Multi-task intelligent coordination execution method and system based on nursing and accompanying robot

Abstract

The invention discloses a multi-task intelligent coordination execution method and system based on a pension accompanying robot, and belongs to the technical field of multi-task intelligent coordination execution, and the method comprises the steps: S1, carrying out the initialization of the system according to a current environment and preset parameters; s2, collecting information of the user and the surrounding environment thereof through an integrated sensor; s3, performing fusion analysis processing according to the collected information, predicting the behavior trend of the user, identifying the current demand of the user according to a data analysis result and a prediction result, and automatically generating a task; by integrating advanced data fusion, behavior pattern recognition and prediction, demand recognition and task evaluation, the intelligent level and service quality of the system are remarkably improved, and more accurate behavior pattern recognition and trend prediction are realized through membership function and sliding window prediction; and in combination with the behavior pattern and the trend prediction result, tasks meeting user requirements are generated, and priorities are set through a multi-dimensional evaluation model, so that the efficiency and rationality of task execution are ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi-task intelligent coordinated execution, and specifically refers to a multi-task intelligent coordinated execution method and system based on an elderly care robot. Background Art

[0002] As the population ages, elderly care has become a focus of social attention. Traditional care methods rely on manpower, which has problems such as low efficiency and unstable service quality. In recent years, with the development of artificial intelligence technology, robots have been increasingly used in the field of elderly care. However, most robots can only perform a single task or require frequent human intervention to switch tasks, which cannot meet the complex and changing care needs.

[0003] However, there are still certain defects in the multi-task coordinated execution of existing elderly care robots. The multi-task coordinated execution of existing elderly care robots relies on a single data source and lacks the ability to comprehensively analyze multi-source heterogeneous data, resulting in incomplete and inaccurate demand identification. Demand identification is usually based on static models, which does not take into account the dynamic changes of user behavior patterns and is difficult to adapt to changes in user needs. It is impossible to identify needs according to the user's actual environment and status, resulting in inaccurate identification results. It only relies on a single standard. Once the task priority is set, it is difficult to dynamically adjust it according to the actual situation. When user needs or environmental conditions change, the system cannot respond in time, ignoring the diversity and complexity of tasks, resulting in inflexible priority setting. Existing task evaluation often only considers one aspect of the factors and ignores other important factors, resulting in incomplete evaluation results. It lacks the ability to dynamically adjust according to specific circumstances and is difficult to cope with changing service needs. For this reason, a multi-task intelligent coordinated execution method and system based on elderly care robots are proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a multi-task intelligent coordinated execution method and system based on an elderly care robot to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a multi-task intelligent coordination execution method based on an elderly care robot, comprising the following steps:

[0006] S1, the system is initialized according to the current environment and preset parameters;

[0007] S2, collects information about the user and his / her surroundings through integrated sensors;

[0008] S3: Perform fusion analysis and processing based on the collected information, predict the user's behavior trend, identify the user's current needs based on the data analysis results and prediction results, and automatically generate tasks;

[0009] S4. Evaluate the generated tasks through a multi-dimensional evaluation model and set priorities;

[0010] S5. Decompose complex tasks into multiple small tasks and optimize the order of tasks;

[0011] S6. Record the results of each task execution and dynamically adjust the model and system parameters based on the feedback information.

[0012] Among them, in S1, the system is initialized according to the current state and preset parameters, obtains the user's current environmental information through integrated sensors, initializes the system parameters according to the user's historical data requirements and preferences, and dynamically updates the system parameters through real-time data and user feedback.

[0013] It not only makes adaptive adjustments based on the current environment, but also integrates the user's historical data and preferences to provide a precise starting point for subsequent task execution. This comprehensive initialization process ensures that the system can quickly and accurately identify and adapt to different elderly care environments, providing more intimate and personalized services for the elderly. At the same time, through dynamic updates of real-time data and user feedback, the system can continuously optimize its parameter settings to better meet the changing needs of users.

[0014] Among them, S3 performs fusion analysis and processing based on the collected information, predicts the user's behavior trend, identifies the user's current needs based on the data analysis results and prediction results, and automatically generates tasks; let the sensor data obtained by S2 be r i (t), assign a trust weight w to the data of each sensor i i (t), the data is fused into F(t), and the implementation formula is:

[0015]

[0016] In the formula, F(t) represents the fusion data at time t, n represents the number of sensors, and w i (t) represents the trust weight of the i-th sensor at time t, r i (t) represents the data correlation coefficient of the i-th sensor at time t, S i (t) represents the data value of the i-th sensor at time t, w i (t) According to the historical data of the sensor, the deterministic dynamic adjustment is made and the trust weight is updated as follows:

[0017] w i (t+1)=w i (t)·exp(-αe i (t)),

[0018] In the formula, α represents the new task adjustment rate constant, e i (t) represents the error of the i-th sensor at time t, r i (t) represents the data correlation coefficient of the i-th sensor at time t, which is implemented as:

[0019]

[0020] In the formula, p ij ( t ) represents the correlation between the i-th sensor and the j-th sensor at time t;

[0021] Feature extraction is performed based on the fused data, and dynamic adjustments are made based on the user's current behavior pattern and current environment. The feature extraction implementation formula is:

[0022] φ(F(t))=F(t)·A(t),

[0023] In the formula, φ(F(t)) represents the extracted feature vector, F(t) represents the fused data, A(t) represents the dynamic feature weight matrix at time t, and each element a of the feature weight matrix A(t) is ij (t) represents the importance coefficient of the i-th feature in the j-th behavior mode, expressed as:

[0024]

[0025] In the formula, b ij (t) represents the unnormalized importance coefficient of the i-th feature in the j-th behavior mode at time t.

[0026] Wherein, S3 identifies the user's behavior pattern according to the extracted feature vector, and assumes that each behavior pattern P i There is a eigenvector, let the membership function μ i (φ(F(t))) indicates that the feature vector φ(F(t)) belongs to the behavior pattern P i The final behavior pattern recognition formula is:

[0027]

[0028] In the formula, B(t) represents the behavior pattern recognition result at time t, P i The characteristic vector representing the i-th behavior mode, μ i (φ(F(t))) represents the membership function of the behavior mode in the ith order, expressed as:

[0029]

[0030] In the formula, θi represents the fuzziness parameter of the i-th behavior mode, || || 2 Indicates calculating the Euclidean distance between two vectors;

[0031] Predict the user's behavior data at the next moment t+1 based on historical data The behavior data B(tk) of the past K time steps is averaged by adaptive weights to obtain the predicted value. The implementation formula is:

[0032]

[0033] In the formula, represents the behavior trend prediction result at time t+1, K represents the sliding window size, and λ k (t) represents the weight factor of the kth time step at time t, B(tk) represents the behavior pattern recognition result at time tk, λ k (t) By dynamically adjusting the weight according to the historical data error, the formula is:

[0034]

[0035] In the formula, γ represents the decay rate of the weight factor, and exp represents the exponential function.

[0036] Among them, the S3 performs comprehensive demand identification based on the identified behavior patterns and behavior trend prediction results, and the implementation formula is:

[0037]

[0038] In the formula, D(t) represents the demand identification result at time t, B i (t) represents the behavior pattern recognition result at time t, represents the behavior trend prediction result at time t+1, ψ i Represents the demand type weight vector corresponding to the i-th behavior mode;

[0039] According to the identified user needs, tasks are generated and the implementation formula is:

[0040]

[0041] In the formula, T represents the generated task, T pool represents the candidate task pool, D i (t) represents the demand intensity of the ith demand type at time t, ψ i Indicates the importance coefficient of the requirement type.

[0042] By integrating and analyzing the collected information and predicting the user's behavior trends, the system not only improves the system's recognition accuracy of user needs and enables the system to predict and meet users' potential needs in advance, but also continuously optimizes its prediction model by dynamically adjusting parameters such as feature weight matrix and membership function to adapt to different environments and changes in user behavior, thereby improving the timeliness and effectiveness of the service.

[0043] Among them, S4 evaluates the generated tasks through a multi-dimensional evaluation model and sets priorities; the evaluation dimensions include urgency U, importance I, security S, resource consumption R and user experience X, and a multi-dimensional evaluation model is constructed according to the evaluation dimensions to comprehensively consider the evaluation results of each dimension. The implementation formula is:

[0044]

[0045] In the formula, E(T) represents the comprehensive evaluation score of task T, U(T) represents the urgency score of task T, I(T) represents the importance score of task T, S(T) represents the security score of task T, R(T) represents the resource occupation score of task T, and X(T) represents the user experience score of task T.

[0046] Among them, S4 sets the priority of the task according to the comprehensive evaluation score, and the implementation formula is:

[0047]

[0048] In the formula, P(T) represents the standardized priority score of task T, min(E) represents the minimum value of all task evaluation scores, and max(E) represents the maximum value of all task evaluation scores. Tasks are sorted according to the standardized evaluation scores, which can be implemented as follows:

[0049]

[0050] In the formula, T sd Represents a list of tasks sorted from high to low priority, T pl Represents the candidate task pool; dynamically adjusts the priority according to the actual situation and task execution status, and is implemented as follows:

[0051] P′(T)=P(T)+ΔP(T),

[0052] In the formula, P′(T) represents the adjusted priority score, and ΔP(T) represents the priority increment adjusted according to the task execution status.

[0053] Among them, S6 records the results of each task execution, and dynamically adjusts the model and system parameters according to the feedback information, records the user's initial state before the task execution and the basic information of the task, including but not limited to physiological indicators, psychological state, and environmental conditions, records the real-time data generated during the task execution and the task execution results, and collects user feedback information and execution results for adjustment and update.

[0054] The multi-dimensional evaluation model enables the system to comprehensively consider multiple dimensions such as urgency, importance, security, resource consumption and user experience to evaluate the generated tasks. The comprehensive and detailed evaluation method ensures that the system can reasonably allocate resources and give priority to key tasks while avoiding waste and redundancy of resources. By setting priorities, the system can efficiently and orderly execute tasks, improving the response speed and efficiency of the service. The function of dynamically adjusting task priorities according to the evaluation results enables the system to flexibly respond to changes in demand in different scenarios, improving the adaptability and flexibility of the system.

[0055] Among them, the multi-task intelligent coordination execution system based on the elderly care robot includes an integrated sensor, a processor, a memory, a processor, an actuator and a communication module. The memory is used to store the system program, and the processor is used to read the system program in the memory. When the actuator executes the system program, the multi-task intelligent coordination execution method of the elderly care robot described in one of claims 1 to 8 is implemented.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] 1. In order to solve the problems of inaccurate demand identification, task priority flexibility and task execution efficiency of existing elderly care robots, and improve accuracy, flexibility and efficiency, the present invention integrates advanced data fusion, behavior pattern recognition and prediction, demand identification and task evaluation, significantly improves the intelligence level and service quality of the system, and realizes more accurate behavior pattern recognition and trend prediction through membership function and sliding window prediction; combines behavior pattern and trend prediction results to generate tasks that meet user needs, and sets priorities through multi-dimensional evaluation models to ensure the efficiency and rationality of task execution;

[0058] 2. The present invention can accurately predict the user's behavior trend by integrating the collected information through analysis and processing, and using the membership function and sliding window prediction algorithm. Combining the results of behavior pattern recognition and behavior trend prediction, it can identify the user's current specific needs, ensure that the generated tasks are closer to the user's actual needs, and improve the relevance of the tasks;

[0059] 3. The present invention reduces the need for manual intervention through the process of automatically generating tasks, enables the system to respond to user needs more efficiently, improves the immediacy and convenience of services, and can provide more accurate services by intelligently identifying needs and generating corresponding tasks, avoiding invalid or redundant operations, and improving the quality and efficiency of services;

[0060] 4. The present invention takes into account multiple aspects such as urgency, importance, security, resource consumption and user experience through a multi-dimensional evaluation model, making task evaluation more comprehensive and objective, avoiding the bias that may be caused by a single standard, setting the priority of tasks according to the comprehensive evaluation score, ensuring that important and urgent tasks are executed first, improving the rationality and effectiveness of task execution, and dynamically adjusting the priority according to the task execution status, so that the system can flexibly respond to emergencies or changes in task execution, thereby enhancing the system's resilience. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 The operation flow of the multi-task intelligent coordination execution method based on the elderly care robot of the present invention is as follows Figure 1 ;

[0062] Figure 2 The operation flow of the multi-task intelligent coordination execution method based on the elderly care robot of the present invention is as follows Figure 2 ;

[0063] Figure 3 The operation flow of the multi-task intelligent coordination execution method based on the elderly care robot of the present invention is as follows Figure 3 ;

[0064] Figure 4 It is a structural schematic diagram of the multi-task intelligent coordinated execution system based on the elderly care robot of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments 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.

[0066] Example

[0067] See also Figure 1-Figure 4 As shown, the present invention provides a technical solution: comprising the following steps:

[0068] S1, the system is initialized according to the current environment and preset parameters;

[0069] S2, collects information about the user and his / her surroundings through integrated sensors;

[0070] S3: Perform fusion analysis and processing based on the collected information, predict the user's behavior trend, identify the user's current needs based on the data analysis results and prediction results, and automatically generate tasks;

[0071] S4. Evaluate the generated tasks through a multi-dimensional evaluation model and set priorities;

[0072] S5. Decompose complex tasks into multiple small tasks and optimize the order of tasks;

[0073] S6. Record the results of each task execution and dynamically adjust the model and system parameters based on the feedback information.

[0074] Among them, in S1, the system is initialized according to the current state and preset parameters, obtains the user's current environmental information through integrated sensors, initializes the system parameters according to the user's historical data requirements and preferences, and dynamically updates the system parameters through real-time data and user feedback.

[0075] It not only makes adaptive adjustments based on the current environment, but also integrates the user's historical data and preferences to provide a precise starting point for subsequent task execution. This comprehensive initialization process ensures that the system can quickly and accurately identify and adapt to different elderly care environments, providing more intimate and personalized services for the elderly. At the same time, through dynamic updates of real-time data and user feedback, the system can continuously optimize its parameter settings to better meet the changing needs of users.

[0076] Among them, S3 performs fusion analysis and processing based on the collected information, predicts the user's behavior trend, identifies the user's current needs based on the data analysis results and prediction results, and automatically generates tasks; let the sensor data obtained by S2 be r i (t), assign a trust weight w to the data of each sensor i i (t), the data is fused into F(t), and the implementation formula is:

[0077]

[0078] In the formula, F(t) represents the fusion data at time t, n represents the number of sensors, and w i (t) represents the trust weight of the i-th sensor at time t, r i (t) represents the data correlation coefficient of the i-th sensor at time t, S i (t) represents the data value of the i-th sensor at time t, w i(t) According to the historical data of the sensor, the deterministic dynamic adjustment is made and the trust weight is updated as follows:

[0079] w i (t+1)=w i (t)·exp(-αe i (t)),

[0080] In the formula, α represents the new task adjustment rate constant, e i (t) represents the error of the i-th sensor at time t, r i (t) represents the data correlation coefficient of the i-th sensor at time t, which is implemented as:

[0081]

[0082] In the formula, p ij (t) represents the correlation between the i-th sensor and the j-th sensor at time t;

[0083] Feature extraction is performed based on the fused data, and dynamic adjustments are made based on the user's current behavior pattern and current environment. The feature extraction implementation formula is:

[0084] φ(F(t))=F(t)·A(t),

[0085] In the formula, φ(F(t)) represents the extracted feature vector, F(t) represents the fused data, A(t) represents the dynamic feature weight matrix at time t, and each element a of the feature weight matrix A(t) is ij (t) represents the importance coefficient of the i-th feature in the j-th behavior mode, expressed as:

[0086]

[0087] In the formula, b ij (t) represents the unnormalized importance coefficient of the i-th feature in the j-th behavior mode at time t.

[0088] Wherein, S3 identifies the user's behavior pattern according to the extracted feature vector, and assumes that each behavior pattern P i There is a eigenvector, let the membership function μ i (φ(F(t))) indicates that the feature vector φ(F(t)) belongs to the behavior pattern P i The final behavior pattern recognition formula is:

[0089]

[0090] In the formula, B(t) represents the behavior pattern recognition result at time t, P iThe characteristic vector representing the i-th behavior mode, μ i (φ(F(t))) represents the membership function of the behavior mode in the ith order, expressed as:

[0091]

[0092] In the formula, θ i represents the fuzziness parameter of the i-th behavior mode, || || 2 Indicates calculating the Euclidean distance between two vectors;

[0093] Predict the user's behavior data at the next moment t+1 based on historical data The behavior data B(tk) of the past K time steps is averaged by adaptive weights to obtain the predicted value. The implementation formula is:

[0094]

[0095] In the formula, represents the behavior trend prediction result at time t+1, K represents the sliding window size, and λ k (t) represents the weight factor of the kth time step at time t, B(tk) represents the behavior pattern recognition result at time tk, λ k (t) By dynamically adjusting the weight according to the historical data error, the formula is:

[0096]

[0097] In the formula, γ represents the decay rate of the weight factor, and exp represents the exponential function.

[0098] Among them, the S3 performs comprehensive demand identification based on the identified behavior patterns and behavior trend prediction results, and the implementation formula is:

[0099]

[0100] In the formula, D(t) represents the demand identification result at time t, B i (t) represents the behavior pattern recognition result at time t, represents the behavior trend prediction result at time t+1, ψ i Represents the demand type weight vector corresponding to the i-th behavior mode;

[0101] According to the identified user needs, tasks are generated and the implementation formula is:

[0102]

[0103] In the formula, T represents the generated task, T poolrepresents the candidate task pool, D i (t) represents the demand intensity of the ith demand type at time t, ψ i Indicates the importance coefficient of the requirement type.

[0104] By integrating and analyzing the collected information and predicting the user's behavior trends, the system not only improves the system's recognition accuracy of user needs and enables the system to predict and meet users' potential needs in advance, but also continuously optimizes its prediction model by dynamically adjusting parameters such as feature weight matrix and membership function to adapt to different environments and changes in user behavior, thereby improving the timeliness and effectiveness of the service.

[0105] Among them, S4 evaluates the generated tasks through a multi-dimensional evaluation model and sets priorities; the evaluation dimensions include urgency U, importance I, security S, resource consumption R and user experience X, and a multi-dimensional evaluation model is constructed according to the evaluation dimensions to comprehensively consider the evaluation results of each dimension. The implementation formula is:

[0106]

[0107] In the formula, E(T) represents the comprehensive evaluation score of task T, U(T) represents the urgency score of task T, I(T) represents the importance score of task T, S(T) represents the security score of task T, R(T) represents the resource occupation score of task T, and X(T) represents the user experience score of task T.

[0108] Among them, S4 sets the priority of the task according to the comprehensive evaluation score, and the implementation formula is:

[0109]

[0110] In the formula, P(T) represents the standardized priority score of task T, min(E) represents the minimum value of all task evaluation scores, and max(E) represents the maximum value of all task evaluation scores. Tasks are sorted according to the standardized evaluation scores, which can be implemented as follows:

[0111]

[0112] In the formula, T sd Represents a list of tasks sorted from high to low priority, T pl Represents the candidate task pool; dynamically adjusts the priority according to the actual situation and task execution status, and is implemented as follows:

[0113] P′(T)=P(T)+ΔP(T),

[0114] In the formula, P′(T) represents the adjusted priority score, and ΔP(T) represents the priority increment adjusted according to the task execution status.

[0115] Among them, S6 records the results of each task execution, and dynamically adjusts the model and system parameters according to the feedback information, records the user's initial state before the task execution and the basic information of the task, including but not limited to physiological indicators, psychological state, and environmental conditions, records the real-time data generated during the task execution and the task execution results, and collects user feedback information and execution results for adjustment and update.

[0116] The multi-dimensional evaluation model enables the system to comprehensively consider multiple dimensions such as urgency, importance, security, resource consumption and user experience to evaluate the generated tasks. The comprehensive and detailed evaluation method ensures that the system can reasonably allocate resources and give priority to key tasks while avoiding waste and redundancy of resources. By setting priorities, the system can efficiently and orderly execute tasks, improving the response speed and efficiency of the service. The function of dynamically adjusting task priorities according to the evaluation results enables the system to flexibly respond to changes in demand in different scenarios, improving the adaptability and flexibility of the system.

[0117] Among them, the multi-task intelligent coordination execution system based on the elderly care robot includes an integrated sensor, a processor, a memory, a processor, an actuator and a communication module. The memory is used to store the system program, and the processor is used to read the system program in the memory. When the actuator executes the system program, the multi-task intelligent coordination execution method of the elderly care robot described in one of claims 1 to 8 is implemented.

[0118] Working principle: By obtaining the user's current environment information through integrated sensors, initializing system parameters according to the user's historical data needs and preferences, and dynamically updating system parameters according to real-time data and user's immediate feedback, adapting to the user's current status and needs, integrating multiple sensors to collect information about the user and his or her surrounding environment in real time, and storing the collected information in the database;

[0119] Through the collected data information, a trust weight and data correlation coefficient are assigned to the data of each sensor, and fusion processing is performed to obtain comprehensive data. The trust weight is dynamically adjusted according to the historical data of the sensor, and the data correlation coefficient is calculated through the correlation between sensors. Feature extraction is performed based on the fused data, and the feature weight matrix is ​​dynamically adjusted according to the user's current behavior pattern and environment. The user's behavior pattern is identified through the membership function, and the membership is calculated based on the feature vector. The user's behavior trend at the next moment is predicted using the sliding window and adaptive weight. The comprehensive demand identification is performed by combining the behavior pattern recognition and behavior trend prediction results. According to the identified needs, tasks are automatically generated from the candidate task pool; the evaluation dimensions are defined, including urgency, importance, security, resource consumption and user experience, and a multi-dimensional evaluation model is constructed. Consider the evaluation results of each dimension to obtain a comprehensive evaluation score, set the priority of the task through standardized processing according to the comprehensive evaluation score, sort the tasks according to the standardized evaluation score, and obtain a task list with high to low priority. Dynamically adjust the priority according to the task execution status, decompose complex tasks into multiple simple subtasks, optimize the order of task execution according to the dependencies and priorities between tasks, and record the results of each task execution, including the user's initial state before task execution, basic task information, real-time data generated during task execution, task execution results, etc. Based on the recorded results of each task execution, including the user's initial state before task execution, basic task information, real-time data generated during task execution, task execution results, etc., dynamically adjust the model and system parameters according to the recorded execution results and user feedback information.

[0120] 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.

[0121] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A multi-task intelligent coordination execution method based on an elderly care robot, characterized in that: The following steps are involved: S1, the system is initialized according to the current environment and preset parameters; S2, collects information about the user and his / her surroundings through integrated sensors; S3: Perform fusion analysis and processing based on the collected information, predict the user's behavior trend, identify the user's current needs based on the data analysis results and prediction results, and automatically generate tasks; S4. Evaluate the generated tasks through a multi-dimensional evaluation model and set priorities; S5. Decompose complex tasks into multiple small tasks and optimize the order of tasks; S6. Record the results of each task execution and dynamically adjust the model and system parameters based on the feedback information.

2. The multi-task intelligent coordination execution method based on the elderly care robot according to claim 1 is characterized in that: In S1, the system is initialized according to the current state and preset parameters, obtains the user's current environment information through integrated sensors, initializes the system parameters according to the user's historical data requirements and preferences, and dynamically updates the system parameters through real-time data and user feedback.

3. The multi-task intelligent coordination execution method based on the elderly care robot according to claim 1 is characterized in that: S3, performs fusion analysis and processing based on the collected information, predicts the user's behavior trend, identifies the user's current needs based on the data analysis results and prediction results, and automatically generates tasks; set up The sensor data obtained by S2 is r i (t), assign a trust weight w to the data of each sensor i i (t), the data is fused into F(t), and the implementation formula is: In the formula, F(t) represents the fusion data at time t, n represents the number of sensors, and w i (t) represents the trust weight of the i-th sensor at time t, r i (t) represents the data correlation coefficient of the i-th sensor at time t, S i (t) represents the data value of the i-th sensor at time t, w i (t) According to the historical data of the sensor, the deterministic dynamic adjustment is made and the trust weight is updated as follows: w i (t+1)=w i (t)·exp(-αe i (t)), In the formula, α represents the new task adjustment rate constant, e i (t) represents the error of the i-th sensor at time t, r i (t) represents the data correlation coefficient of the i-th sensor at time t, which is implemented as: In the formula, p ij ( t ) represents the correlation between the i-th sensor and the j-th sensor at time t; Feature extraction is performed based on the fused data, and dynamic adjustments are made based on the user's current behavior pattern and current environment. The feature extraction implementation formula is: φ(F(t))=F(t)·A(t), In the formula, φ(F(t)) represents the extracted feature vector, F(t) represents the fused data, A(t) represents the dynamic feature weight matrix at time t, and each element a of the feature weight matrix A(t) is ij (t) represents the importance coefficient of the i-th feature in the j-th behavior mode, expressed as: In the formula, b ij (t) represents the unnormalized importance coefficient of the i-th feature in the j-th behavior mode at time t.

4. The multi-task intelligent coordination execution method based on the elderly care robot according to claim 3 is characterized by: S3, identifying the user's behavior pattern according to the extracted feature vector, assuming that each behavior pattern P i There is a eigenvector, let the membership function μ i (φ(F(t))) indicates that the feature vector φ(F(t)) belongs to the behavior pattern P i The final behavior pattern recognition formula is: In the formula, B(t) represents the behavior pattern recognition result at time t, P i The characteristic vector representing the i-th behavior mode, μ i (φ(F(t))) represents the membership function of the behavior mode in the ith order, expressed as: In the formula, θ i represents the fuzziness parameter of the i-th behavior mode, ||||2 represents the calculation of the Euclidean distance between two vectors; Predict the user's behavior data at the next moment t+1 based on historical data The behavior data B(tk) of the past K time steps is averaged by adaptive weights to obtain the predicted value. The implementation formula is: In the formula, represents the behavior trend prediction result at time t+1, K represents the sliding window size, and λ k (t) represents the weight factor of the kth time step at time t, B(tk) represents the behavior pattern recognition result at time tk, λ k (t) By dynamically adjusting the weight according to the historical data error, the formula is: In the formula, γ represents the decay rate of the weight factor, and exp represents the exponential function.

5. The multi-task intelligent coordination execution method based on the elderly care robot according to claim 4 is characterized in that: S3, based on the identified behavior patterns and behavior trend prediction results, performs comprehensive demand identification, and the implementation formula is: In the formula, D(t) represents the demand identification result at time t, B i (t) represents the behavior pattern recognition result at time t, represents the behavior trend prediction result at time t+1, ψ i Represents the demand type weight vector corresponding to the i-th behavior mode; According to the identified user needs, tasks are generated, and the implementation formula is: In the formula, T represents the generated task, T pool represents the candidate task pool, D i (t) represents the demand intensity of the i-th demand type at time t, ψ i Indicates the importance coefficient of the requirement type.

6. The multi-task intelligent coordination execution method based on the elderly care robot according to claim 1 is characterized in that: S4, evaluates the generated tasks through a multi-dimensional evaluation model and sets priorities; the evaluation dimensions include urgency U, importance I, security S, resource consumption R and user experience X, and a multi-dimensional evaluation model is constructed according to the evaluation dimensions to comprehensively consider the evaluation results of each dimension, and the implementation formula is: In the formula, E(T) represents the comprehensive evaluation score of task T, U(T) represents the urgency score of task T, I(T) represents the importance score of task T, S(T) represents the security score of task T, R(T) represents the resource occupation score of task T, and X(T) represents the user experience score of task T.

7. The multi-task intelligent coordination execution method based on the elderly care robot according to claim 6 is characterized in that: S4 sets the priority of the task according to the comprehensive evaluation score, and the implementation formula is: In the formula, P(T) represents the standardized priority score of task T, min(E) represents the minimum value of all task evaluation scores, and max(E) represents the maximum value of all task evaluation scores. Tasks are sorted according to the standardized evaluation scores, which can be implemented as follows: In the formula, T sd Represents a list of tasks sorted from high to low priority, T pl Represents the candidate task pool; dynamically adjusts the priority according to the actual situation and task execution status, and is implemented as follows: P′(T)=P(T)+ΔP(T), In the formula, P′(T) represents the adjusted priority score, and ΔP(T) represents the priority increment adjusted according to the task execution status.

8. The multi-task intelligent coordination execution method based on the elderly care robot according to claim 1 is characterized by: The S6 records the results of each task execution, and dynamically adjusts the model and system parameters according to the feedback information, records the user's initial state before the task execution and the basic information of the task, including but not limited to physiological indicators, psychological state, and environmental conditions, records the real-time data generated during the task execution and the task execution results, and collects the user's feedback information and execution results for adjustment and update.

9. A multi-task intelligent coordination execution system based on an elderly care robot, comprising an integrated sensor, a processor, a memory, an actuator and a communication module, wherein the memory is used to store a system program, the processor is used to read the system program in the memory, and when the actuator executes the system program, the multi-task intelligent coordination execution method of the elderly care robot described in one of claims 1 to 8 is implemented.

Images