Intelligent home environment intelligent regulation and control system and control method thereof

Abstract

According to the intelligent home environment intelligent regulation and control system and the control method thereof, various intelligent sensors are arranged, various environment data such as indoor temperature, humidity and illumination are monitored in real time, the data are preprocessed and then sent to a data analysis layer, data features are extracted according to the monitored data and historical data based on the deep learning technology, and user preferences are deduced; the control parameters are adjusted and sent to the equipment control layer, a specific instruction is sent to an indoor smart home, an air conditioner, a humidifier or a curtain and the like are controlled to accurately control the indoor environment, the living preference of a user is met, and in the process, the effect monitoring and dynamic adjusting layer monitors the adjusting condition in real time, and the user experience is improved. If the ideal effect is not achieved after adjustment, the equipment control layer is controlled to continue adjustment, in the process, the abnormal condition of the system is detected, the power consumption of the equipment and other data are monitored at the same time, normal work of the system is guaranteed, possible faults are predicted, a user is reminded in time, and safe and energy-saving operation of the whole system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of smart home technology, and in particular to an intelligent control system for a smart home environment and a control method thereof. Background Art

[0002] With the rapid development of Internet of Things technology and the popularization of smart home concepts, smart home appliances are becoming more and more popular in the market, and smart homes with multiple functions and high intelligence are increasingly being put into use. With the development of cloud computing, big data, artificial intelligence and other technologies, smart home appliance control methods and systems have gradually become a research hotspot. They optimize the user experience through automated control strategies and achieve effective energy management.

[0003] The development of AI and 5G technologies has brought new opportunities to the smart home market. The advancement of AI technology enables smart home devices to provide more personalized and automated services through technologies such as deep learning and user behavior analysis. With the advancement of technology, consumers have more and more expectations for smart homes, and people tend to choose smart home solutions that can provide seamless experience and full range of services.

[0004] At present, although the development of smart home technology has provided convenience for users, the existing system still has obvious limitations in many aspects.

[0005] Most smart home systems only provide basic automation functions, such as temperature control, humidity regulation or lighting management. These systems often focus on a single environmental parameter, rather than forming a comprehensive environmental management solution. In addition, existing systems are difficult to meet the requirements of intelligent and personalized living environments while improving living comfort and convenience. They cannot make dynamic adjustments at any time according to user preferences, and cannot fully meet users' individual needs for a comfortable living environment.

[0006] Smart home devices and systems often operate as independent units, lacking a unified integrated platform to coordinate and manage various smart devices in the home. This decentralized system architecture leads to poor interoperability between devices, making it impossible to achieve information sharing and linkage control between devices, thereby reducing the overall system efficiency and user experience.

[0007] Since the control system lacks in-depth analysis of user behavior patterns, it often fails to accurately match users’ actual needs. The system may not be able to recognize the user’s activity patterns at different time periods, and thus cannot automatically adjust the environment settings to suit the user’s daily habits, resulting in a poor user experience.

[0008] Existing systems often lack adaptability and are unable to self-learn and optimize based on users' historical behaviors and preferences. Such systems that lack adaptive learning capabilities cannot improve their services over time to adapt to users' changing needs. In addition, traditional home appliance control methods mainly rely on manual operations by users, which is not only inefficient but also unable to meet users' growing demands for comfort, energy saving, etc. Summary of the invention

[0009] In view of this, the present invention proposes an intelligent control system for smart home environment and a control method thereof to provide a more intelligent, personalized and safe smart home environment. This system is divided into different levels according to different functions, equipped with various intelligent sensors, real-time monitoring of various environmental data such as indoor temperature, humidity, and light, and pre-processing the data before sending it to the data analysis layer. According to the monitored data and historical data, based on deep learning technology, data features are extracted, user preferences are inferred, control parameters are adjusted, and sent to the device control layer, and specific instructions are sent to the indoor smart home to control air conditioners, humidifiers or curtains to accurately control the indoor environment to meet the user's living preferences. In this process, the effect monitoring and dynamic adjustment layer monitors the adjustment in real time. If the ideal effect is not achieved after adjustment, the control device control layer continues to adjust. In this process, the abnormal situation of the system is also detected at the same time, and the power consumption of the equipment and other data are monitored to ensure the normal operation of the system, predict possible failures, and remind users in time to ensure the safe and energy-saving operation of the entire system.

[0010] The present invention proposes an Internet smart home control system for regulating indoor environment, comprising:

[0011] 1. Data collection and preprocessing layer: collect key data of indoor environment and user behavior in real time. By integrating a variety of intelligent sensors, such as temperature, humidity, and light sensors, the indoor environment status and user activities are fully monitored. In order to ensure the quality and synchronization of data, the system must collect environmental data in real time, and detect environmental conditions more frequently when the environment changes rapidly. The system also performs preprocessing operations on the collected data, including noise filtering, outlier removal, and data standardization. At the same time, a timestamp alignment algorithm is used to ensure data stream synchronization, remove noise and outliers, and ensure that data streams from different sensors can be uploaded to the cloud platform synchronously.

[0012] 2. Data processing and analysis layer: The system stores real-time data, historical environmental data, user behavior data, and control parameters. It includes servers and cloud computing platforms, which are used to store and process large amounts of data and provide the necessary computing resources to run deep learning models and data analysis algorithms; database systems, which are used to store real-time data, historical environmental data, user behavior data, and control parameters, including relational databases (MySQL, PostgreSQL); data analysis and machine learning libraries, which are used for data preprocessing, feature extraction, and implementation of machine learning algorithms; high-performance computing GPUs, which are used to increase the speed of deep learning model training and reasoning; data synchronization and timestamp alignment algorithm modules, which are used to ensure that data streams from different sensors can be uploaded to the cloud platform synchronously; user interfaces (UIs), which are used by system administrators or users to view data analysis results and adjust system parameters. Through deep learning technology, key features are extracted from key data of indoor environment and user behavior. K-means clustering algorithm and decision tree algorithm are used to classify and refine user behavior patterns, and clustering results and more detailed behavior patterns of user behavior are obtained.

[0013] First, the collected data is standardized, and the database stores real-time data, historical environmental data, user behavior data, and control parameters. A deep learning model (convolutional neural network CNN) is applied to automatically extract key features from the raw data. Including:

[0014] Time feature: time period of a day.

[0015] Environmental parameters: indoor temperature, indoor humidity, indoor light intensity.

[0016] User behavior characteristics: user activity patterns in a specific time period, such as sleep, work, and entertainment. User control habits of smart devices, such as air conditioning temperature settings and light switch times. User feedback on environmental comfort, such as the frequency of adjusting temperature and humidity.

[0017] Device status characteristics: current status of smart devices, energy consumption data of devices, maintenance and fault records of devices.

[0018] User preference characteristics: user's personalized preferences for temperature, humidity, and light, and user's preference for energy-saving mode.

[0019] Health and comfort indicators: User's physiological indicators, such as sleep patterns, activity levels. The impact of the environment on the user's health and comfort.

[0020] Energy usage characteristics: users’ energy consumption patterns, and the correlation between energy usage and user behavior.

[0021] User behavior pattern recognition is performed, combining the user's historical behavior data and personal preferences, including the user's use of smart home devices, such as the time period and frequency of smart light switching, the temperature setting, usage time, and boot time preference of smart air conditioners; interactive behavior, such as the user's preference for interaction with smart homes, whether to prefer automatic adjustment or manual adjustment, and whether the device status is adjusted frequently; environmental preferences, based on historical data, collect user preferences for temperature, light, and humidity, and the approximate range of temperature, humidity, and light that users like; user living habits, based on the time users use smart homes, infer the user's activity time, which period of time is more frequently used, and which period of time is less used. The system collects these data into the database and constructs a user feature vector. Then use the user feature vector to train the deep learning model to predict user behavior. The K-means clustering algorithm is used to cluster user behavior data to identify different user behavior patterns. First, randomly select K data points as the initial cluster center, and assign each data point to the nearest cluster center. Then update and recalculate the center of each cluster. Repeat the assignment and update steps until the cluster center no longer changes significantly or the maximum number of iterations is reached. The results of K-means clustering are further classified to refine user behavior patterns. The features that best distinguish different behavior patterns are selected. The dataset is recursively split into smaller subsets, with each node representing a feature test, each branch representing a test result, and each leaf node representing a behavior pattern. The clustering results are analyzed to identify the probability of a user's behavior in a specific pattern.

[0022] By analyzing these clustering results and behavior patterns, the probability of a user's behavior in a specific pattern can be identified. Based on these probabilities, the user's potential intention can be inferred, where the expression is:

[0023]

[0024] Among them, d represents the user behavior feature vector, β represents the output range of adjusting the entire intention recognition, ω represents the angular frequency of the sine wave, and t represents time. represents the phase shift of the sine function, γ represents the base value of the influence of the adjustment location, δ represents the influence of the adjustment location on the user's intention, η represents the influence of the adjustment device on the user's intention, and e represents the type of device the user is using after numerical encoding. d2, d3 represent the different effects of user behavior characteristics on the intensity of intention in different ranges, d(s) represents the user behavior characteristics at time s, and the integral term represents the cumulative effect from the beginning to the current time, simulating the influence of user behavior characteristics accumulated over time on the intensity of intention.

[0025] Based on the comparison of the collected indoor environment data and the inferred potential intentions of the user, the system continuously optimizes the control strategy based on reinforcement learning, interaction and feedback with the environment, and generates precise control instructions or executes instructions issued by the user. These instructions will intelligently adjust the indoor environment according to the specific needs of the user and the preset target parameters. In this way, the system can provide a more personalized and comfortable living space to ensure that the user's living experience is optimal.

[0026] 3. Device control and interaction layer: When the smart device receives the control instructions generated by the system decision, the control unit of the device will parse the instructions to determine the specific execution steps. These steps are then converted into operation commands for the internal control system. Based on these commands, the smart device generates corresponding control signals. The hardware components of the device perform predetermined operations based on these signals to ensure that the user's environmental adjustment needs are accurately achieved.

[0027] The intelligent device control method in this system is to take adaptive control decisions to meet user intentions based on user historical behaviors and preferences. These analysis results are used to generate user intention instructions and control instructions. Generating control instructions includes the following steps: Based on the results of user intention recognition, a control decision function is defined. The system decides what control measures to take to meet user intentions. The expression is:

[0028]

[0029] Among them, C d is the control decision value for user behavior, v is the influence coefficient of user intention intensity, θ is the influence coefficient of energy efficiency, E(D) is the energy efficiency of user behavior, λ is the adjustment coefficient of the external temperature on energy efficiency, T e is the external ambient temperature, μ is the adjustment coefficient of humidity on energy efficiency, H is the ambient humidity, is the adjustment coefficient of the effect of light on energy efficiency, L is the ambient light intensity, ρ is the influence coefficient of comfort, C f (d) is the comfort factor of user behavior d, ζ is the adjustment coefficient of the external environment on comfort, σ is the sensitivity of the external environment to comfort, S e is the external environmental condition; the sine wave term at the end of the formula can simulate periodic changes, such as changes in user preferences at different times of the day.

[0030] 4. Effect monitoring and dynamic adjustment layer: track user interaction and the operation status of smart devices; compare various parameters before and after the execution of control instructions to evaluate the consistency between actual effects and expected goals; examine the effect of control behavior on energy consumption, and conduct abnormal analysis of the device status after operation, monitor the temperature, energy consumption, switch status and other information of indoor equipment in real time to identify possible problems early. When the temperature of the equipment is too high, the energy consumption is too high or too low, it is judged that the equipment status is abnormal and the equipment needs to be stopped in time to prevent safety hazards or unsatisfactory working effects. In addition, it is necessary to monitor in real time whether the equipment is turned off in time after completing the task to achieve the purpose of energy saving and environmental protection. Based on the evaluation results, the system will optimize the control plan by itself; combined with the user's personalized needs and daily usage patterns, analyze Indoor historical environmental data, analyze user needs, and divide indoor functional areas, such as rest areas, office areas, etc. By using reinforcement learning algorithms, smart devices can learn the optimal behavior strategy through interaction with the environment and tailor exclusive control strategies for each area; by analyzing the real-time status and historical data of home appliances, including the operating status of the equipment, the time when abnormal conditions occur, and the time spent on resolving faults, key features are extracted from historical data, such as the pattern and frequency of fault occurrence, and influencing factors related to faults. Using these features, deep learning technology, especially recurrent neural networks (RNNs) and long short-term memory networks (LSTMs), is used to process time series data, predict equipment failures, and implement measures in advance to reduce the possibility of failures.

[0031] The present invention proposes a control method for an intelligent control system of an intelligent home environment, comprising:

[0032] Step 1: Collect key data of indoor environment and user behavior in real time. By integrating multiple smart sensors, such as temperature, humidity, and light sensors, the indoor environment status and user activities are fully monitored.

[0033] Step 2: The system stores real-time data, historical environment data, user behavior data, and control parameters. Through deep learning technology, key features are extracted from massive data to obtain clustering results of user behavior and more detailed behavior patterns.

[0034] Step 3: Infer the user’s potential intention by analyzing these clustering results and behavior patterns

[0035] Step 4: Adaptive control decisions are made to meet user intent based on user historical behavior and preferences. These analysis results are used to generate user intent instructions and control instructions.

[0036] Step 5: The smart device responds to the control instructions generated by the system decision to ensure that the user's environmental adjustment needs are accurately achieved.

[0037] Step 6: Monitor the user's interactive behavior and the operating status of the smart device; compare the parameters before and after the execution of the control command, and evaluate the degree of compliance between the actual results and the predetermined goals. The system will automatically adjust the control strategy and customize the exclusive control plan according to the user's personalized needs and daily usage habits.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] (1) Using deep learning technology, the system can extract key features from the huge data set in view of the complexity of the smart home data set, including real-time data, historical environmental data, user behavior data and control parameters. The system uses K-means clustering algorithm and decision tree algorithm to carefully classify and deeply analyze the user's behavior patterns, and obtain accurate user behavior clustering results and detailed behavior patterns. By deeply analyzing these clustering results and behavior patterns, the system can calculate the probability of user behavior in a specific mode, and then infer the user's potential intentions. Enhance the interactivity of the system, provide users with a more customized service experience, and bring broader application prospects to the field of smart homes.

[0040] (2) In view of the adaptive and predictive characteristics of smart homes, the system realizes all-round monitoring and data-based management of the living environment by capturing environmental information and user behavior data in real time, significantly enhancing the agility of response and the accuracy of operation. By building a family context model, the system can deeply analyze the user's behavior pattern and generate instructions that reflect the user's intention, so as to accurately grasp and predict the user's personalized needs, intelligently adapt to the user's actual needs, and thus improve the user experience. Based on the user's historical behavior and preferences, the system adopts an adaptive control strategy to match the user's intention and generate corresponding control instructions. Smart devices perform operations according to these instructions, monitor the operation results in real time, and dynamically adjust the device status to ensure the stable and efficient operation of home appliances, while enhancing the reliability and security of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 4 is a system function structure diagram in an embodiment of the present invention.

[0042] Figure 2 The present invention is a flowchart of a smart home appliance control method and device execution operation in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings.

[0044] This example provides an intelligent control system and method for a smart home environment, including:

[0045] The data collection and preprocessing layer is used to collect real-time data indoors through intelligent sensors. The sensors collect data at regular intervals to ensure that the data collected from different sensors and devices are synchronized in time, and to ensure that the data is timely, effective, and reliable for easy analysis and processing. After the data is collected, the data is preprocessed to determine whether to adjust the control parameters of the indoor environment, and store real-time data, preset target parameters, and control parameters. When it is determined that the control parameters of the indoor environment need to be adjusted, the control parameters of the indoor environment (temperature control parameters, humidity control parameters, and light control parameters) need to be adjusted according to the preset target parameters (target temperature, light, and humidity) and the real-time data.

[0046] The data processing and analysis layer stores historical environmental data. Based on the data stored in the data processing layer, the user's historical behavior data and personal characteristic data are collected, mainly including: environmental preferences, behavioral habits, control habits, comfort feedback, equipment usage records, etc., to build user feature vectors. The deep learning model is used to train and predict user feature vectors and item feature vectors, where the item feature vector represents a series of attributes related to smart home devices, such as device energy consumption data, device maintenance and fault records, and device current status. K-means clustering algorithm and decision tree algorithm tree algorithm classify and refine user behavior patterns, and use user interaction data to build a personalized preference model. Interaction data refers to the interaction information between users and smart home systems. Model establishment first conducts data collection and preprocessing, collects interaction data between users and smart home systems, and performs cleaning, normalization and other operations for subsequent analysis and processing; feature engineering, extracts key features from raw data, such as user's work and rest time and device usage frequency, which are closely related to automatic control and user interaction; model training, using machine learning algorithms, trains the extracted features, and builds a model that can achieve automatic control and optimize user interaction; model optimization, by adjusting model parameters and using regularization techniques and other methods to improve the accuracy and generalization ability of the model; finally, model application, deploys the trained model to the actual smart home scene to achieve automatic control of the environment and optimization of user interaction. Match the content feature vector and the user feature vector, infer the user's needs, and generate the user environment preference. At the same time, determine the target adjustment parameters according to the user environment preference. In addition, the data analysis layer receives the collected and processed real-time monitoring data of the indoor environment, analysis results and target adjustment parameters, generates control instructions, and sends them to the device control and interaction layer.

[0047] The device control and interaction layer is connected to the user's handheld terminal. After receiving the instructions from the data analysis layer or the user, the device matches the instructions that are consistent with its own module, parses the instruction content, generates corresponding control signals, and accurately adjusts the indoor environment to meet the user's environmental preference standards.

[0048] The effect monitoring and dynamic adjustment layer monitors the effect of the control layer in real time, compares the current environment with the expected target, determines whether the adjustment expectations are met, and monitors the status of the equipment in real time, such as energy consumption, temperature, and working hours, and makes dynamic adjustments in a timely manner if any abnormality is found.

[0049] This embodiment proposes an advanced smart home environment intelligent control system and method, which aims to effectively control and optimize the indoor environment to improve the comfort and convenience of residents. The system mainly includes the following key levels:

[0050] First, the data collection layer is composed of various sensors, such as temperature, humidity, and light sensors, to ensure all-round monitoring of the indoor environment. The sensors need to be monitored regularly to obtain real-time data. The collection time interval of the sensors can be adjusted dynamically. Under normal circumstances, the collection is scheduled, and the collection frequency is more frequent when an abnormality is found, so as to better monitor the data.

[0051] Next, the data processing layer needs to perform preliminary processing on the data collected by the sensor to prevent the collected data from being inaccurate, such as data cleaning, removing noise, missing values ​​and outliers in the data; data conversion, converting the raw data into a standardized format that can be understood by the machine learning model; feature extraction, extracting meaningful information from the raw sensor data; data integration, combining data from multiple sensors to form a larger data set; data reduction, reducing the original data set to a smaller data set while retaining key information. And because the collected data is real-time, in order to ensure data synchronization, a timestamp alignment algorithm is used to obtain raw data from each sensor device, record the corresponding timestamp, find the corresponding position of different sensor data on the timeline, and perform data interpolation to ensure that the data streams from different sensors can be uploaded to the cloud platform synchronously. Based on these processed data, the data processing layer can determine whether the control parameters of the indoor environment need to be adjusted. Therefore, the data processing layer has a storage module for saving real-time data, preset target parameters and current control parameters. These preset target parameters include target temperature, target humidity and target light, while the control parameters include temperature control parameters, humidity control parameters and light control parameters. When the data processing layer determines that the indoor environment is not good and needs to be adjusted, it will intelligently adjust the control parameters based on the preset target parameters and real-time data to achieve the purpose of optimizing the indoor environment.

[0052] The role of the data analysis layer is to further enhance the intelligence level of the system. It can generate user environmental preferences based on the data stored in the data processing layer, determine the final target adjustment parameters, and meet the personalized needs of users. It also uses adaptive control decisions to generate corresponding control instructions and send them to the device control and interaction layer. The role of this layer is to make the system more humane and adjust the indoor environment according to the user's preferences.

[0053] Then, after collecting data and determining the target adjustment parameters, the device control and interaction layer will be used to perform specific functions. This layer is directly connected to the user terminal. Users can view the monitored data and adjustment parameters in real time on the terminal, and can also adjust the parameters and preferences through the terminal, and manually control the indoor equipment to adjust the indoor environment. The device control layer parses the instructions received from the upper layer, generates different control signals according to different home appliances, and controls the indoor equipment to achieve the purpose of adjusting the indoor environment. For example, the indoor temperature can be adjusted by controlling the air conditioner temperature, the humidity can be adjusted by controlling the humidifier, and the light curtain switch can be adjusted. The device control layer can accurately match the home appliances that need to be adjusted according to the user's personalized needs to ensure that the indoor environment is always kept in the best state.

[0054] Finally, the effect monitoring and dynamic adjustment layer monitors the above layers, ensures the normal operation of the smart home system and can make timely adjustments. It needs to analyze the gap between the current indoor environment and the target value in real time. When the two are not matched, it will promptly notify the device control layer to make corresponding adjustments. The effect monitoring layer needs to monitor whether each device is operating normally, identify abnormal situations, distinguish between normal and suspicious behaviors, and issue alarms in time. The effect monitoring layer is mainly responsible for real-time monitoring and management of the operating status, performance and effect of the smart home system to ensure efficient, stable and safe operation of the system.

[0055] In addition, the system also sets some security measures, such as user identity authentication, using user account password verification and biometric recognition such as face and fingerprint; setting system management permissions to allow specific system administrators, family members and visitors to view and change system parameters; recording system usage time, setting restrictions to allow or prohibit users from accessing the system, and user login, operation and exit times; supporting security protocols such as encrypted transmission and cloud storage to ensure data security. Configure a multimedia junction box for telephone, data, cable TV and other cables entering the house and connecting to various information sockets in the house to achieve visual intercom function and control the switches of lights, curtains, humidifiers, air conditioners, etc.

[0056] To sum up, the smart home environment intelligent control system of this embodiment realizes comprehensive monitoring and intelligent regulation of the indoor environment through multi-level data collection, processing, parameter adjustment and equipment control. Through these functions, it not only improves the comfort and convenience of living, but also helps to save energy and reduce emissions, realize green living, and also enhances the safety of the family.

[0057] The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiment. Any equivalent modifications or changes made by ordinary technicians in this field based on the contents disclosed by the present invention should be included in the protection scope recorded in the claims.

Claims

1. An intelligent control system for a smart home environment, characterized in that: The system includes a data collection and preprocessing layer, a data processing and analysis layer, a device control and interaction layer, and an effect monitoring and dynamic adjustment layer connected in sequence, wherein: The data collection and preprocessing layer collects key data on indoor environment and user behavior in real time; it monitors indoor environment status and user activities through integrated intelligent sensors, and performs preprocessing operations on the collected data, including noise filtering, outlier removal and data standardization. At the same time, it uses a timestamp alignment algorithm to synchronize data streams from different sensors and upload them to the cloud platform; In the data processing and analysis layer, deep learning technology is used to extract key features from the collected data, and then the K-means clustering algorithm and decision tree algorithm are used to classify and refine user behavior patterns to obtain clustering results and detailed behavior patterns of user behaviors; In the device control and interaction layer, when the smart devices in the smart home environment receive the control instructions generated by the system decision, the control unit of the device will parse the instructions to determine the specific execution steps; these steps are then converted into operation commands of the internal control system; based on these commands, the smart devices generate corresponding control signals; the hardware components of the device perform predetermined operations based on these signals to achieve the user's environmental adjustment needs; In the effect monitoring and dynamic adjustment layer, the user's interactive behavior and the operating status of the smart device are tracked; the various parameters before and after the execution of the control command are compared to evaluate the consistency between the actual effect and the expected goal; the abnormal status of the equipment after the operation is analyzed, and the operating information of the indoor equipment is monitored in real time to identify possible problems.

2. The intelligent control system for an intelligent home environment according to claim 1, characterized in that: The data processing and analysis layer includes: Servers and cloud computing platforms are used to store and process large amounts of data and provide computing resources to run deep learning models and data analysis algorithms; The database system uses a relational database to store real-time data, historical environmental data, user behavior data, and control parameters; Data analysis and machine learning libraries for data preprocessing, feature extraction, and implementation of machine learning algorithms; Graphics processor GPU, for deep learning model training and reasoning; data synchronization and timestamp alignment algorithm module, to ensure that data streams from different sensors can be uploaded to the cloud platform synchronously; User interface UI is used by system administrators or users to view data analysis results and adjust system parameters.

3. The intelligent control system for an intelligent home environment according to claim 1, characterized in that: The data processing and analysis layer uses a deep learning model based on a convolutional neural network (CNN) to automatically extract key features from the raw data, including: Time characteristics: time period of a day; Environmental parameters: indoor temperature, indoor humidity, indoor light intensity; User behavior characteristics: user activity patterns in specific time periods, such as sleep, work, and entertainment; user control habits of smart devices, such as air conditioning temperature settings and light switch times; user feedback on environmental comfort, such as the frequency of adjusting temperature and humidity; Equipment status characteristics: current status of smart devices, energy consumption data of devices, maintenance and fault records of devices; User preference characteristics: user's personalized preferences for temperature, humidity, and light, and user's preference for energy-saving mode; Health and comfort indicators: physiological indicators of users, such as sleep patterns, activity levels, and the impact of the environment on user health and comfort; Energy usage characteristics: users’ energy consumption patterns, and the correlation between energy usage and user behavior.

4. The intelligent control system for an intelligent home environment according to claim 3, characterized in that: The data processing and analysis layer identifies user behavior patterns, combining historical behavior data and personal preferences of users, including the use of smart home devices by users, such as the time periods and frequencies of smart light on and off, temperature settings, usage duration, and boot time preferences of smart air conditioners; interactive behaviors, such as the user's preference for interaction with smart homes, whether automatic adjustment or manual adjustment is preferred, and whether the device status is adjusted frequently; environmental preferences, collecting user preferences for temperature, light, and humidity based on historical data, and the approximate range of temperature, humidity, and light that users like; user living habits, inferring the user's activity time based on the time the user uses the smart home, which period of time is used more frequently and which period of time is used less frequently; the system collects this data into the database and constructs a user feature vector; then uses the user feature vector to train a deep learning model to predict user behavior; uses the K-means clustering algorithm to cluster user behavior data to identify different user behavior patterns; first, randomly selects K data points as the initial cluster centers, and assigns each data point to the nearest cluster center; Then, the update is performed to recalculate the center of each cluster; the assignment and update steps are repeated until the cluster center no longer changes significantly or the maximum number of iterations is reached; the results of K-means clustering are further classified to refine user behavior patterns; and the features that best distinguish different behavior patterns are selected; Recursively split the dataset into smaller subsets, where each node represents a feature test, each branch represents a test result, and each leaf node represents a behavior pattern; Analyze clustering results and identify the probability of user behavior in specific patterns.

5. The intelligent control system for an intelligent home environment according to claim 4, characterized in that: The data processing and analysis layer identifies the probability of user behavior in a specific mode by analyzing clustering results and behavior patterns; Based on these probabilities, the user's potential intention can be inferred, where the expression is: Among them, d represents the user behavior feature vector, β represents the output range of adjusting the entire intention recognition, ω represents the angular frequency of the sine wave, and t represents time. represents the phase shift of the sine function, γ represents the base value of the influence of the adjustment location, δ represents the influence of the adjustment location on the user's intention, η represents the influence of the adjustment device on the user's intention, and e represents the type of device the user is using after numerical encoding; d2, d3 represent the different effects of user behavior characteristics on the intensity of intention in different ranges, d(s) is the user behavior characteristics at time s, and the integral term represents the cumulative effect from the beginning to the current time, simulating the influence of user behavior characteristics on the intensity of intention over time.

6. The intelligent control system for an intelligent home environment according to claim 1, characterized in that: In the device control and interaction layer, the intelligent device control method is to take adaptive control decisions to meet user intentions based on user historical behaviors and preferences. These analysis results are used to generate user intention instructions and control instructions. Generating control instructions includes the following steps: Based on the results of user intention recognition, a control decision function is defined. The system decides what control measures to take to meet user intentions. The expression is: Among them, C d is the control decision value for user behavior, v is the influence coefficient of user intention intensity, θ is the influence coefficient of energy efficiency, E(D) is the energy efficiency of user behavior, λ is the adjustment coefficient of the influence of external temperature on energy efficiency, T e is the external ambient temperature, μ is the adjustment coefficient of humidity on energy efficiency, H is the ambient humidity, is the adjustment coefficient of the effect of light on energy efficiency, L is the ambient light intensity, ρ is the influence coefficient of comfort, C f (d) is the comfort factor of user behavior d, ζ is the adjustment coefficient of the external environment on comfort, σ is the sensitivity of the external environment to comfort, S e is the external environmental condition; the sine wave term at the end of the formula can simulate periodic changes, such as changes in user preferences at different times of the day.

7. The intelligent control system for an intelligent home environment according to claim 1, characterized in that: In the effect monitoring and dynamic adjustment layer, when the temperature of the equipment is too high, the energy consumption is too high or too low, it is determined that the equipment is in an abnormal state, and the equipment is stopped in time, and real-time monitoring is carried out to see whether the equipment is turned off in time after completing the task; based on the evaluation results, the system will optimize the control plan by itself; combined with the user's personalized needs and daily usage patterns, analyze the indoor historical environmental data, analyze user needs, divide the indoor functional areas, etc. By using the reinforcement learning algorithm, the smart device learns the optimal behavior strategy through interaction with the environment, and tailors a unique control strategy for each area; By analyzing the real-time status and historical data of home appliances, including the operating status of the equipment, the time when abnormal conditions occur, and the time spent on resolving faults, key features are extracted from the historical data, including the patterns and frequency of fault occurrence, as well as influencing factors related to the faults. These features are used together with deep learning technology to process time series data, predict equipment failures, and implement measures in advance to reduce the possibility of failures.

8. A control method based on the intelligent control system of the smart home environment according to any one of claims 1 to 7, characterized in that: The method comprises the following steps: Step 1: By integrating multiple smart sensors, including temperature, humidity, and light sensors, the indoor environment status and user activities are fully monitored, and key data of indoor environment and user behavior are collected in real time; Step 2: The system stores real-time data, historical environmental data, user behavior data, and control parameters; through deep learning technology, it extracts key features from massive data to obtain clustering results and detailed behavior patterns of user behaviors; Step 3: Infer the user's potential intention by analyzing these clustering results and behavior patterns; Step 4: Adaptive control decisions are made to meet user intent based on user historical behavior and preferences. These analysis results are used to generate user intent instructions and control instructions. Step 5: The smart device responds to the control instructions generated by the system decision to ensure that the user's environmental adjustment needs are accurately achieved; Step 6: Monitor the user's interactive behavior and the operating status of the smart device; compare the parameters before and after the execution of the control command, and evaluate the degree of compliance between the actual results and the predetermined goals; automatically adjust the control strategy, and customize the exclusive control plan according to the user's personalized needs and daily usage habits.

Images