Robot data processing method and device, intelligent robot

Abstract

The present disclosure provides a robot data processing method and device, and an intelligent robot, belonging to the field of data processing technology. The method includes: fusing first data and second data to obtain first feature data, the first data is the user's vital sign monitoring data, and the second data is the user's basic information; extracting features from third data to obtain second feature data; the third data is the user's motor function monitoring data; extracting features from fourth data to obtain third feature data; the fourth data is the user's body composition monitoring data; searching a set strategy library based on the first feature data, the second feature data, and the third feature data, and determining the robot's target control strategy based on the search result; the target control strategy is used to control the robot. The robot data processing method and device, and the intelligent robot provided by the present disclosure can improve the accuracy of the intelligent robot's output decision.

Description

Technical Field

[0001] The present disclosure belongs to the field of data processing technology, and more specifically, relates to a robot data processing method and device, and an intelligent robot. Background Art

[0002] With the development of artificial intelligence technology, more and more nursing homes are using smart robots to assist in the care of the elderly. Smart robots can record the elderly’s daily living information, health data, nursing conditions, etc. in real time, and formulate daily rehabilitation plans for the elderly, including dietary recommendations, exercise plans, etc.

[0003] Due to differences in physiological conditions, psychological states, life experiences, economic conditions and social environments, the needs of the elderly are diverse. In order to better meet the diverse needs of the elderly, it is necessary to further improve the accuracy of the output decisions of intelligent robots. Summary of the invention

[0004] The purpose of the present disclosure is to provide a robot data processing method and device, and an intelligent robot, so as to improve the accuracy of the output decision of the intelligent robot.

[0005] A first aspect of the embodiments of the present disclosure provides a robot data processing method, comprising:

[0006] Fusing the first data with the second data to obtain first feature data, wherein the first data is the vital sign monitoring data of the user, and the second data is the basic information of the user;

[0007] Extracting features from the third data to obtain second feature data; the third data is the motor function monitoring data of the user;

[0008] Extracting features from the fourth data to obtain third feature data; the fourth data is the body composition monitoring data of the user;

[0009] A set strategy library is searched based on the first feature data, the second feature data and the third feature data, and a target control strategy of the robot is determined based on the search result; the target control strategy is used to control the robot.

[0010] A second aspect of the embodiments of the present disclosure provides a robot data processing device, including:

[0011] A first feature extraction module is used to fuse the first data and the second data to obtain first feature data, wherein the first data is the vital sign monitoring data of the user, and the second data is the basic information of the user;

[0012] A second feature extraction module is used to extract features from the third data to obtain second feature data; the third data is the motor function monitoring data of the user;

[0013] A third feature extraction module is used to extract features from the fourth data to obtain third feature data; the fourth data is the body composition monitoring data of the user;

[0014] A data matching module is used to search a set strategy library based on the first feature data, the second feature data and the third feature data, and determine a target control strategy of the robot based on the search result; the target control strategy is used to control the robot.

[0015] According to a third aspect of an embodiment of the present disclosure, an intelligent robot is provided, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the above-mentioned robot data processing method when executing the computer program.

[0016] According to a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned robot data processing method are implemented.

[0017] The robot data processing method and device, and the intelligent robot provided by the embodiments of the present disclosure have the following beneficial effects:

[0018] In the disclosed embodiment, by integrating the vital sign monitoring data with the basic information of the user, the first characteristic data obtained can accurately reflect the vital signs of the user; the second characteristic data can reflect the motor ability of the user's limbs; and the third characteristic data can reflect the metabolic level and nutritional status of the user. By searching the pre-built strategy library based on the first characteristic data, the second characteristic data, and the third characteristic data, it is possible to quickly find a control strategy suitable for the user's current condition, achieve accurate decision-making, and provide personalized services for the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 A schematic diagram of a flow chart of a robot data processing method provided by an embodiment of the present disclosure;

[0021] Figure 2A structural block diagram of a robot data processing device provided by an embodiment of the present disclosure;

[0022] Figure 3 A schematic block diagram of an intelligent robot provided in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0024] In order to make the purpose, technical solutions and advantages of the present disclosure more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0025] Please refer to Figure 1 , Figure 1 A schematic diagram of a flow chart of a robot data processing method provided by an embodiment of the present disclosure, the method comprising:

[0026] S101: Fusing first data and second data to obtain first feature data, where the first data is the user's vital sign monitoring data and the second data is the user's basic information.

[0027] In this embodiment, the user may be an elderly person in a nursing home, and the user's vital sign monitoring data may include heart rate, blood pressure, body temperature, etc. The user's basic information includes age, gender, medical history, etc. Considering that different users may have different judgment criteria for their vital sign monitoring data, for example, the same heart rate value may be judged as abnormal for a healthy user, but may be normal for a user with heart disease. Therefore, by fusing the user's vital sign monitoring data with the user's basic information, the first feature data obtained can accurately reflect the user's vital signs and provide richer data information for subsequent analysis.

[0028] Specifically, the existing data fusion method can be used to fuse the first data and the second data. For example, rules can be set based on prior knowledge and experience to integrate data from different sources; or the feature splicing method in machine learning can be used to splice the data features of the vital sign monitoring data and the basic information in terms of dimensions to form the first feature data.

[0029] S102: Extract features from the third data to obtain second feature data; the third data is the user's motor function monitoring data.

[0030] In this embodiment, the motor function monitoring data may include motor ability data, motor coordination data, joint mobility data, etc. Key features are extracted from the motor function monitoring data, such as muscle strength data, movement speed, muscle endurance and other features related to the motor ability data, static balance, dynamic balance and other features related to the motor coordination data, shoulder joint mobility, knee joint mobility, spinal flexibility, wrist joint flexibility and other features related to the joint mobility data, and the above feature data are spliced ​​to obtain the second feature data.

[0031] S103: Extract features from the fourth data to obtain third feature data; the fourth data is the user's body composition monitoring data.

[0032] In this embodiment, the body composition monitoring data can reflect the user's metabolic level and nutritional status. By extracting key features from the body composition monitoring data, such as body fat percentage, muscle mass, water content, bone density, etc., the third feature data can be obtained.

[0033] S104: searching a set strategy library based on the first feature data, the second feature data and the third feature data, and determining a target control strategy of the robot based on the search result; the target control strategy is used to control the robot.

[0034] In this embodiment, the robot is an intelligent robot installed in a nursing home, and a strategy library can be pre-built, in which a first mapping relationship between various feature data combinations and robot control strategies is stored. By searching the strategy library based on the first feature data, the second feature data, and the third feature data, a matching robot control strategy can be quickly found.

[0035] For example, in a rehabilitation scenario, if the first feature data of a user indicates that he is older and has a history of cardiovascular disease, the second feature data shows that his motor function recovers slowly, and the third feature data shows that his body fat percentage is high, by searching the strategy library, the control strategies such as the robot rehabilitation training intensity, auxiliary methods, and dietary recommendations suitable for the user can be determined, thereby providing users with accurate and effective services.

[0036] The control parameters of the robot can be adjusted according to the above target control strategy. For example, when user A uses the robot for rehabilitation training, the rehabilitation training intensity and auxiliary method included in the target control strategy can be determined as the working parameters of the rehabilitation equipment to provide targeted rehabilitation training for the user.

[0037] The above target control strategy can be sent to the mobile phone of the guardian, who can monitor the user's execution according to the target control strategy, such as monitoring whether the patient eats according to the prescribed diet plan, whether the rehabilitation training is carried out on time, etc., to ensure the effective implementation of the target control strategy.

[0038] From the above, it can be concluded that the first characteristic data obtained by integrating the vital sign monitoring data with the basic information of the user in this embodiment can accurately reflect the vital signs of the user; the second characteristic data can reflect the motor ability of the user's limbs; and the third characteristic data can reflect the metabolic level and nutritional status of the user. By searching the pre-built strategy library based on the first characteristic data, the second characteristic data, and the third characteristic data, it is possible to quickly find a control strategy suitable for the user's current condition, achieve accurate decision-making, and provide personalized services for the user.

[0039] Furthermore, an image acquisition device may be provided to acquire a tongue image of the user, and the tongue image may be input into an image recognition model to obtain fifth data, which may represent the user's physical condition from a TCM perspective.

[0040] When training the image recognition model, a large number of users' tongue images can be collected, and Chinese medicine experts can annotate the tongue images. The annotated tongue images are used as training sets to train neural network models (such as convolutional neural networks, etc.). The loss function of the model on the training set is minimized by adjusting the parameters, and finally the trained image recognition model is obtained through testing, evaluation and optimization.

[0041] In one embodiment of the present disclosure, performing data fusion on the first data and the second data includes:

[0042] Determine a threshold value corresponding to each first data based on the second data;

[0043] Compare each first data with a corresponding threshold value to obtain evaluation data corresponding to each first data;

[0044] The first data and the evaluation data are concatenated to obtain first characteristic data.

[0045] In this embodiment, considering that different basic information (such as age, gender, and medical history) of individuals will affect the normal range of their vital sign data, the threshold corresponding to the vital sign monitoring data (referred to as the vital sign data threshold) is determined based on the basic information of the user, so that the subsequent evaluation of the vital sign monitoring data can be more in line with the actual situation of the individual user.

[0046] Specifically, the second mapping relationship between each dimension of basic information and the threshold value of vital sign data can be pre-constructed. For example, according to fields such as age, gender and common disease types, the reasonable threshold range of corresponding vital signs such as heart rate and blood pressure can be statistically obtained from existing large-scale medical data. On this basis, the second mapping relationship can be searched according to the input user basic information to obtain the corresponding threshold value of vital sign data.

[0047] By comparing the actually monitored vital sign data with the vital sign data threshold, it is possible to evaluate whether each vital sign monitoring data is within the normal range, thereby generating corresponding evaluation data (such as "normal" or "abnormal"). The above evaluation data helps to quickly determine the user's current health status.

[0048] By combining the vital sign monitoring data with the evaluation data obtained based on the threshold comparison, the first feature data with greater analytical value can be formed. The above fusion method not only retains the detailed information of the vital sign monitoring data, but also incorporates the evaluation results of the vital sign monitoring data, providing a more comprehensive basis for subsequent analysis and decision-making.

[0049] From the above, it can be concluded that this embodiment determines the vital sign data threshold based on the user's basic information, compares the vital sign monitoring data with the vital sign data threshold to obtain evaluation data, and splices the vital sign monitoring data with the evaluation data to form the first feature data which can accurately reflect the user's vital signs.

[0050] In one embodiment of the present disclosure, the robot data processing method further includes:

[0051] determining a motion monitoring item for the user based on the second data;

[0052] The third data is collected according to the motion monitoring project.

[0053] In this embodiment, considering the different conditions of the users, the corresponding motion monitoring items will be different. For example, for stroke patients, it is necessary to focus on testing the motor function and balance ability of the limbs; for orthopedic surgery patients, it is necessary to focus on measuring the recovery of joint mobility and muscle strength.

[0054] Therefore, a third mapping relationship between the user's basic information (specifically, disease type) and the exercise monitoring items can be pre-constructed, and on this basis, the third mapping relationship can be searched according to the input user basic information to obtain the corresponding exercise monitoring items.

[0055] Collecting the motor function monitoring data (third data) according to the above-mentioned sports monitoring items can ensure that the collected data is targeted and effective.

[0056] It can be concluded from the above that the present embodiment collects the third data according to the specific motion monitoring items, which can avoid collecting a large amount of irrelevant or redundant data and improve the efficiency and quality of data collection. At the same time, this targeted data collection method can better support the subsequent data processing and analysis work, so as to make the target control strategy more accurate.

[0057] In one embodiment of the present disclosure, searching a set strategy library based on the first feature data, the second feature data, and the third feature data, and determining a control strategy of the robot based on the search result, includes:

[0058] Fusing the first feature data, the second feature data, and the third feature data to obtain a first feature vector;

[0059] The distances between the first feature vector and the M second feature vectors in the strategy library are calculated respectively; the strategy library includes M classified sample data, the M second feature vectors are feature vectors corresponding to the M first sample data, and the M first sample data are sample data selected from the M classified sample data respectively;

[0060] Select K first sample data whose corresponding distance is less than the distance threshold; where M and K are both natural numbers, and K <M;

[0061] The target control strategy of the robot is determined based on the control strategies corresponding to the K sample data.

[0062] In this embodiment, by vector splicing the first feature data, the second feature data and the third feature data, a first feature vector can be obtained. The first feature vector integrates different feature data and can provide a comprehensive information basis for determining the target control strategy, so that the target control strategy matches the actual needs of the user.

[0063] Correspondingly, each sample data in the strategy library contains historical data of the first feature data, historical data of the second feature data, historical data of the third feature data and historical data of the control strategy. By performing vector splicing on the historical data of the first feature data, the historical data of the second feature data and the historical data of the third feature data, a third feature vector can be obtained. By using the same method for vector splicing, multiple third feature vectors corresponding to multiple sample data in the strategy library can be obtained.

[0064] On this basis, the distances between the first eigenvector and multiple third eigenvectors can be calculated respectively. The closer the distances are, the higher the matching degree between the first eigenvector and the third eigenvector. Therefore, the K third eigenvectors with the nearest distances are selected, and the target control strategy of the robot is determined based on the control strategies corresponding to the K third eigenvectors.

[0065] Furthermore, considering that the amount of sample data in the strategy library is large, the amount of data of the corresponding third eigenvector is large. In order to reduce the amount of calculation of the distance calculation, this embodiment classifies the sample data in the strategy library to obtain M classified sample data, and then randomly selects one sample data from each classified sample data as the first sample data to obtain M first sample data. M third eigenvectors can be obtained based on the M first sample data, and the M third eigenvectors are recorded as second eigenvectors. The distance between the first eigenvector and the M second eigenvectors is calculated respectively, and the K second eigenvectors with the closest distance are selected accordingly. The target control strategy of the robot is determined based on the control strategy corresponding to the K second eigenvectors. Since the number of second eigenvectors is much smaller than the number of third eigenvectors, the above method can reduce the amount of calculation of the distance calculation process. Among them, the classification of sample data in the strategy library can be implemented based on existing decision trees, support vector machines, index structure construction and other methods.

[0066] Exemplarily, determining the target control strategy of the robot based on the control strategy corresponding to the K second eigenvectors can be described in detail as follows: For classification problems in the target control strategy, for example, when determining a specific rehabilitation method, the number of times the rehabilitation methods corresponding to the K second eigenvectors appear can be counted, and the method with the largest number of appearances can be determined as the rehabilitation method in the target control strategy. For numerical problems in the target control strategy, for example, when determining the intensity of rehabilitation training, different weights can be assigned to the training intensities corresponding to the K second eigenvectors according to the distance between the K second eigenvectors and the first eigenvector. The closer the distance, the greater the weight. ,in, represents the weight, represents the distance, and then the training intensities corresponding to the K second eigenvectors are weighted summed to obtain the training intensity in the target control strategy.

[0067] It can be concluded from the above that the method of this embodiment realizes the accurate search of the target control strategy. At the same time, by classifying the sample data in the strategy library, M representative second eigenvectors are obtained, and then the distance between the first eigenvector and the M second eigenvectors is calculated, and the K second eigenvectors with the nearest distance are selected. The target control strategy of the robot is determined based on the control strategy corresponding to the K second eigenvectors, which reduces the amount of data calculation while ensuring the search accuracy.

[0068] In one embodiment of the present disclosure, the robot data processing method further includes:

[0069] Build an index structure for N sample data in the strategy library;

[0070] Based on the index structure, N sample data in the strategy library are classified to obtain M classified sample data, where N is a natural number and N>M.

[0071] In this embodiment, a specific implementation method for classifying sample data in the policy library is provided. Specifically, an index structure, such as a tree index (such as a KD tree, a ball tree, etc.), a hash index, etc., can be constructed for the sample data in the policy library.

[0072] Taking tree index as an example, the tree structure includes nodes of different levels. According to actual needs, the nodes of the set level can be used as the classification of sample data in the policy library. For example, the nodes of the last level, that is, the leaf nodes, are used as the M sample classifications of the policy library.

[0073] It can be concluded from the above that this embodiment classifies sample data through an index structure, can deeply explore the internal connections and rules between data, can ensure that data with similar characteristics are classified into the same category, and is conducive to improving the accuracy of classification.

[0074] In one embodiment of the present disclosure, an index structure is constructed for N sample data in a policy library, including:

[0075] The sample data partitioning operation is performed multiple times until the stopping condition is met to obtain an index structure;

[0076] The sample data partitioning operations include:

[0077] Calculate the correlation coefficients between each dimension data in the target subtree and other dimension data respectively; wherein any dimension data is a collection of data of multiple sample data in the target subtree in the dimension;

[0078] The dimension with the lowest corresponding correlation coefficient is determined as the split axis, and the target subtree is divided into two left and right subtrees;

[0079] When the sample data partitioning operation is performed for the first time, the policy library is determined as the target subtree, and when the sample data partitioning operation is performed for other times, the left and right subtrees are respectively determined as the target subtrees;

[0080] The stopping condition is: the number of sample data in the target subtree is less than the set number.

[0081] In this embodiment, a KD tree is specifically used to construct an index structure for the sample data in the strategy library. According to the segmentation dimension of the root node, the data points (i.e., sample data) are divided into left and right parts. Data points smaller than the root node are divided into the left subtree, and data points greater than or equal to the root node are divided into the right subtree. Each sample data has multiple dimensions. The data of multiple sample data in the same dimension are combined as data of the corresponding dimension to obtain data of each dimension, calculate the correlation coefficient between the data of each dimension, and select the dimension with the lowest correlation coefficient as the segmentation axis. At the same time, the median of all data points on the segmentation dimension is used as the data of the root node.

[0082] Repeat the above steps for the left and right subtrees, that is, select new split dimensions and medians in their respective data sets as the root nodes of the subtrees, and continue to divide the left and right subtrees until the stopping condition is met (that is, the number of sample data in the target subtree is less than the set number), and the index structure of the strategy library can be obtained.

[0083] The process of determining the dimension with the lowest correlation coefficient includes: calculating the correlation coefficient between any dimension data and other dimension data respectively to obtain multiple correlation coefficients corresponding to the any dimension data, calculating the average of the absolute values ​​of the multiple correlation coefficients corresponding to the any dimension data, and obtaining the correlation coefficient corresponding to the any dimension data. The same method can be used to calculate the correlation coefficient corresponding to each dimension data, compare the correlation coefficients corresponding to each dimension data, and determine the dimension with the lowest correlation coefficient as the dimension with the lowest correlation coefficient.

[0084] From the above, it can be concluded that in the process of constructing the KD tree in multiple iterations, this embodiment selects a dimension with low correlation with other dimensions as the splitting axis, which can avoid excessive overlap between the dimension corresponding to the selected splitting axis and the divided dimension information, thereby better dividing the sample data and making the structure of the KD tree better.

[0085] In one embodiment of the present disclosure, calculating the distance between the first feature vector and the second feature vector corresponding to each sample data in the strategy library includes:

[0086] If the basic information of the user shows that the health status of the user is a first status, the distance between the first feature vector and a plurality of second feature vectors in the strategy library is calculated based on the Manhattan distance; the first status is used to indicate that the user suffers from a chronic disease;

[0087] If the basic information of the user shows that the health status of the user is the second status, the distance between the first feature vector and a plurality of second feature vectors in the strategy library is calculated based on the Euclidean distance; the second status is used to indicate that the user is in good health.

[0088] In this embodiment, the distance between the first feature vector and the plurality of second feature vectors in the strategy library may be calculated using existing calculation methods such as Euclidean distance, Manhattan distance, and cosine similarity.

[0089] Considering that elderly people with chronic diseases often involve multiple interrelated factors, such as elderly people with diabetes, their vital signs such as blood sugar and blood pressure, as well as body composition such as insulin resistance and muscle fat ratio, the above factors affect each other. At this time, the Manhattan distance is used to calculate the distance between the first eigenvector and multiple second eigenvectors in the strategy library, and the differences in each dimension can be calculated and accumulated separately, which can take into account the differences between these factors more comprehensively.

[0090] At the same time, considering that the vital signs, motor function, body composition and other characteristics of the elderly without obvious diseases are relatively stable and independent, at this time, the Euclidean distance is used to calculate the distance between the first eigenvector and multiple second eigenvectors in the strategy library. This can comprehensively consider the differences in all feature dimensions. By calculating the square of the difference in each dimension and then taking the square root, it can better reflect the overall similarity between the data, which is conducive to finding the most similar sample data.

[0091] It can be concluded from the above that this embodiment selects a suitable distance calculation method according to actual conditions, which can more effectively mine potential information in the data and provide a more reliable basis for subsequent decision-making.

[0092] In one embodiment of the present disclosure, the robot data processing method further includes:

[0093] determining a first evaluation value based on the first feature data;

[0094] determining a second evaluation value based on the second feature data;

[0095] determining a third evaluation value based on the third feature data;

[0096] The first evaluation value, the second evaluation value and the third evaluation value are weighted and summed to obtain a fourth evaluation value;

[0097] Determine a monitoring period based on the fourth evaluation value; wherein the monitoring period is negatively correlated with the fourth evaluation value;

[0098] Prompt information is sent to the first device according to the monitoring period to indicate the collection time of the first data, the third data and the fourth data.

[0099] In this embodiment, the first characteristic data can reflect the vital signs of the user, the second characteristic data can reflect the motor ability of the user's limbs, and the third characteristic data can reflect the metabolic level and nutritional status of the user. By respectively determining the first evaluation value corresponding to the first characteristic data, the second evaluation value corresponding to the second characteristic data, and the third evaluation value corresponding to the third characteristic data, and weighted summing the first evaluation value, the second evaluation value and the third evaluation value to obtain the fourth evaluation value, the user's physical condition can be fully reflected. When the fourth evaluation value is large, it indicates that the user is in good physical condition, and the number of monitoring of the first data, the third data and the fourth data can be reduced, that is, the monitoring period of the first data, the third data and the fourth data can be increased. On the contrary, when the fourth evaluation value is small, it indicates that the user is in poor physical condition, and the number of monitoring of the first data, the third data and the fourth data can be increased, that is, the monitoring period of the first data, the third data and the fourth data can be reduced.

[0100] Specifically, determining the monitoring period based on the fourth evaluation value can be described in detail as follows:

[0101] If the fourth evaluation value is less than the first threshold, the monitoring period is determined to be the first value;

[0102] If the fourth evaluation value is between the first threshold and the second threshold, the monitoring period is determined to be the second value;

[0103] If the fourth evaluation value is greater than the second threshold, the monitoring period is determined to be a third value;

[0104] The first value, the second value and the third value are thereby reduced.

[0105] The first threshold, the second threshold, the first value, the second value and the third value are all preset constants, and those skilled in the art can determine specific values ​​according to actual needs. For example, the first threshold is set to 0.3, the second threshold is set to 0.6, the first value can be 3 months, the second value can be one month, and the third value can be one week.

[0106] When determining the first evaluation value, the normal range of each vital sign such as heart rate and blood pressure can be set, and the monitored value of each vital sign can be compared with the normal range. The signs within the normal range are determined to be normal, and the signs beyond the normal range are determined to be abnormal, and the proportion of normal indicators is determined as the first evaluation value. The second evaluation value and the third evaluation value can be obtained by the same calculation method.

[0107] From the above, it can be concluded that this embodiment converts various types of feature data into specific evaluation values ​​and weighted sums them, quantifies the user's physical condition, and determines the monitoring period of the first data, the third data, and the fourth data based on the quantified fourth evaluation value. This can reduce the number of data collection times as much as possible while ensuring the monitoring effect.

[0108] In one embodiment of the present disclosure, the robot data processing method further includes:

[0109] If the basic information of the user shows that the user is in the early stage of rehabilitation, the weights of the first evaluation value, the second evaluation value and the third evaluation value are set to decrease in sequence;

[0110] If the basic information of the user shows that the user is in the mid-term recovery stage, the weights of the second evaluation value, the first evaluation value and the third evaluation value are set to decrease in sequence;

[0111] If the basic information of the user shows that the user is in a late recovery stage and the user suffers from a specified disease, the weights of the third evaluation value, the first evaluation value and the second evaluation value are set to decrease in sequence.

[0112] In this embodiment, considering that in the early stage of rehabilitation, the user's physical condition is unstable, and changes in vital signs directly reflect the development and risk of the disease; at the same time, rehabilitation training has just begun, and motor function has not been significantly improved; body composition is not the primary factor of concern, but mainly serves as a basic data record to provide a reference for interventions such as nutrition and exercise in the subsequent rehabilitation process. Therefore, at this time, the weight of the first evaluation value can be set to be greater than the weight of the second evaluation value, and the weight of the second evaluation value can be set to be greater than the weight of the third evaluation value.

[0113] In the middle stage of rehabilitation, the user's physical condition is relatively stable, vital signs are basically stable, and rehabilitation training is in full swing, with motor function recovery becoming the focus. For example, for stroke patients, the recovery of limb motor function is the key to evaluating rehabilitation effects and adjusting control strategies, so the second evaluation value has the largest weight.

[0114] In the later stage of rehabilitation, the user's motor function is basically restored and vital signs are relatively stable. At this time, if the user suffers from a specified disease, such as obesity, accompanied by metabolic disorders such as high blood lipids and high blood sugar, various indicators of body composition such as body fat percentage, muscle mass, bone density, etc. are crucial for assessing the user's nutritional status and health risks. At this time, the weight of the third evaluation value is set to the maximum.

[0115] It can be concluded from the above that this embodiment sets different weights according to different rehabilitation stages of the user, so that the calculation of the fourth evaluation value is more in line with the actual situation of the user, and further determines a reasonable monitoring period.

[0116] Corresponding to the robot data processing method of the above embodiment, Figure 2 This is a block diagram of a robot data processing device provided by an embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 2The robot data processing device 20 includes: a first feature extraction module 21, a second feature extraction module 22, a third feature extraction module 23 and a data matching module 24.

[0117] The first feature extraction module 21 is used to fuse the first data and the second data to obtain the first feature data, where the first data is the user's vital sign monitoring data and the second data is the user's basic information;

[0118] The second feature extraction module 22 is used to extract features from the third data to obtain second feature data; the third data is the user's motor function monitoring data;

[0119] The third feature extraction module 23 is used to extract features from the fourth data to obtain third feature data; the fourth data is the user's body composition monitoring data;

[0120] The data matching module 24 is used to search the set strategy library based on the first feature data, the second feature data and the third feature data, and determine the target control strategy of the robot based on the search result; the target control strategy is used to control the robot.

[0121] In one embodiment of the present disclosure, the first feature extraction module 21 is specifically used for:

[0122] Determine a threshold value corresponding to each first data based on the second data;

[0123] Compare each first data with a corresponding threshold value to obtain evaluation data corresponding to each first data;

[0124] The first data and the evaluation data are concatenated to obtain first characteristic data.

[0125] In one embodiment of the present disclosure, the second feature extraction module 22 is specifically used for:

[0126] determining a motion monitoring item for the user based on the second data;

[0127] The third data is collected according to the motion monitoring project.

[0128] In one embodiment of the present disclosure, the data matching module 24 is specifically used to:

[0129] Fusing the first feature data, the second feature data, and the third feature data to obtain a first feature vector;

[0130] The distances between the first feature vector and the M second feature vectors in the strategy library are calculated respectively; the strategy library includes M classified sample data, the M second feature vectors are feature vectors corresponding to the M first sample data, and the M first sample data are sample data selected from the M classified sample data respectively;

[0131] Select K first sample data whose corresponding distance is less than the distance threshold; where M and K are both natural numbers, and K <M;

[0132] The target control strategy of the robot is determined based on the control strategies corresponding to the K sample data.

[0133] In one embodiment of the present disclosure, the data matching module 24 is further configured to:

[0134] Build an index structure for N sample data in the strategy library;

[0135] Based on the index structure, N sample data in the strategy library are classified to obtain M classified sample data, where N is a natural number and N>M.

[0136] In one embodiment of the present disclosure, the data matching module 24 is further configured to:

[0137] The sample data partitioning operation is performed multiple times until the stopping condition is met to obtain an index structure;

[0138] The sample data partitioning operations include:

[0139] Calculate the correlation coefficients between each dimension data in the target subtree and other dimension data respectively; wherein any dimension data is a collection of data of multiple sample data in the target subtree in the dimension;

[0140] The dimension with the lowest corresponding correlation coefficient is determined as the split axis, and the target subtree is divided into two left and right subtrees;

[0141] When the sample data partitioning operation is performed for the first time, the policy library is determined as the target subtree, and when the sample data partitioning operation is performed for other times, the left and right subtrees are respectively determined as the target subtrees;

[0142] The stopping condition is: the number of sample data in the target subtree is less than the set number.

[0143] In one embodiment of the present disclosure, the data matching module 24 is specifically used to:

[0144] determining a first evaluation value based on the first feature data;

[0145] determining a second evaluation value based on the second feature data;

[0146] determining a third evaluation value based on the third feature data;

[0147] The first evaluation value, the second evaluation value and the third evaluation value are weighted and summed to obtain a fourth evaluation value;

[0148] Determine a monitoring period based on the fourth evaluation value; wherein the monitoring period is negatively correlated with the fourth evaluation value;

[0149] Prompt information is sent to the first device according to the monitoring period to indicate the collection time of the first data, the third data and the fourth data.

[0150] See also Figure 3 , Figure 3 This is a schematic block diagram of an intelligent robot provided by an embodiment of the present disclosure. Figure 3 The intelligent robot 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303 and one or more memories 304. The processors 301, input devices 302, output devices 303 and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of each module / unit in the above-mentioned device embodiments, such as Figure 2 The functions of modules 21 to 24 are shown.

[0151] It should be understood that in the embodiment of the present disclosure, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0152] The input device 302 may include a touch panel, a fingerprint collection sensor (for collecting the user's fingerprint information and fingerprint direction information), a microphone, etc., and the output device 303 may include a display (LCD, etc.), a speaker, etc.

[0153] The memory 304 may include a read-only memory and a random access memory, and provide instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store information about the device type.

[0154] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present disclosure can execute the implementation methods described in the first and second embodiments of the robot data processing method provided in the embodiments of the present disclosure, and can also execute the implementation methods of the intelligent robot described in the embodiments of the present disclosure, which will not be repeated here.

[0155] In another embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by the processor, all or part of the processes in the above-mentioned embodiment method are implemented, and the computer program can also be completed by instructing the relevant hardware through the computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, the steps of each of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0156] The computer-readable storage medium may be an internal storage unit of the intelligent robot of any of the aforementioned embodiments, such as a hard disk or memory of the intelligent robot. The computer-readable storage medium may also be an external storage device of the intelligent robot, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the intelligent robot. Furthermore, the computer-readable storage medium may also include both an internal storage unit and an external storage device of the intelligent robot. The computer-readable storage medium is used to store computer programs and other programs and data required by the intelligent robot. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0157] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0158] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the intelligent robot and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0159] In the several embodiments provided in the present application, it should be understood that the disclosed intelligent robots and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or it can be an electrical, mechanical or other form of connection.

[0160] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present disclosure.

[0161] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0162] The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present disclosure, and these modifications or replacements should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. A robot data processing method, characterized in that: include: Fusing the first data with the second data to obtain first feature data, wherein the first data is the vital sign monitoring data of the user, and the second data is the basic information of the user; Extracting features from the third data to obtain second feature data; the third data is the motor function monitoring data of the user; Extracting features from the fourth data to obtain third feature data; the fourth data is the body composition monitoring data of the user; Searching a set strategy library based on the first feature data, the second feature data, and the third feature data, and determining a target control strategy of the robot based on the search result; The target control strategy is used to control the robot; The fusing the first data and the second data comprises: Determine a threshold value corresponding to each first data based on the second data; Compare each first data with a corresponding threshold value to obtain evaluation data corresponding to each first data; Concatenate the first data with the evaluation data to obtain the first feature data; Searching a set strategy library based on the first feature data, the second feature data, and the third feature data, and determining a control strategy of the robot based on the search result, including: Fusing the first feature data, the second feature data, and the third feature data to obtain a first feature vector; respectively calculating the distances between the first feature vector and M second feature vectors in the strategy library; the strategy library includes M classified sample data, the M second feature vectors are feature vectors corresponding to the M first sample data, and the M first sample data are sample data selected from the M classified sample data; Select K first sample data whose corresponding distance is less than the distance threshold; where M and K are both natural numbers, and K <M; The target control strategy of the robot is determined based on the control strategies corresponding to the K sample data.

2. The robot data processing method according to claim 1, characterized in that: Also includes: determining a sports monitoring item for the user based on the second data; The third data is collected according to the motion monitoring project.

3. The robot data processing method according to claim 1, characterized in that: Also includes: Constructing an index structure for the N sample data in the strategy library; Based on the index structure, N sample data in the policy library are classified to obtain M classified sample data, where N is a natural number and N>M.

4. The robot data processing method according to claim 1, characterized in that: Also includes: determining a first evaluation value based on the first feature data; determining a second evaluation value based on the second feature data; determining a third evaluation value based on the third feature data; Obtain a fourth evaluation value by weighted summing the first evaluation value, the second evaluation value and the third evaluation value; Determining a monitoring period based on the fourth evaluation value; wherein the monitoring period is negatively correlated with the fourth evaluation value; Prompt information is sent to the first device according to the monitoring period to indicate the collection time of the first data, the third data and the fourth data.

5. A robot data processing device, characterized in that: include: A first feature extraction module is used to fuse the first data and the second data to obtain first feature data, wherein the first data is the vital sign monitoring data of the user, and the second data is the basic information of the user; A second feature extraction module is used to extract features from the third data to obtain second feature data; the third data is the motor function monitoring data of the user; A third feature extraction module is used to extract features from the fourth data to obtain third feature data; the fourth data is the body composition monitoring data of the user; A data matching module, used to search a set strategy library based on the first feature data, the second feature data and the third feature data, and determine a target control strategy of the robot based on the search result; the target control strategy is used to control the robot; The first feature extraction module is specifically used for: Determine a threshold value corresponding to each first data based on the second data; Compare each first data with a corresponding threshold value to obtain evaluation data corresponding to each first data; Concatenate the first data with the evaluation data to obtain first feature data; The data matching module is specifically used for: Fusing the first feature data, the second feature data, and the third feature data to obtain a first feature vector; The distances between the first feature vector and the M second feature vectors in the strategy library are calculated respectively; the strategy library includes M classified sample data, the M second feature vectors are feature vectors corresponding to the M first sample data, and the M first sample data are sample data selected from the M classified sample data respectively; Select K first sample data whose corresponding distance is less than the distance threshold; where M and K are both natural numbers, and K <M; The target control strategy of the robot is determined based on the control strategies corresponding to the K sample data.

6. An intelligent robot, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.

Images