Intelligent Internet of Things man-machine interaction system and method with edge AI processing and dynamic function arrangement

Abstract

The invention relates to the field of interaction systems, and discloses an intelligent Internet of Things man-machine interaction system with edge AI processing and dynamic function arrangement, the intelligent Internet of Things man-machine interaction system comprises a plurality of microcontrollers and edge AI processing modules integrated in the microcontrollers, each microcontroller is connected with at least one endpoint device and is used for collecting data of the endpoint device, and the edge AI processing modules are connected with the edge AI processing modules. The system comprises a plurality of microcontrollers, the microcontrollers are electrically connected with a gateway device, the gateway device coordinates the microcontrollers, the gateway device is electrically connected with a function ball library, the function ball library stores predefined function logic templates, and the gateway device is electrically connected with a dynamic function arrangement module and is used for visually creating and managing system control logic. And the gateway equipment is electrically connected with the intelligent recommendation module. Through dynamic function arrangement, a user can quickly adapt to demand changes of different application scenes, the system flexibility is improved, meanwhile, through edge AI processing, the equipment has local intelligent analysis and decision-making capabilities, and the intelligent degree of the system is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of interactive systems, and specifically to an intelligent Internet of Things human-computer interaction system and method with edge AI processing and dynamic function orchestration. Background Art

[0002] With the rapid development of Internet of Things technology, the number and types of smart devices have shown explosive growth.

[0003] However, current IoT systems face the following major challenges in practical applications:

[0004] 1. Lack of flexibility: Existing IoT systems are usually designed for specific scenarios and lack flexibility. When user needs change, the system is difficult to adapt quickly and often requires complex reprogramming and configuration by professionals.

[0005] 2. Centralized processing bottleneck: Most IoT systems use centralized data processing methods, and all data needs to be transmitted to the cloud for analysis and decision-making. This not only increases the network burden, but may also cause system response delays and cannot meet application scenarios with high real-time requirements.

[0006] 3. High complexity: As IoT application scenarios become more complex, the interaction logic between devices becomes more and more complex. Traditional programming methods are difficult to intuitively express and manage these complex logical relationships, increasing the difficulty of system development and maintenance.

[0007] 4. Poor user-friendliness: Existing IoT configuration and management tools usually require professional knowledge and are not user-friendly enough for ordinary users. This limits the widespread application and popularization of IoT technology.

[0008] 5. Low level of intelligence: Although artificial intelligence technology has made significant progress in various fields, its application in IoT systems is still limited. Most systems lack intelligent data analysis and decision-making capabilities and cannot fully tap the value of IoT data.

[0009] 6. Security and privacy issues: Transferring all data to the cloud for processing increases the risk of data leakage and also raises users' concerns about privacy protection.

[0010] 7. Limited scalability: As the number of connected devices increases, the scalability of existing systems faces challenges. It is difficult to achieve efficient management and coordination of large-scale devices.

[0011] Although some existing technologies attempt to solve some of the above problems, there is no comprehensive solution that can address these challenges at the same time. For example, although edge computing technology can reduce the pressure of centralized processing, it usually lacks sufficient intelligence. Although visual programming tools improve user-friendliness, they often have difficulty handling complex logical relationships.

[0012] Therefore, there is an urgent need for an innovative IoT human-computer interaction system and method that can comprehensively solve the above problems and provide a flexible, intelligent, easy-to-use and secure IoT application development and management platform. Summary of the invention

[0013] In view of the shortcomings of the prior art, the present invention provides an intelligent IoT human-computer interaction system and method with edge AI processing and dynamic function orchestration, which solves the problems of insufficient flexibility, delays and burdens of centralized processing, difficulty in complexity management, poor user-friendliness, low intelligence, limited scalability, and security and privacy protection in the existing human-computer interaction system.

[0014] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent IoT human-computer interaction system with edge AI processing and dynamic function orchestration, comprising multiple microcontrollers and edge AI processing modules integrated in the multiple microcontrollers, each microcontroller is connected to at least one endpoint device for collecting endpoint device data, the microcontroller is electrically connected to a gateway device, the gateway device coordinates multiple microcontrollers, the gateway device is electrically connected to a function ball library, the function ball library stores predefined function logic templates, the gateway device is electrically connected to a dynamic function orchestration module for visually creating and managing system control logic, and the gateway device is electrically connected to an intelligent recommendation module for providing optimization suggestions based on usage patterns and system status.

[0015] Preferably, the endpoint device includes at least one sensor or actuator, wherein the sensor includes at least a temperature sensor, a humidity sensor, and a light sensor, and the actuator executes control decisions generated based on rules defined by dynamic function orchestration and AI analysis results.

[0016] Preferably, the edge AI processing module includes a lightweight AI model, a model compression mechanism and a local inference engine, which are used for local intelligent analysis and decision-making of the microcontroller, and the lightweight AI model runs on the microcontroller.

[0017] Preferably, the intelligent recommendation module includes a user behavior analyzer, a performance evaluator, and a recommendation generator, and recommends appropriate function balls based on user historical operations and current device configuration.

[0018] A processing method for an intelligent IoT human-computer interaction system with edge AI processing and dynamic function orchestration, comprising the following steps:

[0019] S1. Initialize the system and identify the microcontrollers and endpoint devices in the network;

[0020] S2. Receive configuration instructions from the user through the human-machine interface;

[0021] S3. Create control logic using dynamic function orchestration modules;

[0022] S4. Perform edge AI processing on the microcontroller for local data analysis and decision making;

[0023] S5. Aggregate and analyze data from multiple microcontrollers on a gateway device;

[0024] S6. Generate control instructions based on the analysis results and predefined logic;

[0025] S7. Execute control instructions and monitor system performance;

[0026] S8. Adaptively optimize system configuration based on performance evaluation results and user feedback.

[0027] Preferably, the above steps also include the following steps:

[0028] S9. Generate system optimization suggestions using the intelligent recommendation module;

[0029] S10. Present optimization suggestions to the user and adjust the system configuration based on user feedback.

[0030] Preferably, the edge AI processing step specifically includes collecting sensor data on a microcontroller, then using a lightweight AI model to perform real-time analysis on the data, and finally making a quick response decision locally based on the analysis results.

[0031] Preferably, the dynamic function arrangement step specifically includes creating control logic on a visual interface by dragging and dropping, verifying the correctness and efficiency of the logic in real time, and converting the verified logic into executable control instructions.

[0032] A computer-readable storage medium stores a computer program, which implements the above method when executed by a processor.

[0033] The present invention provides an intelligent IoT human-computer interaction system and method with edge AI processing and dynamic function arrangement. It has the following beneficial effects:

[0034] 1. The present invention enables users to quickly adapt to changing requirements of different application scenarios through dynamic function arrangement, thereby improving system flexibility.

[0035] 2. The present invention uses edge AI processing to enable the device to have local intelligent analysis and decision-making capabilities, thereby enhancing the intelligence of the system.

[0036] 3. The present invention reduces the complexity of system configuration by designing a visual interface and intelligent recommendation function, making it more convenient for users to use the system and improving the user experience.

[0037] 4. The present invention reduces the overall burden of the system through the setting of edge processing, improves the system performance, and increases the response speed of the system.

[0038] 5. The present invention reduces the transmission of user sensitive data through the design of local data processing, improves system security, and enhances the effect of protecting user privacy.

[0039] 6. The present invention makes the system easier to expand and manage through modular design and dynamic arrangement, further improving the scalability of the system.

[0040] 7. The present invention uses a code-free development environment to enable non-professionals to quickly develop IoT applications, reducing the development cost of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a three-dimensional diagram of the overall structure of the present invention;

[0042] Figure 2 It is a schematic diagram of the internal structure of the microcontroller in the present invention;

[0043] Figure 3 A schematic diagram of the main screen of the human-machine interface of the present invention;

[0044] Figure 4 A schematic diagram of a workflow editor interface of a dynamic function arrangement module of the present invention;

[0045] Figure 5 A schematic diagram of the function ball configuration interface of the present invention;

[0046] Figure 6 This is an operation flow chart of the intelligent recommendation module of the present invention;

[0047] Figure 7 It is a schematic diagram of an embodiment of the intelligent greenhouse control system of the present invention;

[0048] Figure 8 A flow chart of the steps for configuring an intelligent greenhouse control system for the method of the present invention;

[0049] Fig. 9 The figure is an overall flow chart of the method of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in 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 embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] Example:

[0052] Please see attached Figure 1 -Attached Fig. 9 , an embodiment of the present invention provides an intelligent IoT human-computer interaction system with edge AI processing and dynamic function orchestration, including the following components:

[0053] Microcontroller (MCU), gateway device, function ball library, edge AI processing module, dynamic function orchestration module, intelligent recommendation module and human-machine interface. These components are interconnected through the network to form a complete intelligent IoT system. The microcontroller is connected to the cloud server through the gateway to realize long-term data storage and advanced analysis functions.

[0054] The microcontroller (MCU) not only undertakes traditional data acquisition and device control functions, but also integrates edge AI processing capabilities, greatly enhancing the intelligence and real-time response capabilities of the system. Each microcontroller contains the following main components: processor, memory to store programs and data, communication interface to communicate with gateways and endpoint devices, edge AI processing module containing lightweight machine learning models and inference engines

[0055] The processor performs the main computing tasks and control logic, and adopts the low-power, high-performance ARM Cortex-M series processor, such as Cortex-M4 or Cortex-M7, to meet the needs of edge AI processing;

[0056] The memory stores programs and data, including sufficient Flash memory (for storing program code and AI models) and RAM (for runtime data processing), with a configuration of 512KB Flash and 128KB RAM;

[0057] The communication interface communicates with the gateway and endpoint devices, and supports multiple communication protocols, such as Wi-Fi, Bluetooth Low Energy (BLE), and Zigbee, to adapt to different application scenarios;

[0058] The edge AI processing module includes a lightweight machine learning model and an inference engine. This is one of the core innovations of the present invention, enabling the microcontroller to have local intelligent analysis and decision-making capabilities;

[0059] The implementation method of the edge AI processing module includes the following steps:

[0060] a) Lightweight AI model design:

[0061] - Use models suitable for running on microcontrollers, such as decision trees, random forests, and small neural networks.

[0062] -Use AI frameworks optimized for embedded devices such as TensorFlow Lite for Microcontrollers or CMSIS-NN.

[0063] b) Model compression:

[0064] -Use quantization technology to compress model parameters from 32-bit floating point numbers to 8-bit integers, significantly reducing the model size.

[0065] -Apply pruning techniques to remove neural network connections that have little impact on the prediction results.

[0066] c) Optimize the inference engine:

[0067] -Use the hardware features of microcontrollers (such as DSP instruction sets) to optimize matrix operations.

[0068] -Implement data pipelining to improve the parallelism of the reasoning process.

[0069] d) Dynamic model update:

[0070] -Design incremental learning algorithms that allow models to be locally updated based on new data.

[0071] -Implement a secure model transmission and verification mechanism to ensure the security of the update process.

[0072] e) Task Scheduling:

[0073] - Implement a lightweight real-time operating system (RTOS), including FreeRTOS, to efficiently manage AI processing and other tasks.

[0074] -Design priority scheduling strategies to ensure timely execution of critical tasks.

[0075] Through this design, the microcontroller can perform tasks such as data preprocessing, anomaly detection, and simple prediction locally, which greatly reduces the burden on the gateway device and improves the response speed and reliability of the system.

[0076] The human-machine interface (HMI) includes: main interface layout, device topology map, function ball panel, system status bar, interactive design, adaptive layout, theme customization, auxiliary functions, data visualization, and user authority management;

[0077] Main interface layout: Responsive design is adopted to adapt to display devices of different sizes, from smartphones to large-screen monitors. The main interface is divided into three main areas: device topology map, function ball panel and system status bar;

[0078] Device topology diagram: Microcontroller representation: Each microcontroller is represented by a unique icon, and the color of the icon indicates the online status of the microcontroller (green for online, red for offline);

[0079] Endpoint device representation: Displays connected sensors and actuators in a satellite layout around the microcontroller icon, using different icons to distinguish device types;

[0080] Connection lines: Use lines of different styles to represent physical connections and logical relationships between devices.

[0081] Function ball panel: Function ball representation: Use icons of different colors and shapes to represent different types of function balls, such as control, analysis, optimization, etc.; Drag operation: Users can apply function balls to specific devices in the device topology map by dragging; Quick preview: Hover over a function ball to display a brief description of the function;

[0082] System status bar: displays the overall system operation status, key performance indicators, alarm information, etc.;

[0083] Interactive design: Click on the device icon to view detailed information and historical data; double-click the function ball to open the detailed configuration interface; support multi-touch operations, such as pinching and zooming the device topology map;

[0084] Adaptive layout: Automatically adjust the layout according to the number of connected devices to ensure good readability even when there are a large number of devices;

[0085] Theme customization: provides day and night modes, as well as a variety of color themes to meet the preferences and usage environment requirements of different users;

[0086] Accessibility: integrated voice control and text-to-speech functions to improve system accessibility;

[0087] Data visualization: Integrate multiple data visualization components, such as real-time curve charts, bar charts, pie charts, etc., to intuitively display system operation data;

[0088] User authority management: supports multi-level user authority settings. Users with different authorities can see different interface elements and perform different levels of operations.

[0089] Through this intuitive and interactive interface design, even users with non-technical backgrounds can easily understand and operate complex IoT systems, greatly lowering the threshold for using the system and improving user experience.

[0090] The dynamic function arrangement module includes: visual workflow editor, multi-level condition setting, logical operation support, timing control, data flow management, exception handling mechanism, version control and collaboration, simulation and debugging, API integration, and performance optimization.

[0091] Visual workflow editor: provides a drag-and-drop interface: users can create control processes by dragging function balls and connecting lines; supports multi-level nesting: allows the creation of sub-processes to improve the manageability of complex logic; real-time preview function: visually represents the process.

[0092] Multi-level condition setting: supports setting multiple trigger conditions and executing actions; provides condition combiner, allowing users to create complex condition logic, including AND, OR, NOT; supports multiple condition types such as threshold setting, range judgment, status check, etc.

[0093] Logical operation support: built-in common logical operators, including AND, OR, NOT, and XOR; provides custom operator functions to meet special business needs; supports priority setting to ensure the correct execution order of complex logic.

[0094] Timing control: supports setting the time sequence and delay of action execution; provides timing trigger function, allowing specific operations to be performed based on time conditions; realizes periodic task scheduling to meet the needs of cyclic execution.

[0095] Data flow management: Visually display the flow of data between different functional balls; support data conversion and formatting to ensure the compatibility of data between different components; provide data cache and queue management to optimize data flow processing efficiency.

[0096] Exception handling mechanism: allows users to define error handling and alternative processes in abnormal situations; provides system-level error capture and logging functions; supports setting alarm conditions and notification mechanisms.

[0097] Version control and collaboration: Integrated version control system to support process version management and rollback; provides team collaboration function, allowing multiple users to edit and review processes at the same time; implements change tracking and comparison functions to facilitate viewing of process modification history.

[0098] Simulation and debugging: Provides a workflow simulator that allows users to test process logic before deployment; supports breakpoint setting and single-step execution to facilitate debugging of complex processes; provides real-time monitoring tools to visualize data flow and execution status.

[0099] API integration: supports API integration with external systems and services to expand system functionality; provides an API management interface to facilitate user configuration and testing of API calls; implements dynamic API adaptation to automatically handle differences in different API versions and formats.

[0100] Performance optimization: Automatically analyze process complexity and provide optimization suggestions; support parallel processing to improve the execution efficiency of complex processes; achieve dynamic load balancing and optimize resource utilization.

[0101] Through this highly visual and flexible function arrangement method, users can quickly create and adjust complex IoT application logic, greatly improving the configurability and adaptability of the system. At the same time, this method also lowers the threshold for development and maintenance, allowing non-professionals to participate in the customization of smart IoT systems.

[0102] The specific components of the function balls in the function ball panel include parameter setting, condition setting, action definition, data mapping, execution strategy, exception handling, authority control, version management, testing and verification, and document generation;

[0103] Parameter setting: Provides an intuitive graphical interface for setting various parameters of the function ball; supports multiple parameter types, such as numeric, text, Boolean, and enumeration; implements parameter value range checking and format verification to prevent incorrect input; provides default values ​​and preset templates to speed up the configuration process.

[0104] Condition setting: allows users to define the conditions that trigger the execution of the function ball; supports multiple condition combinations, such as "and", "or", and "not" logical operations; provides a visual condition editor, such as a slider, switch, and drop-down menu; implements dynamic condition preview to instantly display the validity of the condition.

[0105] Action definition: configure the specific action to be performed after the function ball is triggered; support multiple action sequences and define the execution order of actions; provide an action template library that includes common control commands and data processing operations; allow custom action scripts to meet special needs.

[0106] Data mapping: defines the mapping relationship between input data and internal variables of the function ball; supports data type conversion and unit conversion; provides data preprocessing options such as filtering and normalization; implements data validity checks to ensure the quality of input data.

[0107] Execution strategy: Set the execution frequency of the function ball, such as periodicity and event triggering; define execution priority and allocate system resources reasonably; configure timeout processing and retry mechanism to improve system stability; support conditional execution and dynamically adjust the execution strategy according to the system status.

[0108] Exception handling: define error handling logic, such as retry, skip, and alarm; configure exception logging level and notification method; support custom exception types and processing procedures; provide exception simulation tools to test exception handling mechanisms.

[0109] Permission control: Set access and modification permissions for function balls; support role-based permission allocation; provide operation audit function to record all configuration changes; implement multiple authorization mechanisms for sensitive operations.

[0110] Version management: supports version control of function ball configuration; provides configuration comparison tools to intuitively display the differences between different versions; allows quick rollback to the previous configuration version; implements configuration template saving and sharing functions.

[0111] Testing and verification: Provide configuration validity check tools; support independent testing of single-function balls; implement parameter sensitivity analysis to evaluate the impact of parameter changes on results; provide performance evaluation tools to estimate the demand for system resources by function balls.

[0112] Document generation: Automatically generate function ball configuration documents; support adding custom comments and instructions; provide configuration snapshot function to facilitate team collaboration and problem tracking; realize configuration report export and support multiple formats such as PDF and HTML.

[0113] Through this comprehensive and flexible configuration interface, users can precisely control the behavior of each function ball and implement complex business logic. This approach not only ensures the powerful functions of the system, but also maintains a low threshold for use, allowing users at all levels to customize IoT applications according to their needs.

[0114] The intelligent recommendation system includes: data collection, feature extraction, model training, real-time recommendation, feedback learning, context awareness, personalized customization, cold start processing, recommendation explanation, security and privacy protection;

[0115] Data collection: Continuously record the user's operation history, including the frequency of use and configuration mode of the function ball; collect system operation data, such as equipment status, environmental parameters, and performance indicators; integrate external data sources, such as weather forecasts, energy prices, and other related information.

[0116] Feature extraction: Extract key features from the collected data, such as user preferences, usage patterns, and system load; apply natural language processing (NLP) technology to analyze text information entered by users; implement time series analysis to capture the time correlation and periodicity characteristics of the data.

[0117] Model training: Use supervised learning algorithms (such as random forest and gradient boosting tree) to train the feature ball recommendation model; apply collaborative filtering technology to make recommendations based on the behaviors of similar users; implement deep learning models (such as LSTM network) to capture complex temporal dependencies.

[0118] Real-time recommendation: Generate a real-time recommendation list based on the current system status and user context; implement multi-objective optimization to balance the relevance, diversity, and novelty of recommendations; provide explainability of recommendation reasons to help users understand the recommendation logic.

[0119] Feedback learning: Record users’ acceptance or rejection of recommendations as a basis for model optimization; implement online learning mechanisms to allow the model to dynamically adjust based on the latest feedback; design an A / B testing framework to continuously evaluate and improve the recommendation algorithm.

[0120] Context-aware: Consider the user's current task and system environment to provide context-related recommendations; implement multimodal context understanding, integrating visual, textual, and sensor data; dynamically adjust recommendation strategies to adapt to different usage scenarios and user states.

[0121] Personalized customization: Establish a personalized recommendation model for each user to adapt to different usage habits and preferences; support users to manually adjust recommendation parameters, such as recommendation frequency, type preference, etc.; achieve progressive personalization and continuously optimize personal models as usage time increases.

[0122] Cold start processing: Design a rule-based initial recommendation strategy to solve the cold start problem of new users or new functions; implement transfer learning technology to accelerate model adaptation by leveraging knowledge in similar fields; provide a guided question-and-answer interface to quickly collect user preference information.

[0123] Recommendation explanation: Provide clear and concise explanations for each recommendation to enhance user trust; implement an interactive explanation interface to allow users to gain in-depth understanding of the logic behind the recommendation; provide visualization of the impact of the recommendation to show the system changes that may be brought about by accepting the recommendation.

[0124] Security and privacy protection: Implement data anonymization and encryption technology to protect user privacy; provide configurable data sharing options to allow users to control the scope of use of personal data; design anti-fraud mechanisms to prevent malicious operations from affecting the fairness of the recommendation system.

[0125] Through this intelligent recommendation system, users can find the required functions and configurations more quickly and accurately, greatly improving the system's ease of use and work efficiency. At the same time, the system can also continuously learn and adapt to user needs, providing more and more accurate recommendations, forming a virtuous circle.

[0126] The processing method of the intelligent IoT human-computer interaction system with edge AI processing and dynamic function orchestration, this section will describe the overall method flow of the present invention in detail, showing the complete process from system initialization to continuous optimization. This process summarizes the core operation mode of the system, embodies the collaborative work of edge AI processing, dynamic function orchestration and intelligent recommendation system, including the following steps:

[0127] S1. System initialization: Start the gateway device and all connected microcontrollers. The gateway automatically scans the network, identifies all online microcontrollers and connected endpoint devices (sensors and actuators), loads the preset function ball library and AI model, initializes the system database, and prepares to record operation data and user operations;

[0128] S2. User interaction and configuration: The user views the system topology through the human-machine interface. The user selects the target microcontroller and required functions and creates preliminary control logic. The system recommends suitable function balls and configuration parameters based on user selection and historical data. The user adjusts the function ball parameters and logical relationships according to needs.

[0129] S3. Dynamic function orchestration: Users use the visual workflow editor to create complex control strategies. The system verifies the logical correctness and efficiency of the workflow in real time. Users can use the simulator to test the performance of the workflow in different scenarios. After confirmation, the system converts the workflow into executable control instructions.

[0130] S4. Edge AI processing: A pre-trained lightweight AI model is deployed on each microcontroller. The microcontroller continuously collects sensor data and uses the local AI model for preliminary analysis and prediction. Based on the output of the AI ​​model, the microcontroller can make a quick response locally, such as adjusting the actuator status. Important analysis results and abnormal conditions will be reported to the gateway device.

[0131] S5. Data aggregation and analysis: The gateway device continuously receives data and analysis results from each microcontroller. The advanced AI model on the gateway conducts in-depth analysis of the aggregated data to identify long-term trends and patterns. The system stores the analysis results in the database for subsequent decision support and system optimization.

[0132] S6. Intelligent decision-making and execution: Based on the rules defined by dynamic function orchestration and AI analysis results, the system generates control decisions. The gateway sends control instructions to the corresponding microcontroller, which receives the instructions and controls the connected actuators to achieve the expected system behavior. The system records each decision and its results for subsequent performance evaluation.

[0133] S7. Performance monitoring and evaluation: The system continuously monitors various performance indicators, such as response time, energy efficiency, control accuracy, etc., and regularly generates performance reports to highlight the system's strengths and potential areas for improvement, compare actual performance with preset targets, and identify any deviations or anomalies;

[0134] S8. Adaptive optimization: Based on the performance evaluation results, the system automatically adjusts the AI ​​model parameters and control strategies. The intelligent recommendation system generates optimization suggestions and provides them to users for review. Users can choose to accept, modify or reject the optimization suggestions. The system applies the confirmed optimization measures and monitors their effects.

[0135] S9. User feedback and learning: Collect user feedback on system performance and recommendations, use user feedback to adjust recommendation algorithms and optimization strategies, update user preference models, and provide a more personalized experience;

[0136] S10. Security and Maintenance: Continuously monitor system security status, detect potential threats or abnormal access, perform system diagnosis regularly, identify potential hardware or software problems, automatically perform software updates and security patch installations, and remind users to perform maintenance or replacement of physical equipment when necessary.

[0137] This method flow shows how the system can achieve efficient and flexible IoT control through edge computing, dynamic configuration and continuous optimization. It emphasizes the system's adaptability and user-friendliness, making complex IoT applications easy to manage and optimize.

[0138] Application examples:

[0139] Intelligent greenhouse control system, in order to better illustrate the practical application of the present invention, we take the intelligent greenhouse control system as an example to describe in detail the composition and operation mode of the system.

[0140] System composition:

[0141] The intelligent greenhouse control system consists of the following main parts:

[0142] a) Microcontroller 1: Connecting the temperature sensor and heater

[0143] -Temperature sensor: accuracy ±0.5℃, sampling frequency 1 time / minute

[0144] - Heater: adjustable power, range 0-2000W

[0145] b) Microcontroller 2: Connecting humidity sensor and irrigation system

[0146] -Humidity sensor: accuracy ±3%RH, sampling frequency 1 time / minute

[0147] -Irrigation system: flow rate adjustable, range 0-10L / min

[0148] c) Microcontroller 3: Connecting light sensor and sunshade system

[0149] -Light sensor: range 0-100000l ux, sampling frequency 1 time / minute

[0150] - Shading system: blinds, opening and closing adjustable 0-100%

[0151] d) Gateway device: coordinates the work of each microcontroller and provides a human-machine interface

[0152] -Processor: Quad-core ARM Cortex-A72

[0153] - Memory: 4GB RAM

[0154] - Storage: 64GB eMMC

[0155] - Display: 7-inch touch screen, resolution 1024x600

[0156] Function ball example:

[0157] a) Temperature control ball:

[0158] -Condition: When the temperature is below 18℃ or above 28℃

[0159] -Action: Turn on the heater when it is below 18℃, and turn off the heater when it is above 28℃

[0160] - Execution frequency: Check every 5 minutes

[0161] b) Humidity control ball:

[0162] -Condition: When humidity is below 60%

[0163] -Action: Turn on the irrigation system for 30 seconds

[0164] - Execution frequency: Check every 10 minutes

[0165] c) Light control ball:

[0166] -Condition: When the light intensity is higher than 50000l ux

[0167] -Action: Adjust the sunshade system, close 50%

[0168] - Execution frequency: Check every 15 minutes

[0169] Edge AI Applications:

[0170] a) AI model on microcontroller 1: Temperature trend prediction

[0171] - Input: Temperature data for the past 24 hours

[0172] - Output: Temperature forecast for the next 2 hours

[0173] -Model: Lightweight LSTM network

[0174] - Application: Adjust heater power in advance based on predictions to optimize energy usage b) AI model on microcontroller 2: Optimal irrigation strategy

[0175] -Input: soil moisture, ambient humidity, crop growth stage

[0176] - Output: Optimal irrigation time and water volume

[0177] -Model: Decision Tree

[0178] -Application: Balancing water use efficiency and crop growth needs

[0179] c) AI Models on Microcontroller 3: Lighting Optimization

[0180] - Input: current light intensity, weather forecast data, crop light requirements

[0181] -Output: Optimal adjustment strategy for the shading system

[0182] -Model: Reinforcement Learning Algorithm

[0183] -Application: Maximize the use of natural light and reduce the need for artificial lighting

[0184] Dynamic function orchestration example:

[0185] Users can create complex control strategies by dragging function balls, such as:

[0186] a) Temperature and humidity linkage control:

[0187] Connect the temperature control ball and the humidity control ball, and add the following conditions:

[0188] "If the temperature is above 28°C and the humidity is below 50%, start the misting system instead of the irrigation system"

[0189] b) Intelligent energy-saving mode:

[0190] Create a new feature ball that combines the temperature forecast and energy price data:

[0191] "If the temperature is predicted to rise naturally within 2 hours and it is currently a peak electricity usage period, delay starting the heater."

[0192] Intelligent recommended applications:

[0193] The system provides intelligent recommendations based on usage patterns and environmental data, such as:

[0194] a) "The humidity has been high for three consecutive days. It is recommended to adjust the irrigation strategy. Do you want to create a new humidity control rule?"

[0195] b) "Based on the current crop growth stage, it is recommended to increase the light threshold. Do you want to apply this change?"

[0196] c) "The temperature control efficiency is found to be inefficient. AI analysis recommends optimizing the heater power curve. Do you want to view detailed suggestions?"

[0197] Through this example of a smart greenhouse control system, we can see how the present invention combines edge AI processing, dynamic function orchestration, and an intelligent recommendation system to create a highly intelligent, easy-to-configure, and self-optimizing IoT application. This approach not only improves the efficiency and flexibility of the system, but also greatly reduces the threshold for using complex IoT systems.

[0198] System configuration method:

[0199] This section will explain in detail how to configure and use the intelligent IoT human-computer interaction system, taking the intelligent greenhouse control system as an example to show the complete process from initial setup to advanced function configuration.

[0200] Initialization and device identification:

[0201] a) After the system starts, it automatically scans and identifies all microcontroller devices in the network.

[0202] b) The identified microcontrollers are displayed on the main interface, with different icons indicating their type and connection status.

[0203] c) The user can view the detailed information of the microcontroller such as device ID, connected sensors and actuator types, etc. by touching the microcontroller icon on the screen.

[0204] Microcontroller and function ball association:

[0205] a) The user moves the microcontroller icon to a preset position on the interface by dragging and dropping to complete the logical grouping of the microcontrollers.

[0206] b) The system automatically analyzes the devices connected to each microcontroller and recommends the appropriate function ball type.

[0207] c) The user can select the required function balls from the function ball library, drag and drop them between the corresponding microcontrollers, and establish preliminary logical associations.

[0208] Function ball configuration:

[0209] a) Double-click the function ball icon to open the detailed configuration interface.

[0210] b) In the configuration interface, users can set:

[0211] -Trigger conditions: such as "temperature below 18°C"

[0212] - Execute an action: such as "turn on the heater and set the power to 1500W"

[0213] - Execution frequency: such as "check every 5 minutes"

[0214] c) For complex conditions, a visual condition editor is provided, which supports drag-and-drop operations to create multiple conditions.

[0215] d) The system validates user input in real time and provides immediate feedback, such as potential logic conflicts or optimization suggestions.

[0216] Edge AI model configuration:

[0217] a) In the microcontroller configuration interface, an AI model selection option is provided.

[0218] b) Users can select preset AI models, such as temperature prediction, irrigation optimization, etc.

[0219] c) For advanced users, an AI model parameter adjustment interface is provided, allowing fine-tune model performance.

[0220] d) The system provides a model performance simulation tool, and users can use historical data to test the effectiveness of AI models.

[0221] Dynamic Function Orchestration:

[0222] a) In the main interface, provide an entry to the workflow editor.

[0223] b) Users can create complex control logic in the workflow editor:

[0224] -Drag multiple function balls onto the canvas

[0225] -Use connecting lines to define the logical relationship between functional balls

[0226] -Add advanced process control elements such as conditional nodes and loop structures

[0227] c) The system automatically verifies the logical correctness of the workflow and marks potential problem areas.

[0228] d) Provides a workflow simulator that allows users to test complex logic before deployment.

[0229] Smart recommended applications:

[0230] a) The system continuously analyzes user operations and system operation data.

[0231] b) Provide optimization suggestions in the form of pop-ups or notifications when appropriate.

[0232] c) The user can choose to accept, reject or modify the recommended configuration.

[0233] d) For accepted recommendations, the system automatically applies the new configuration and records the change history.

[0234] System Monitoring and Fine-tuning:

[0235] a) Once configured, the system enters operational mode, displaying key metrics in real time on the dashboard.

[0236] b) Users can view the historical trends and current status of various parameters through the dashboard.

[0237] c) Provide quick tuning options that allow users to fine-tune system behavior without going into deep configuration.

[0238] d) The system continuously learns and self-optimizes, and regularly generates performance reports and optimization suggestions.

[0239] Remote access and control:

[0240] a) Provides web interface and mobile applications to support remote access to the system.

[0241] b) Implement multi-level authority control to ensure system security.

[0242] c) Support remote configuration and fault diagnosis to reduce the need for on-site maintenance.

[0243] d) Provide API interface to allow integration and data exchange with other systems.

[0244] This intuitive and flexible configuration method allows even non-technical users to quickly set up and optimize complex IoT systems. The system's intelligent recommendations and self-learning capabilities further reduce management difficulties, ensuring that the system can continuously optimize and adapt to changing needs.

[0245] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. Intelligent IoT human-computer interaction system with edge AI processing and dynamic function orchestration, characterized by: It includes multiple microcontrollers and edge AI processing modules integrated in the multiple microcontrollers, each microcontroller is connected to at least one endpoint device for collecting endpoint device data, the microcontroller is electrically connected to a gateway device, the gateway device coordinates multiple microcontrollers, the gateway device is electrically connected to a function ball library, the function ball library stores predefined function logic templates, the gateway device is electrically connected to a dynamic function orchestration module for visually creating and managing system control logic, and the gateway device is electrically connected to an intelligent recommendation module for providing optimization suggestions based on usage patterns and system status.

2. The intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to claim 1, characterized in that: The endpoint device includes at least one sensor or actuator, wherein the sensor includes at least a temperature sensor, a humidity sensor, and a light sensor, and the actuator executes control decisions generated based on rules defined by dynamic function orchestration and AI analysis results.

3. The intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to claim 1, characterized in that: The edge AI processing module includes a lightweight AI model, a model compression mechanism and a local inference engine, which are used for local intelligent analysis and decision-making of the microcontroller. The lightweight AI model runs on the microcontroller.

4. The intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to claim 1, characterized in that: The dynamic function orchestration module includes a visual workflow editor, a logic validator, and a simulator. The visual workflow editor is used to create complex control logic, and the simulator is used to test the performance of the workflow in different scenarios.

5. The intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to claim 1, characterized in that: The intelligent recommendation module includes a user behavior analyzer, a performance evaluator, and a recommendation generator, and recommends appropriate function balls based on user historical operations and current device configuration.

6. The processing method of the intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to any one of claims 1 to 5, characterized in that: The following steps are involved: S1. Initialize the system and identify the microcontrollers and endpoint devices in the network; S2. Receive configuration instructions from the user through the human-machine interface; S3. Create control logic using dynamic function orchestration modules; S4. Perform edge AI processing on the microcontroller for local data analysis and decision making; S5. Aggregate and analyze data from multiple microcontrollers on a gateway device; S6. Generate control instructions based on the analysis results and predefined logic; S7. Execute control instructions and monitor system performance; S8. Adaptively optimize system configuration based on performance evaluation results and user feedback.

7. The processing method of the intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to claim 6, characterized in that: The above steps also include the following steps: S9. Generate system optimization suggestions using the intelligent recommendation module; S10. Present optimization suggestions to the user and adjust the system configuration based on user feedback.

8. The processing method of the intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to claim 6, characterized in that: The edge AI processing steps specifically include collecting sensor data on a microcontroller, then using a lightweight AI model to analyze the data in real time, and finally making quick response decisions locally based on the analysis results.

9. The processing method of the intelligent IoT human-computer interaction system with edge AI processing and dynamic function arrangement according to claim 6, characterized in that: The dynamic function arrangement step specifically includes creating control logic on a visual interface by dragging and dropping, verifying the correctness and efficiency of the logic in real time, and converting the verified logic into executable control instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 6 to 9 is implemented.

Images