An edge-computing-based ion cabin robot AI agent running system and method

Abstract

This invention discloses an AI-powered operating system and method for an ion chamber robot based on edge computing. The system includes an ion chamber body module, which integrates multiple physiological intervention units. These physiological intervention units include at least two or more of the following: a massage component, a bio-detection component, a red light therapy component, a negative oxygen ion generator component, a terahertz wave resonance component, and a graphene heating component. This edge-computing-based AI-powered operating system for an ion chamber robot addresses the pain point of real-time control, ensuring safety and continuity. Local deployment of edge computing nodes enables real-time local processing of user data, achieving millisecond-level latency in controlling the physiological intervention units, thus completely resolving the latency issue of cloud-based control. A built-in lightweight neural network model allows for independent safety judgment and emergency protection during network outages / weak network conditions. Combined with local priority switching logic for multi-mode communication, it ensures continuous and stable operation of the device, guaranteeing service continuity and user safety.

Description

Technical Field

[0001] This invention relates to the field of rehabilitation equipment technology, specifically to an ion chamber robot AI intelligent body operation system and method based on edge computing. Background Technology

[0002] With the development of the smart health industry, the ion chamber, as an integrated device for multiple physiological interventions, has the core requirement of providing users with personalized, efficient, and safe conditioning services. However, existing technologies have the following problems that urgently need to be addressed: Insufficient real-time control capabilities: Existing ion chambers rely on remote cloud control, resulting in high data transmission latency and an inability to achieve real-time control of physiological intervention units. In emergencies (abnormal user physiological indicators), this can easily lead to safety risks, and the device cannot operate normally during network outages / weak network conditions, compromising service continuity; Poor personalization and coordination: The lack of intelligent decision-making capabilities based on the user's real-time physiological state makes it impossible to build accurate user health models and achieve seamless automated coordination of multiple physiological intervention units, leading to timing conflicts, electromagnetic interference, and poor conditioning effects; Privacy risks: Sensitive user biometric data needs to be uploaded to the cloud for processing, which can easily lead to data leakage and fail to meet the privacy protection compliance requirements in the medical and health field; Lack of multi-device coordination capabilities: There is no effective coordination mechanism between multiple ion chambers, making it impossible to achieve load balancing and fault redundancy, resulting in low equipment utilization and insufficient service stability. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to overcome the existing defects and provide an ion chamber robot AI intelligent agent operation system and method based on edge computing, which greatly improves the real-time performance, personalization, privacy protection and multi-device collaboration capabilities of the ion chamber system, and can effectively solve the problems in the background technology.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] The edge computing-based AI-powered operation system for ion chamber robots includes:

[0006] The ion chamber module integrates multiple physiological intervention units, which include at least two or more of the following: massage component, biodetection component, red light therapy component, negative oxygen ion generating component, terahertz wave resonance component, and graphene heating component.

[0007] Edge computing nodes are deployed locally or nearby to the ion chamber module and are connected to it for real-time collection of the operation data and user status data of the physiological intervention unit, execution of localized data processing and logical judgment, and real-time control of the physiological intervention unit. The real-time control of the physiological intervention unit by the edge computing nodes is at the millisecond level.

[0008] The AI ​​intelligent agent control module is communicatively connected to the edge computing node. It is used to construct a user digital twin based on a pre-trained multimodal large model, generate dynamic conditioning strategies based on the user status data, and send control commands to the edge computing node to realize the automatic and seamless switching and collaborative operation of different physiological intervention units.

[0009] The Internet of Things (IoT) communication module communicates bidirectionally with the edge computing node and the remote cloud server, respectively, for remote monitoring of system status, incremental model updates, multi-device collaboration, and data interaction.

[0010] The human-computer interaction interface module communicates bidirectionally with the edge computing node to enable user-system command interaction, display of operating status, and transparent transmission of user commands.

[0011] Preferably, the edge computing node includes a data acquisition subunit, a local inference subunit, and a task scheduling subunit. The data acquisition subunit is connected to each physiological intervention unit and is used to acquire user physiological parameters and environmental parameters within the ion chamber module in real time through sensors. The local inference subunit has a built-in lightweight neural network model, which is used to determine the user's safety threshold based on the physiological parameters in the event of a network outage or weak network environment, and to execute emergency shutdown or protection mode in priority over cloud commands when an anomaly is detected. The task scheduling subunit is used to receive the dynamic conditioning strategy, decompose it into specific execution sequences for each physiological intervention unit, and coordinate the working sequence of each unit to avoid electromagnetic interference or energy conflicts.

[0012] Preferably, the bio-detection component includes a non-contact vital sign monitoring sensor and an impedance analysis sensor. The non-contact vital sign monitoring sensor is used to automatically detect the user's heart rate, respiratory rate, and body surface temperature after the user enters the ion chamber module. The impedance analysis sensor is used to assess the user's body water and metabolic status in real time by combining the changes in human body impedance when the graphene heating component is working. The AI ​​intelligent body control module is equipped with an adaptive adjustment unit, which is used to dynamically adjust the spectral wavelength of the red light therapy component or the release concentration of the negative oxygen ion generating component based on the real-time feedback data of the bio-detection component.

[0013] Preferably, the human-computer interaction interface module includes a voice interaction unit and a visualization display unit. The voice interaction unit supports voice wake-up and natural language command recognition, and is used to guide user operation, confirm the physiotherapy plan, and provide feedback on subjective feelings through natural dialogue. The visualization display unit is used to display the user's key physiological indicators and the current progress of the physiotherapy project after processing by the edge computing node in real time. When the user issues an interruption or change command via voice, the human-computer interaction interface module directly transmits the command to the edge computing node for priority response.

[0014] Preferably, the IoT communication module adopts 5G / 4G / WiFi / Bluetooth multi-mode communication and has a local edge communication priority switching logic. When the external network is interrupted, it automatically switches to the local communication mode of the edge computing node, and automatically synchronizes the data to the remote cloud server after the network is restored. The IoT communication module realizes multi-machine collaboration through the edge gateway. When multiple ion chamber robots are on the same local area network, they exchange the load status of each chamber. The AI ​​intelligent agent control module dynamically allocates new user service queues according to the global load distribution, or migrates user data to nearby idle devices and synchronizes the physiotherapy progress when the ion chamber robot fails.

[0015] Preferably, the system is configured with a privacy protection mechanism, which specifically includes: the user's original biometric data is stored and processed locally on the edge computing node and is not uploaded to the remote cloud server; only the anonymized statistical feature values ​​or model gradient update parameters are uploaded to the remote cloud server; the AI ​​intelligent agent control module adopts a federated learning architecture, which uses the model gradient parameters uploaded by multiple terminals to aggregate and update the global model in the cloud, and then distributes the updated model parameters to each edge computing node.

[0016] Preferably, the edge computing-based ion chamber robot AI intelligent body operation system further includes a functional control module. The functional control module is electrically connected to each physiological intervention unit of the edge computing node and the ion chamber body module, and includes a microcontroller, a drive circuit, and a status acquisition unit. The microcontroller receives control commands from the edge computing node and drives each physiological intervention unit to start, stop, and adjust parameters through the drive circuit. The status acquisition unit collects the operating status data of each physiological intervention unit in real time and feeds it back to the edge computing node to form a closed-loop control.

[0017] Preferably, the edge computing node further includes a data preprocessing unit and a data synchronization unit. The data preprocessing unit cleans, reduces noise, and normalizes the collected human body data and equipment operation data. The data synchronization unit encrypts and synchronizes the data stored locally on the edge computing node to a remote cloud server according to a preset time interval or triggering condition, and supports the download of cloud model update packages and model iteration of the AI ​​intelligent agent control module.

[0018] Preferably, the process by which the AI ​​intelligent agent control module generates the dynamic conditioning strategy is as follows:

[0019] First, it receives user voice commands or historical health record data through a natural language processing interface;

[0020] Second, by using multimodal perception algorithms to fuse the current data collected by the biological detection components with historical health records, a real-time user health status vector is constructed.

[0021] Third, based on reinforcement learning algorithms, with the reward function being the maximization of user comfort and the optimization of physiotherapy effect, the activation sequence, workload and duration of each physiological intervention unit are planned;

[0022] Fourth, the generated control command sequence is encrypted and sent to the edge computing node.

[0023] The method for operating an ion chamber robot AI intelligent agent based on edge computing includes the following steps:

[0024] Step 1: The device starts up. The edge computing node, IoT communication module, and AI intelligent agent control module complete self-tests. The IoT communication module establishes a communication connection with the remote cloud server. If the connection fails, it switches to the local edge communication mode. The user enters the ion chamber module. The biological detection component automatically collects the user's initial physiological data, or the user inputs usage instructions through the human-machine interface module.

[0025] Step 2: The edge computing node preprocesses and performs security checks on the initial physiological data and user command-related data, extracts feature data, and sends it to the AI ​​intelligent agent control module;

[0026] Step 3: The AI ​​intelligent agent control module combines the user's historical health record data to generate a dynamic conditioning strategy that includes multiple physiological intervention units working together. The user can manually adjust the strategy through the human-computer interaction interface module. The adjusted parameters are synchronized to the AI ​​intelligent agent control module to optimize the user's digital twin.

[0027] Step 4: After receiving the dynamic conditioning strategy, the edge computing node drives each physiological intervention unit to run in sequence through the function control module, and monitors the user status and the running status of each unit in real time through the status acquisition unit during the operation.

[0028] Step 5: If the edge computing node detects that the user's status exceeds the safety threshold, detects abnormal unit operation, or receives a user's voice change / interruption command, it immediately interrupts the current physiotherapy process, executes the preset protection strategy, fault alarm, or responds to new commands.

[0029] Step Six: After the physiotherapy is completed, the edge computing node generates a physiotherapy report, which is displayed to the user through the human-computer interaction interface module and synchronized to the user's terminal. At the same time, the desensitized physiotherapy data and equipment operation data are uploaded to the remote cloud server for model optimization of the AI ​​intelligent agent control module.

[0030] Compared with the prior art, the beneficial effects of the present invention are:

[0031] 1. Addressing the pain points of real-time control and ensuring security and continuity, edge computing nodes are deployed locally to achieve real-time local processing of user data. The control latency of physiological intervention units reaches the millisecond level, completely solving the problem of cloud control latency. The built-in lightweight neural network model can independently perform security judgment and emergency protection when the network is down / weak. Combined with the local priority switching logic of multi-mode communication, it ensures the continuous and stable operation of the device, ensuring service continuity and user safety.

[0032] 2. Addressing the pain points of personalization and collaboration, and improving the treatment effect, the AI ​​intelligent agent constructs a digital twin of the user based on a multimodal big model, integrates real-time physiological data and historical records, and generates dynamic treatment strategies through reinforcement learning, realizing the automatic and seamless switching and temporal collaboration of multiple physiological intervention units to avoid interference; the adaptive adjustment unit can optimize intervention parameters according to real-time physiological feedback, accurately match individual differences of users, and greatly improve the treatment effect and comfort.

[0033] 3. To address privacy concerns and meet compliance requirements, a privacy protection mechanism of "local storage + de-identified upload + federated learning" is adopted. The user's original biometric data is only processed and stored locally, and the cloud only receives de-identified statistical data and model gradient parameters, realizing "no data sharing and joint model optimization" and preventing data leakage from the source.

[0034] 4. Solve the pain points of multi-device collaboration and improve service stability. The IoT communication module realizes multi-device collaboration through the edge gateway, balances the load of the cabin in real time, and can quickly migrate user data and treatment progress in case of failure, avoid service interruption, and improve equipment utilization and overall system reliability. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the operation method of the present invention. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Please see Figure 1 The present invention provides a technical solution:

[0038] The edge computing-based AI-powered operation system for ion chamber robots includes:

[0039] The ion chamber module integrates multiple physiological intervention units, which include at least two or more of the following: massage components, biological detection components, red light therapy components, negative oxygen ion generating components, terahertz wave resonance components, and graphene heating components.

[0040] Edge computing nodes are deployed locally or nearby to the ion chamber module and communicate with it. They are used to collect real-time operating data and user status data of the physiological intervention unit, perform localized data processing and logical judgment, and realize real-time control of the physiological intervention unit.

[0041] The AI ​​intelligent agent control module communicates with edge computing nodes and is used to build a user's digital twin based on a pre-trained multimodal large model. It generates dynamic conditioning strategies based on user status data and sends control commands to the edge computing nodes.

[0042] The IoT communication module communicates bidirectionally with both edge computing nodes and remote cloud servers for remote system status monitoring, incremental model updates, multi-device collaboration, and data interaction.

[0043] The human-computer interaction interface module communicates bidirectionally with the edge computing nodes.

[0044] The edge computing node includes a data acquisition subunit, a local inference subunit, and a task scheduling subunit; the data acquisition subunit is connected to each physiological intervention unit; the local inference subunit has a built-in lightweight neural network model; and the task scheduling subunit is used to receive dynamic conditioning strategies.

[0045] The biological detection components include non-contact vital sign monitoring sensors and impedance analysis sensors; the AI ​​intelligent agent control module is equipped with an adaptive adjustment unit.

[0046] The human-computer interaction interface module includes a voice interaction unit and a visual display unit.

[0047] The IoT communication module adopts 5G / 4G / WiFi / Bluetooth multi-mode communication and has a local edge communication priority switching logic. The IoT communication module realizes multi-machine collaboration through the edge gateway. Multiple ion chamber robots exchange the load status of each chamber when they are on the same local area network. The AI ​​intelligent agent control module dynamically allocates new user service queues according to the global load distribution.

[0048] The system is equipped with a privacy protection mechanism, and the AI ​​agent control module adopts a federated learning architecture.

[0049] The edge computing-based ion chamber robot AI intelligent body operation system also includes a functional control module, which is electrically connected to the edge computing node and each physiological intervention unit of the ion chamber body module, and includes a microcontroller, a drive circuit and a status acquisition unit.

[0050] Edge computing nodes also include data preprocessing units and data synchronization units.

[0051] The process by which the AI ​​intelligent agent control module generates dynamic conditioning strategies is as follows:

[0052] First, it receives user voice commands or historical health record data through a natural language processing interface;

[0053] Second, by using multimodal perception algorithms to fuse the current data collected by the biological detection components with historical health records, a real-time user health status vector is constructed.

[0054] Third, based on reinforcement learning algorithms, with the reward function being the maximization of user comfort and the optimization of physiotherapy effect, the activation sequence, workload and duration of each physiological intervention unit are planned;

[0055] Fourth, the generated control command sequence is encrypted and sent to the edge computing node.

[0056] A method for operating an AI-powered ion chamber robot based on edge computing includes the following steps:

[0057] Step 1: The device starts up. The edge computing node, IoT communication module, and AI intelligent agent control module complete self-tests. The IoT communication module establishes a communication connection with the remote cloud server. If the connection fails, it switches to the local edge communication mode. The user enters the ion chamber module. The biological detection component automatically collects the user's initial physiological data, or the user inputs usage instructions through the human-machine interface module.

[0058] Step 2: The edge computing node preprocesses and performs security checks on the initial physiological data and user command-related data, extracts feature data, and sends it to the AI ​​intelligent agent control module;

[0059] Step 3: The AI ​​intelligent agent control module combines the user's historical health record data to generate a dynamic conditioning strategy that includes multiple physiological intervention units working together. The user can manually adjust the strategy through the human-computer interaction interface module. The adjusted parameters are synchronized to the AI ​​intelligent agent control module to optimize the user's digital twin.

[0060] Step 4: After receiving the dynamic conditioning strategy, the edge computing node drives each physiological intervention unit to run in sequence through the function control module, and monitors the user status and the running status of each unit in real time through the status acquisition unit during the operation.

[0061] Step 5: If the edge computing node detects that the user's status exceeds the safety threshold, detects abnormal unit operation, or receives a user's voice change / interruption command, it immediately interrupts the current physiotherapy process, executes the preset protection strategy, fault alarm, or responds to new commands.

[0062] Step Six: After the physiotherapy is completed, the edge computing node generates a physiotherapy report, which is displayed to the user through the human-computer interaction interface module and synchronized to the user's terminal. At the same time, the desensitized physiotherapy data and equipment operation data are uploaded to the remote cloud server for model optimization of the AI ​​intelligent agent control module.

[0063] Example 1: Specific Implementation of the Edge Computing-Based Ion Chamber Robot AI Intelligent Agent Operation System

[0064] This embodiment provides an edge computing-based AI intelligent agent operation system for an ion chamber robot, including an ion chamber body module, an edge computing node, an AI intelligent agent control module, an IoT communication module, a human-machine interface module, and a function control module. These modules work together to solve the aforementioned core technical problems, as specifically implemented below:

[0065] The ion chamber module integrates massage components, bio-detection components, red light therapy components, negative oxygen ion generating components, and graphene heating components. The bio-detection components use non-contact sensors and impedance analysis sensors to accurately collect key physiological data such as the user's heart rate, respiratory rate, and body water content, providing data support for personalized treatment and safety monitoring, and solving the problem of poor treatment effects caused by inaccurate physiological data collection.

[0066] The edge computing nodes utilize Intel NUC 12 Pro embedded edge gateways, deployed locally within the ion chamber control box, primarily addressing real-time control and offline operation issues.

[0067] Data Acquisition and Preprocessing: Physiological and device data are acquired at a frequency of 10Hz, and the data quality is improved after cleaning and noise reduction to support real-time decision-making; Local Inference: A built-in lightweight neural network model (≤50MB) can independently determine the user's physiological safety threshold when the network is disconnected and execute emergency protection within 100ms to solve the safety risks of network disconnection; Task Scheduling: The AI ​​conditioning strategy is broken down and the timing of each intervention unit is coordinated to avoid electromagnetic interference and energy conflicts, thus solving the problem of poor coordination; Data Synchronization: Anonymized data is encrypted and synchronized to the cloud at 30-minute intervals to balance privacy protection and model optimization.

[0068] The AI ​​intelligent agent control module is deployed on edge computing nodes, and its core solutions are personalized conditioning and model optimization: Digital twin: It integrates multimodal data to build a real-time health model for users, accurately matching individual differences; Dynamic strategy generation: Based on reinforcement learning, it plans the operating parameters and sequence of intervention units with the goal of comfort and conditioning effect; Federated learning: It aggregates the model gradients of multiple devices in the cloud to achieve collaborative model optimization without sharing the original data, thus balancing privacy and model accuracy.

[0069] The IoT communication module adopts the Quectel EC200S multi-mode communication module, which solves the problems of real-time communication and multi-device collaboration: Multi-mode switching: Supports 5G / 4G / WiFi / Bluetooth, prioritizes local communication, and automatically switches when the network is disconnected to ensure service continuity; Multi-device collaboration: Real-time load exchange between devices on the same local area network, dynamic allocation of service queues, and rapid migration of user data in case of failure, improving stability.

[0070] The human-computer interaction interface module supports voice wake-up and command pass-through, responding to user operations within 50ms, lowering the barrier to entry; the function control module forms a closed-loop control with the edge computing node, monitoring the device status in real time, handling anomalies promptly, extending device life, and further ensuring operational safety.

[0071] Example 2: A Method for the Operation System of an Ion Chamber Robotic AI Agent Based on Edge Computing

[0072] Step 1: Startup initialization, device self-test, IoT module prioritizes establishing 5G communication, if it fails, switch to local mode; bio-detection component collects initial physiological data of user, user inputs conditioning needs, solves the problem of failure to start when the network is disconnected;

[0073] Step 2: Data preprocessing, edge computing nodes clean and normalize data, extract features and verify them to provide high-quality data for accurate decision-making and solve the problem of processing deviation caused by data messiness;

[0074] Step 3: Strategy generation and confirmation. The AI ​​agent combines historical records and real-time data to generate personalized treatment strategies. Users can manually adjust and optimize the digital twin to solve the problem of insufficient personalization.

[0075] Step 4: Real-time execution and monitoring, edge computing node decomposition strategy, drive the collaborative operation of each intervention unit, monitor physiological and equipment status in real time, fine-tune parameters, and solve the problems of poor coordination and insufficient real-time performance;

[0076] Step 5: Abnormal response. Upon detecting physiological abnormalities, equipment malfunctions, or user commands, immediately interrupt the process and execute protection, alarm, or response operations to resolve potential safety hazards.

[0077] Step Six: End and Synchronize. Generate a physiotherapy report and synchronize it to the user's terminal. Upload the anonymized data to the cloud for model optimization, balancing privacy protection and model iteration.

[0078] Example 3: System Performance Testing

[0079] The core performance of this system was tested to verify the effectiveness of the solution to the core technical problems. The test results are as follows:

[0080] Real-time control: Average control latency of 85ms, 100% network outage operation rate, solving the pain points of real-time and continuity; Privacy protection: No data leakage, meets compliance requirements, and solves privacy risks; Treatment effect: 88% of users reported improved physiological state, and 76% of users recognized personalized strategies, solving the pain points of collaboration and personalization; Multi-machine collaboration: Even load distribution, fault migration response ≤5s, solving the pain point of insufficient collaboration.

[0081] Test results show that the system of the present invention can effectively achieve real-time and precise physiological intervention, protect user privacy and security, improve conditioning effect and system stability, and meet the usage needs of diverse health conditioning scenarios.

[0082] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An edge computing-based AI-powered ion chamber robot operating system, characterized in that, include: The ion chamber module integrates multiple physiological intervention units, which include at least two or more of the following: massage component, biodetection component, red light therapy component, negative oxygen ion generating component, terahertz wave resonance component, and graphene heating component. Edge computing nodes are deployed locally or in the vicinity of the ion chamber module and are connected to it for communication. They are used to collect the operation data and user status data of the physiological intervention unit in real time, perform localized data processing and logical judgment, and realize real-time control of the physiological intervention unit. The AI ​​intelligent agent control module is communicatively connected to the edge computing node and is used to construct a user digital twin based on a pre-trained multimodal large model, generate dynamic conditioning strategies based on the user status data, and send control commands to the edge computing node. The Internet of Things (IoT) communication module communicates bidirectionally with the edge computing node and the remote cloud server, respectively, for remote monitoring of system status, incremental model updates, multi-device collaboration, and data interaction. The human-computer interaction interface module communicates bidirectionally with the edge computing node.

2. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 1, characterized in that, The edge computing node includes a data acquisition subunit, a local inference subunit, and a task scheduling subunit; the data acquisition subunit is connected to each physiological intervention unit; the local inference subunit has a built-in lightweight neural network model; and the task scheduling subunit is used to receive the dynamic conditioning strategy.

3. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 1, characterized in that, The biological detection component includes a non-contact vital sign monitoring sensor and an impedance analysis sensor; the AI ​​intelligent agent control module is equipped with an adaptive adjustment unit.

4. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 1, characterized in that, The human-computer interaction interface module includes a voice interaction unit and a visualization display unit.

5. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 1, characterized in that, The IoT communication module adopts 5G / 4G / WiFi / Bluetooth multi-mode communication and has a switching logic that prioritizes local edge communication. The IoT communication module realizes multi-machine collaboration through the edge gateway. Multiple ion chamber robots exchange the load status of each chamber when they are on the same local area network. The AI ​​intelligent agent control module dynamically allocates new user service queues according to the global load distribution.

6. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 1, characterized in that, The system is equipped with a privacy protection mechanism, and the AI ​​agent control module adopts a federated learning architecture.

7. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 1, characterized in that, It also includes a functional control module, which is electrically connected to each physiological intervention unit of the edge computing node and the ion chamber body module, and includes a microcontroller, a drive circuit and a status acquisition unit.

8. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 2, characterized in that, The edge computing node also includes a data preprocessing unit and a data synchronization unit.

9. The edge computing-based ion chamber robot AI intelligent agent operation system according to claim 1, characterized in that, The process by which the AI ​​intelligent agent control module generates the dynamic conditioning strategy is as follows: First, it receives user voice commands or historical health record data through a natural language processing interface; Second, by using multimodal perception algorithms to fuse the current data collected by the biological detection components with historical health records, a real-time user health status vector is constructed. Third, based on reinforcement learning algorithms, with the reward function being the maximization of user comfort and the optimization of physiotherapy effect, the activation sequence, workload and duration of each physiological intervention unit are planned; Fourth, the generated control command sequence is encrypted and sent to the edge computing node.

10. A method for operating an ion chamber robot AI intelligent agent based on edge computing as described in any one of claims 1-9, characterized in that, Includes the following steps: Step 1: The device starts up. The edge computing node, IoT communication module, and AI intelligent agent control module complete self-tests. The IoT communication module establishes a communication connection with the remote cloud server. If the connection fails, it switches to the local edge communication mode. The user enters the ion chamber module. The biological detection component automatically collects the user's initial physiological data, or the user inputs usage instructions through the human-machine interface module. Step 2: The edge computing node preprocesses and performs security checks on the initial physiological data and user command-related data, extracts feature data, and sends it to the AI ​​intelligent agent control module; Step 3: The AI ​​intelligent agent control module combines the user's historical health record data to generate a dynamic conditioning strategy that includes multiple physiological intervention units working together. The user can manually adjust the strategy through the human-computer interaction interface module. The adjusted parameters are synchronized to the AI ​​intelligent agent control module to optimize the user's digital twin. Step 4: After receiving the dynamic conditioning strategy, the edge computing node drives each physiological intervention unit to run in sequence through the function control module, and monitors the user status and the running status of each unit in real time through the status acquisition unit during the operation. Step 5: If the edge computing node detects that the user's status exceeds the safety threshold, detects abnormal unit operation, or receives a user's voice change / interruption command, it immediately interrupts the current physiotherapy process, executes the preset protection strategy, fault alarm, or responds to new commands. Step Six: After the physiotherapy is completed, the edge computing node generates a physiotherapy report, which is displayed to the user through the human-computer interaction interface module and synchronized to the user's terminal. At the same time, the desensitized physiotherapy data and equipment operation data are uploaded to the remote cloud server for model optimization of the AI ​​intelligent agent control module.

Images