; The real-time energy consumption optimization module analyzes real-time energy consumption data through the AICA module, generates energy-saving strategies and dynamically adjusts the operating parameters of the air-conditioning system.

7. The split-type air conditioning load management system according to claim 1, characterized in that: The user terminal device includes: a personalized service module, based on the user's historical usage data and preference analysis, the adaptive integrated control module AICA generates personalized control options and suggestions, including optimal temperature settings and energy-saving mode selection.

8. A split air conditioner load management method according to any one of claims 1 to 7, characterized in that: The method comprises the following steps: S1: Real-time monitoring of the temperature, humidity, energy consumption and operation mode of the air-conditioning unit through multiple temperature, humidity, energy consumption and operation status sensors installed in each air-conditioning unit; S2: The central controller receives and fuses the monitoring data of the sensors, and uses the adaptive integrated control module AICA to generate a unified environment description and optimized control instructions. The AICA combines contrastive learning, deep Q network DQN, Transformer model and stable entropy optimization strategy SEOP to make control decisions; S3: The central controller dynamically adjusts the operating parameters of the air conditioning system based on the real-time monitoring data of individual air conditioning units and the overall working status through the local optimization module and the global optimization module; S4: Based on the intelligent algorithm and user behavior prediction module in the AICA module, user needs are predicted in advance and corresponding control instructions are generated to achieve load balancing and optimized control; S5: Through the health management module and personalized scene configuration module, provide customized air conditioning control strategies for user health and scenes.

9. The split air conditioner load management method according to claim 8, characterized in that: The implementation steps of the adaptive integrated control module AICA include: 1) Collect indoor and outdoor environmental parameters and operating status data of air-conditioning units through multiple sensors, integrate these data to form a unified environmental description, and use contrastive learning to pre-train multimodal data to obtain richer feature representations suitable for dynamic load management of air-conditioning systems; 2) Use the deep learning model in the AICA module to preprocess the collected time series data, including noise filtering and missing data completion, and standardize and normalize the preprocessed data; 3) Based on the deep Q network DQN, the air conditioning load demand under different states is calculated, and the difference between the current state and the ideal state is analyzed through the Transformer model. The Q value function is defined as follows: Among them, Q(s t ,a t ) is state s t Take action a t The expected cumulative return value that can be obtained; t Represents the current environmental status, including indoor and outdoor temperature, humidity, and air conditioning operation status information; a t Represents in state s t The actions taken under the control of the controller include temperature setting and wind speed adjustment; t represents the immediate reward obtained at time step t, which is related to user comfort and energy consumption; γ represents the discount factor (0≤γ≤1), which is used to balance current and future rewards; E represents the expectation operator, which is used to calculate the expected return of the action; 4) The stable entropy optimization strategy SEOP is used for control decision-making, combining the stability of the proximal strategy optimization PPO and the maximum entropy exploration characteristics of the soft actor critic algorithm SAC to strike a balance between stability and adaptability and select the optimal operation mode of the air-conditioning system. The core loss function of SEOP includes: The PPO loss function is: The SAC loss function is: L SAC (θ)=E t [Qφ(s t ,a t )-alogπ θ (a t |s t )] Among them, θ represents the parameters of the policy network; rt(θ) represents the ratio of the new and old policies; represents the potential function, which is used to evaluate the quality of the current action relative to the average level; ∈ represents the clipping range to prevent the strategy from being updated too much; Qφ(s t ,a t ) represents the Q value estimation based on SAC, which indicates the expected return of action at in state St; α represents the entropy adjustment coefficient, which controls the degree of exploration; logπ θ (a t |s t ) represents the log probability of the policy network output; 5) Generate control instructions based on the optimization decision and send them to the air conditioning unit to perform corresponding operations; 6) Monitor the response of the air conditioning system and use the adaptive adjustment mechanism of SEOP to dynamically optimize the model weights; 7) Continuously collect new data during system operation, use SEOP for online learning and real-time update of model parameters.

10. The split air conditioner load management method according to claim 9, characterized in that: Further comprising the steps of: Through the multi-source data fusion technology of the big data analysis module and the AICA module, historical energy consumption data and real-time operation data are combined to predict future energy consumption trends, and energy-saving tips and operation suggestions are generated based on the prediction results and pushed to user terminal devices to prompt users to adjust the air conditioning usage mode during high energy consumption periods; Based on the user behavior prediction module and personalized service module in the AICA module, the user's preferences and historical comfort data are analyzed to generate a personalized comfort curve, and the operating parameters of the air-conditioning system are automatically adjusted. At the same time, the current comfort status and adjustment suggestions are displayed on the user terminal device; When the system detects changes in external environmental conditions, including high temperature, high humidity, and air pollution, it automatically enables environmentally friendly operation mode to reduce the energy consumption of the air-conditioning system and its impact on the environment, and gives priority to natural ventilation or reduces the operating load of the air-conditioning. At the same time, the energy-saving effect and environmental impact of the current mode are prompted on the user terminal device.

Images