Implementation method of intelligent robot based on integrated care framework

Abstract

The invention discloses a realization method of an intelligent robot based on an integrated care framework, and relates to the technical field of artificial intelligence, and the method comprises the steps: context awareness and intelligent triggering: continuously monitoring a user state through a finite state machine model, and intelligently triggering deep analysis only when entering a preset evaluation sensitive state; the multi-layer evidence fusion and elderly integrated care framework evaluation engine is used for processing the collected data after intelligently triggering deep analysis, and converting daily behaviors of the user into quantitative and probabilistic evaluation on intrinsic capabilities of the user through a dynamic Bayesian network model; closed-loop self-adaptive intervention and feedback: planning and executing a personalized intervention task according to an evaluation result, and adjusting the task difficulty in real time through a closed-loop self-adaptive mechanism; the core of the invention lies in a unique and logic closed-loop computing architecture, and the architecture integrates context awareness, multi-layer time sequence evidence fusion and closed-loop adaptive intervention into a whole to form a complete and continuous self-optimization workflow.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method for implementing an intelligent robot based on an integrated care framework. Background Technology

[0002] Traditional care frame intelligent robots mainly include the following categories: Category 1: General-purpose companion robots, representative product: Temi personal robot (Temi Global Ltd.); Structure and principle: This type of robot typically has a mobile chassis, display screen, voice assistant, and camera. Its main functions are video calls, information retrieval, schedule reminders, and simple navigation. Its core function is as a mobile smart speaker and communication terminal. Shortcomings: ① Lack of proactive health assessment capabilities: Its perception system is mainly used for navigation, obstacle avoidance, and facial recognition, and cannot perform professional, quantitative analysis of health indicators such as gait speed, sitting / standing ability, and cognitive state. ② Interaction remains at the command response level: The robot's behavior is driven by user commands; it is a passive service provider and lacks the ability to proactively intervene based on health insights. ③ Lack of integrated professional care framework: Its functions are fragmented and do not follow any systematic medical or care theory framework, thus failing to provide integrated care.

[0003] The second category: Single-point health monitoring devices, represented by products such as millimeter-wave radar vital sign monitors, fall alarms, and smart mattresses. Structure and principle: These devices monitor specific indicators (such as heart rate, respiration, fall status, and in / out-of-bed status) using a single sensor (e.g., radar, accelerometer, pressure sensor). Disadvantages: ① Limited data dimensions, unable to be integrated for evaluation: They can only monitor isolated physiological indicators or events. For example, a fall alarm only sounds after a fall but cannot assess the declining activity level that increases the risk of falls. ② Lack of interaction and intervention capabilities: These devices are purely data collection or alarm tools; they cannot interact with users or implement any health promotion or rehabilitation interventions.

[0004] The third category: Telemedicine software platforms, represented by various health management apps and remote consultation platforms; structure and principle: Through software applications, users can manually enter health data (such as blood pressure and blood sugar) and conduct video consultations with doctors. Shortcomings: ① Reliance on user participation: Poor continuity and objectivity of data collection, heavily reliant on user compliance. ② Lack of contextualized data: Unable to obtain behavioral data about users in their real-life environments, such as home activity levels, social interactions, and nutritional intake. ③ Lack of physical execution capabilities: Unable to provide services such as companionship, guided training, and item delivery in the physical world.

[0005] In summary, existing technologies are either single-function monitoring tools or general-purpose robots lacking in-depth health insights. Specifically, this manifests in the following ways: ① Passivity and lag in health monitoring: Traditional health monitoring is mostly event-driven (e.g., fall alarms) or periodic physical examinations, lacking continuous and proactive quantitative monitoring of the elderly's intrinsic abilities (e.g., activity and cognitive abilities), making effective intervention in the early stages of functional decline impossible. ② Intrusiveness and fragmentation of the assessment process: Existing functional assessments (e.g., gait analysis, cognitive scales) typically require professionals to conduct assessments in specific scenarios, disrupting the elderly's daily lives. Furthermore, the assessment data is isolated and discontinuous, making it difficult to form a complete health trend view. ③ Impersonalized and static care plans: Care plans are often standardized, lacking mechanisms for dynamic adjustment based on continuously changing individual data, resulting in poor care efficiency and effectiveness. ④ Fragmentation of data and services: Data collected by various smart devices is scattered and fails to be integrated with a systematic, internationally recognized health management framework (e.g., ICOPE), making it difficult to translate into actionable, clinically guiding care actions. It is evident that no technology can combine a mobile, multimodal sensing robot with an authoritative, multidimensional integrated care framework for the elderly (ICOPE) to form a closed-loop system from "unobtrusive assessment" to "intelligent intervention" and then to "collaborative management". Summary of the Invention

[0006] To address the problems existing in the prior art, the purpose of this invention is to provide a method for implementing an intelligent robot based on an integrated care framework. This invention deeply integrates the robot's multimodal perception capabilities with the Integrated Care for the Elderly (ICOPE) framework for proactive, non-invasive assessment, intervention, and management of the elderly's intrinsic abilities.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is: a method for implementing an intelligent robot based on an integrated care framework, comprising the following steps: Step 1, Context Awareness and Intelligent Triggering: Continuously monitor the user's state through a finite state machine model, and intelligently trigger in-depth analysis only when the user enters a preset evaluation-sensitive state; Step 2, Multi-layered evidence fusion and integrated elderly care framework assessment engine: After intelligent triggering and deep analysis, the collected data is processed, and the user's daily behavior is transformed into a quantitative and probabilistic assessment of their intrinsic abilities through a dynamic Bayesian network model. Step 3, Closed-loop adaptive intervention and feedback: Based on the assessment results, plan and execute personalized intervention tasks, and adjust the task difficulty in real time through a closed-loop adaptive mechanism.

[0008] As a further improvement to the present invention, step 1 is specifically as follows: The multimodal raw data stream from the robot terminal is input into the finite state machine model, which is then formally defined as a quintuple finite state machine M = (S, Σ, δ, ... , F), where S represents the set of home scenarios, Σ represents the sensor input symbol, and δ is the state transition function, F is the initial state, and F is the set of final states. The finite state machine model continuously determines whether the user has entered a predefined evaluation sensitive state. If not, only low-frequency recording or ignoring is performed. If the user enters, deep analysis is triggered, and the current context is used as a data label to be passed to the next process stage.

[0009] As a further improvement of the present invention, in the sensor input symbols, each input symbol is a vector σ =<t,l,v,a,b> Where t represents a timestamp, l represents the robot's position, v represents a visual event, a represents an auditory event, and b represents a biosignal.

[0010] As a further improvement of the present invention, the state transition function δ is specifically: δ: S × Σ → S, is used to define how the robot transitions to the next state based on the current state and sensor inputs.

[0011] As a further improvement of the present invention, step 2 specifically includes the following steps: Step 2.1: Upon receiving the trigger signal, the user's continuous activity flow is first decomposed into quantifiable parameters through the behavior primitive extraction module, and each primitive B_i is associated with a parameter vector V_i. The parameters are then sent as evidence to the evidence update module of the dynamic Bayesian network model. Step 2.2: The evidence update module iteratively updates the risk probability stored in the user's intrinsic ability probability model using Bayesian formula. When the risk probability and confidence level exceed the threshold, a structured assessment finding is generated. Step 2.3: The parallel cross-domain causal relationship analysis module performs in-depth mining of long-term data and generates in-depth insight reports.

[0012] As a further improvement of the present invention, step 2.2 is specifically as follows: Treating the user's intrinsic capability domain H_k as a latent variable, and the behavioral primitive parameter V_i(t) observed at time t as a piece of evidence E_t, the belief in the state of the latent variable is iteratively updated using Bayes' theorem: P(H_k | E_1, ..., E_t) = η * P(E_t | H_k) * P(H_k | E_1, ..., E_{t-1}); Where: P(H_k | E_1, ..., E_t) is the posterior probability that the capability domain H_k is in a risky state after observing evidence at time t; P(E_t | H_k) is the likelihood function, representing the probability of observing evidence E_t when H_k is in a risky state; P(H_k | E_1, ..., E_{t-1}) is the posterior probability at time t-1, serving as the prior probability at the current time; and η is a normalization constant.

[0013] As a further improvement to the present invention, step 2.3 is specifically as follows: Given time series data with two capability domains, X(t) and Y(t), establish two autoregressive models: Constrained model: Y(t) = Σ_{i=1}^{p} α_i * Y(ti) + ε_t; Unrestricted model: Y(t) = Σ_{i=1}^{p} α_i * Y(ti) + Σ_{j=1}^{q} β_j * X(tj) + η_t; Where: p is the lag order of Y(t), an integer representing how many past time points need to be looked back when predicting the current value Y(t); Y(ti) is the lag term of Y(t), representing the value of Y at the i-th past time point; α_i is the autoregressive coefficient, representing the weight or importance of the value Y(ti) at the i-th past time point on the current value Y(t); ε_t is the residual term, representing the difference between the predicted value and the true value Y(t) at time point t; q is the lag order of X(t), an integer representing how many past time points of X need to be included when predicting Y(t); X(tj) is the lag term of X(t), representing the value of X at the j-th past time point; β_j is the Granger causality coefficient, representing the weight of the value X(tj) at the j-th past time point on the current value Y(t); η_t is the residual term, representing the value of X added at time point t. Based on historical information, the difference between the new predicted value and the true value Y(t); An F-statistic is constructed by comparing the sum of squared residuals of the two models to test whether X is a Granger cause of Y.

[0014] As a further improvement to the present invention, step 3 is specifically as follows: The error between the user's real-time performance and the preset optimal challenge point is used as input to dynamically adjust the difficulty parameter of the next intervention task; the difficulty level D(t+1) of the next intervention is determined by the following formula: D(t+1) = D(t) + K_p * e(t) + K_i * Σ_{i=0}^{t} e(i) + K_d * (e(t) -e(t-1)); Where: D(t) is the current difficulty level, e(t) = Perf_target - Perf(t) is the error between the current performance and the target performance, Perf(t) is the quantified real-time performance score, Perf_target is the preset optimal performance target, and K_p, K_i, K_d are three control parameters: proportional, integral, and derivative, which are used to respond to the current error, cumulative error, and error change trend, respectively.

[0015] As a further improvement of the present invention, the method for calculating the real-time performance breakdown Perf(t) is as follows: Perf(t)=w1·C + w2·B - w3·F; Where: C is the degree of completion, B is the degree of standardization, F is the degree of fatigue, and w1, w2, and w3 are dimensionless weighting coefficients.

[0016] As a further improvement of the present invention, step 3 also includes: feeding back the performance data during the intervention process as new evidence to the evaluation engine, forming a closed loop of continuous learning and optimization.

[0017] The beneficial effects of this invention are: 1. Achieved a shift from passive response to proactive prevention: Through continuous, unobtrusive assessment based on the ICOPE framework, risks can be identified early in the course of signs of decline in the intrinsic abilities of older adults, and proactive interventions can be carried out to effectively delay the onset of disability.

[0018] 2. Significantly improves the objectivity, continuity, and convenience of the assessment: Integrates professional assessment into daily life, avoids the "white coat effect," obtains more authentic and continuous health data, and does not interfere with the user's normal life at all.

[0019] 3. It achieves highly personalized and dynamic care: the care plan is generated based on continuously updated individual data, truly achieving "a thousand people, a thousand faces", and can be dynamically adjusted according to the intervention effect, greatly improving the accuracy and effectiveness of care.

[0020] 4. A systematic integrated care model has been constructed: For the first time, the authoritative ICOPE integrated care framework has been implemented as specific technical products, services and processes, enabling fragmented data and functions to serve a unified and clear health management goal, providing strong technical support for the integration of medical care and elderly care.

[0021] 5. Reduced care costs and improved service efficiency: By automating most daily monitoring and intervention tasks through robots, professional personnel are freed from repetitive labor, allowing them to focus more on complex decision-making and humanistic care, thus optimizing the overall allocation of care resources. Attached Figure Description

[0022] Figure 1 This is a flowchart of an embodiment of the present invention. Detailed Implementation

[0023] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0024] Example

[0025] like Figure 1 As shown, a method for implementing an intelligent robot based on an integrated care framework includes the following steps: Step 1, Context Awareness and Intelligent Triggering: The system continuously monitors the user's state through a finite state machine (FSM) model. It only intelligently triggers in-depth analysis when the user enters a preset "evaluation-sensitive state", thus solving the problem of "when to evaluate and what to evaluate".

[0026] The starting point of this embodiment is an efficient context awareness and event triggering mechanism. The multimodal raw data stream from the robot's terminal is fed into a DL-FSM state monitoring module. This module is formally defined as a finite state machine M = (S, Σ, δ, , F), where S represents the set of home scenarios, Σ represents the sensor input symbol, and δ is the state transition function. The FSM continuously determines whether the user has entered a predefined "evaluation sensitive state". If not, the system only records at a low frequency or ignores it; if it has entered (for example, recognizing the user's action of getting up from the chair), the FSM will trigger deep analysis and pass the current scenario context as a data label to the next process stage.

[0027] Specifically, this embodiment proposes a context-aware & event-triggered data acquisition and analysis mechanism to optimize computing resources and improve the accuracy of data analysis. Formal definition: The robot's working state is modeled as a finite state machine (FSM) M = (S, Σ, δ, Σ, where: S: A finite set of states representing typical home scenarios for the elderly. For example: S = {s_sleep, s_meal, s_walk, s_sit, s_exercise, ...}, where s_sleep = sleep state; s_meal = eating state; s_walk = walking state; s_sit = sitting state; s_exercise = exercise state. Σ: A finite set of input symbols, composed of multimodal sensor data. Each input symbol is a vector σ =<t, l, v, a, b> , where t = timestamp, l = robot position, v = visual event, a = auditory event, b = biosignal. δ: State transition function, δ: S × Σ → S. It defines how the robot transitions to the next state based on the current state and sensor inputs. For example, a transition rule can be expressed as: δ(s_sit, σ) → s_walk Condition: σ.v = 'get up action' ∧ σ.t ∈ [06:00, 22:00] This rule is an "IF-THEN" logic. It describes a scenario in formal language: if the (IF) algorithm believes the user is in a 'seated static' state, and (AND) a new set of sensor signals is received that simultaneously meet two conditions: 1) the vision system recognizes a 'getting up' action, and 2) this action occurs during normal activity hours (e.g., between 6 am and 10 pm); then (THEN) the understanding of the user's state should be immediately updated from 'seated static' to 'walking state'. In short, this rule is an intelligent trigger for the system to switch from a passive, low-power observation state to an active, evaluation-sensitive state ready for evaluation. Initial state, for example = s_idle (standby). F: Set of termination states (can be ignored in this application).

[0028] The system only triggers the corresponding deep analysis module when the FSM enters a predefined "evaluation-sensitive state" (such as s_walk, s_meal). This mechanism ensures that high-energy-consuming data processing is performed only in the most valuable scenarios, avoiding the indiscriminate analysis of routinely useless information.

[0029] Step 2, Multi-layer Evidence Fusion and ICOPE Assessment: After being triggered, the system processes the collected data and uses a dynamic Bayesian network (DBN) model to transform the user's daily behavior into a quantitative and probabilistic assessment of their intrinsic capabilities, thus solving the problem of "how to assess accurately and deeply".

[0030] This step is the core analysis engine of the system, responsible for transforming raw behavioral data into clinically meaningful insights. Upon receiving the trigger signal from the previous step, the system first decomposes the activity into quantifiable parameters through the behavioral primitive extraction module. These parameters are then fed as evidence into the Dynamic Bayesian Network (DBN) evidence update module, which iteratively updates the risk probabilities stored in the user's intrinsic ability probability model using the Bayesian formula P(H_k | E_t) = ... When the risk probability and confidence level exceed a threshold, the system generates a structured "assessment finding." Simultaneously, a parallel cross-domain causal association analysis module performs in-depth mining of long-term data to generate a "deep insight report." The "assessment finding" and "deep insight report" produced in this stage serve as the decision-making basis for initiating the next stage of the intervention process.

[0031] The evaluation engine in this embodiment is a multi-level temporal evidence fusion model, the core of which is a Dynamic Bayesian Network (DBN) for ICOPE evaluation. Specifically, it includes: Layer 1: Behavioral Primitives Extraction Module: This module decomposes the user's continuous activity flow into standardized, quantifiable "behavioral primitives." For example, a "Timed Stand-Walk Test (TUG)" will be decomposed into five primitives: [sit to stand], [walk forward 3 meters], [turn 180 degrees], [walk back to the chair], and [stand to sit]. Each primitive B_i is associated with a parameter vector V_i.

[0032] The second layer: Accumulation of temporal evidence and update of confidence: The user's intrinsic ability H_k (e.g., k represents "activity level") is treated as a latent variable. Each observed behavioral primitive parameter V_i(t) (at time t) is treated as a piece of evidence E_t. The belief in the state of the latent variable is iteratively updated using Bayes' theorem.

[0033] Formula: The posterior probability update for the risk state of any ICOPE capability field H_k follows the formula: P(H_k | E_1, ..., E_t) = η * P(E_t | H_k) * P(H_k | E_1, ..., E_{t-1}) Where: P(H_k | E_1, ..., E_t) is the posterior probability that the capability domain H_k is in a risky state after observing evidence at time t. P(E_t | H_k) is the likelihood function, representing the probability of observing evidence E_t (e.g., walking speed <0.8m / s) when H_k is in a risky state. This model is pre-trained from clinical data. P(H_k | E_1, ..., E_{t-1}) is the posterior probability at time t-1, serving as the prior probability at the current time. η is a normalization constant.

[0034] This iterative process smoothly integrates discrete, potentially noisy, observational data into a stable and reliable long-term assessment of the user's intrinsic capabilities.

[0035] The third layer: Cross-domain causal relationship analysis module: Using the vector autoregression (VAR) model and Granger causality test, it explores the potential lead-lag relationship between indicators of different ICOPE capability domains.

[0036] Formula: Assume there are two time series data points for different ability domains, X(t) (e.g., nutrient intake) and Y(t) (e.g., activity level score). Establish two autoregressive models: 1. Restricted model (Restricted): Y(t) = Σ_{i=1}^{p} α_i * Y(ti) + ε_t; 2. Unrestricted model: Y(t) = Σ_{i=1}^{p} α_i * Y(ti) + Σ_{j=1}^{q} β_j * X(tj) + η_t; p: Lag order of Y(t). This is an integer representing how many past time points need to be looked back to predict the current value Y(t). For example, if p=3, it means we use data from the past 3 days (Y(t-1), Y(t-2), Y(t-3)) to predict today's data. Y(ti): Lag term of Y(t). Represents the value of Y at the i-th past time point. For example, Y(t-1) is the value of Y at the previous time point (yesterday). α_i: Autoregressive coefficient. These are parameters the model needs to learn, representing the weight or importance of the value Y(ti) at the i-th past time point to the current value Y(t). ε_t: Residual term. It represents the difference between the model's predicted value and the true value Y(t) at time point t. ε_t contains all information that cannot be explained by the historical values ​​of Y itself.

[0037] q: Lag order of X(t). Similar to p, this is an integer representing how many past time points of X data need to be incorporated when predicting Y(t). X(tj): Lagged term of X(t). Represents the value of X at the j-th past time point. For example, X(t-1) is the value of X at the previous time point (yesterday). β_j: Granger causality coefficient. This is a key parameter that the model needs to learn. It represents the weight of the influence of the value X(tj) of X at the j-th past time point on the current value Y(t). If these β coefficients are statistically significant and not zero, it indicates that the historical values ​​of X are helpful in predicting Y. η_t: Residual term. It represents the difference between the predicted value of the new model and the true value Y(t) at time point t after incorporating the historical information of X.

[0038] By comparing the sum of squared residuals (RSS) of the two models, an F-statistic can be constructed to test whether X is a Granger cause of Y. If it is statistically significant, it indicates that a change in one capability domain may predict future changes in the other capability domain, providing data support for fundamental intervention.

[0039] Step 3: Closed-Loop Adaptive Intervention and Feedback: Based on the evaluation results, the system plans and executes personalized intervention tasks, and adjusts the task difficulty in real time through a closed-loop adaptive mechanism (such as a PID controller). More importantly, the performance data during the intervention process will serve as new evidence to be fed back to the evaluation engine in Step 2, forming a closed loop of continuous learning and optimization.

[0040] This step is responsible for translating assessment insights into concrete care actions, forming a closed loop. The intervention decision-making and task planning module plans the intervention task based on the input from the previous step. The human-computer interaction module is responsible for executing this task. During execution, the system collects the user's performance data Perf(t) in real time and inputs it into the Dynamic Difficulty Adjustment (DDA) engine. This engine uses the PID controller concept, adjusting the task difficulty in real time using the formula D(t+1) = D(t) + ..., forming a tight internal adaptive loop through the adjustment circuit.

[0041] Core closed-loop feedback: The user performance data collected during the intervention process is itself a valuable new piece of evidence regarding user capabilities. This data is transmitted back to the DBN evidence update module in step 2 to more accurately update the user capability probability model. It is this feedback mechanism that integrates three independent innovations into an organic whole capable of continuous learning and dynamic evolution, forming the core technological barrier of this embodiment. Ultimately, all data and reports will be synchronized to the cloud platform for use by family members and doctors.

[0042] This embodiment introduces the PID (Proportional-Integral-Derivative) controller concept from classical control theory to adaptively adjust the difficulty of the intervention task in a closed-loop manner. The error between the user's real-time performance and the preset "optimal challenge point" is used as input to dynamically adjust the difficulty parameters of the next intervention task.

[0043] Formula: The difficulty level D(t+1) of the next intervention is determined by the following formula: D(t+1) = D(t) + K_p * e(t) + K_i * Σ_{i=0}^{t} e(i) + K_d * (e(t) -e(t-1)); Where: D(t) is the current difficulty level (e.g., number of training repetitions, reaction time requirements for cognitive games, etc.). e(t) = Perf_target - Perf(t) is the error between the current performance and the target performance. Perf(t) is the quantified real-time performance score, calculated by w1•C + w2•B - w3•F (C=Completion, B=Standard, F=Fatigue). w1, w2, and w3 are configurable, dimensionless weighting coefficients. Their core function is to adjust the relative importance of the three dimensions "Completion (C)", "Standard (B)" and "Fatigue (F)" in the final performance score, so that the DDA (Dynamic Difficulty Adjustment) engine can make the most appropriate difficulty adjustment for tasks of different natures and users with different abilities. These weighting coefficients are all positive numbers (w1, w2, w3 > 0), and their sum is not necessarily 1. Perf_target is a preset optimal performance target (e.g., 0.85), representing a state of "slightly challenging but achievable". K_p, K_i, and K_d are three control parameters: proportional, integral, and derivative, responsible for responding to the current error, cumulative error, and error change trend, respectively, ensuring the speed, stability, and predictability of difficulty adjustment. Technical advantages: This mechanism ensures that the intervention task is always kept within the user's "zone of proximal development," avoiding both ineffectiveness due to excessive simplicity and user frustration or injury due to excessive difficulty, thus achieving a scientific, safe, and personalized rehabilitation closed loop.

[0044] The embodiments described above are merely illustrative of specific implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A method for implementing an intelligent robot based on an integrated care framework, characterized in that, Includes the following steps: Step 1, Context Awareness and Intelligent Triggering: Continuously monitor the user's state through a finite state machine model, and intelligently trigger in-depth analysis only when the user enters a preset evaluation-sensitive state; Step 2, Multi-layered evidence fusion and integrated elderly care framework assessment engine: After intelligent triggering and deep analysis, the collected data is processed, and the user's daily behavior is transformed into a quantitative and probabilistic assessment of their intrinsic abilities through a dynamic Bayesian network model. Step 3, Closed-loop adaptive intervention and feedback: Based on the assessment results, plan and execute personalized intervention tasks, and adjust the task difficulty in real time through a closed-loop adaptive mechanism.

2. The implementation method of the intelligent robot based on the integrated care framework according to claim 1, characterized in that, Step 1 is described in detail as follows: The multimodal raw data stream from the robot terminal is input into the finite state machine model, which is then formally defined as a quintuple finite state machine M = (S, Σ, δ, ... , F), where S represents the set of home scenarios, Σ represents the sensor input symbol, and δ is the state transition function, F is the initial state, and F is the set of final states. The finite state machine model continuously determines whether the user has entered a predefined evaluation sensitive state. If not, only low-frequency recording or ignoring is performed. If the user enters, deep analysis is triggered, and the current context is used as a data label to be passed to the next process stage.

3. The implementation method of the intelligent robot based on the integrated care framework according to claim 2, characterized in that, In the sensor input symbols, each input symbol is a vector σ =<t,l,v,a,b> Where t represents a timestamp, l represents the robot's position, v represents a visual event, a represents an auditory event, and b represents a biosignal.

4. The implementation method of the intelligent robot based on the integrated care framework according to claim 3, characterized in that, The state transition function δ is specifically: δ: S × Σ → S, is used to define how the robot transitions to the next state based on the current state and sensor inputs.

5. The method for implementing an intelligent robot based on an integrated care framework according to claim 1 or 4, characterized in that, Step 2 specifically includes the following steps: Step 2.1: Upon receiving the trigger signal, the user's continuous activity flow is first decomposed into quantifiable parameters through the behavior primitive extraction module, and each primitive B_i is associated with a parameter vector V_i. The parameters are then sent as evidence to the evidence update module of the dynamic Bayesian network model. Step 2.2: The evidence update module iteratively updates the risk probability stored in the user's intrinsic ability probability model using Bayesian formula. When the risk probability and confidence level exceed the threshold, a structured assessment finding is generated. Step 2.3: The parallel cross-domain causal relationship analysis module performs in-depth mining of long-term data and generates in-depth insight reports.

6. The implementation method of the intelligent robot based on the integrated care framework according to claim 5, characterized in that, Step 2.2 is as follows: Treating the user's intrinsic capability domain H_k as a latent variable, and the behavioral primitive parameter V_i(t) observed at time t as a piece of evidence E_t, the belief in the state of the latent variable is iteratively updated using Bayes' theorem: P(H_k | E_1, ..., E_t) = η * P(E_t | H_k) * P(H_k | E_1, ..., E_{t-1}); Where: P(H_k | E_1, ..., E_t) is the posterior probability that the capability domain H_k is in a risky state after observing evidence at time t; P(E_t | H_k) is the likelihood function, representing the probability of observing evidence E_t when H_k is in a risky state; P(H_k | E_1, ..., E_{t-1}) is the posterior probability at time t-1, serving as the prior probability at the current time; and η is a normalization constant.

7. The implementation method of the intelligent robot based on the integrated care framework according to claim 5, characterized in that, Step 2.3 is as follows: Given time series data with two capability domains, X(t) and Y(t), establish two autoregressive models: Constrained model: Y(t) = Σ_{i=1}^{p} α_i * Y(ti) + ε_t; Unrestricted model: Y(t) = Σ_{i=1}^{p} α_i * Y(ti) + Σ_{j=1}^{q} β_j * X(tj) + η_t; Where: p is the lag order of Y(t), an integer representing how many past time points need to be backtracked when predicting the current value Y(t); Y(ti) is the lag term of Y(t), representing the value of Y at the i-th past time point; α_i is the autoregressive coefficient, representing the weight or importance of the value Y(ti) at the i-th past time point on the current value Y(t); ε_t is the residual term, representing the difference between the predicted value and the true value Y(t) at time point t; q is the lag order of X(t), an integer representing how many past time points of X need to be incorporated when predicting Y(t); X(tj) is the lag term of X(t), representing the value of X at the j-th past time point; β_j is the Granger causality coefficient, representing the weight of the value X(tj) at the j-th past time point on the current value Y(t); η_t is the residual term, representing the value of X added at time point t. Based on historical information, the difference between the new predicted value and the true value Y(t); An F-statistic is constructed by comparing the sum of squared residuals of the two models to test whether X is a Granger cause of Y.

8. The method for implementing an intelligent robot based on an integrated care framework according to claim 5, characterized in that, Step 3 is as follows: The error between the user's real-time performance and the preset optimal challenge point is used as input to dynamically adjust the difficulty parameter of the next intervention task; the difficulty level D(t+1) of the next intervention is determined by the following formula: D(t+1) = D(t) + K_p * e(t) + K_i * Σ_{i=0}^{t} e(i) + K_d * (e(t) - e(t-1)); Where: D(t) is the current difficulty level, e(t) = Perf_target - Perf(t) is the error between the current performance and the target performance, Perf(t) is the quantified real-time performance score, Perf_target is the preset optimal performance target, and K_p, K_i, K_d are three control parameters: proportional, integral, and derivative, which are used to respond to the current error, cumulative error, and error change trend, respectively.

9. The implementation method of the intelligent robot based on the integrated care framework according to claim 8, characterized in that, The calculation method for the real-time performance score Perf(t) is as follows: Perf(t)=w1·C + w2·B - w3·F; Where: C is the degree of completion, B is the degree of standardization, F is the degree of fatigue, and w1, w2, and w3 are dimensionless weighting coefficients.

10. The implementation method of the intelligent robot based on the integrated care framework according to claim 8, characterized in that, Step 3 further includes: feeding performance data from the intervention process back to the evaluation engine as new evidence, forming a closed loop of continuous learning and optimization.

Images