Robot communication method and device and medical care robot

Abstract

The invention provides a robot communication method and device and a medical care robot, and belongs to the technical field of robot communication, and the method comprises the steps: determining the comprehensive complexity of target text information; in response to the fact that the comprehensive complexity of the target text information is smaller than a first threshold value, feature extraction is conducted on the target text information based on the first processor to obtain state information of the target person, and the state information is sent to the target device; in response to the condition that the comprehensive complexity of the target text information is greater than or equal to a first threshold value, transmitting the target text information and corresponding identification information to a second processor based on a target strategy, the identification information being used for instructing the second processor to perform feature extraction on the target text information to obtain state information of the target person, and sending the state information to the target device. According to the robot communication method and device and the medical care robot provided by the invention, the quality and efficiency of medical and old-age care services can be improved.

Description

Technical Field

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

[0002] In the field of medical care and elderly care, medical robots are playing an increasingly important role. They can interact with the target person through voice and assist medical and elderly care service personnel in their work. However, the existing communication methods of medical robots have certain limitations. Existing robots use a single processing mode, and no matter how complex the voice information is, it will be processed by the local processor. For complex voice information, the local processor may not be able to accurately analyze and process it, resulting in the inability to obtain the status information of the target person in a timely and accurate manner. This makes it difficult for family members and caregivers to understand the situation of the target person in a timely and comprehensive manner, affecting the quality and efficiency of medical and elderly care services. Summary of the invention

[0003] The purpose of the present disclosure is to provide a robot communication method and device, and a medical care robot to improve the quality and efficiency of medical and elderly care services.

[0004] A first aspect of an embodiment of the present disclosure provides a robot communication method, comprising: Determining the comprehensive complexity of target text information, wherein the target text information is text information obtained by converting audio information of a target person; In response to the comprehensive complexity of the target text information being less than a first threshold, extracting features from the target text information based on the first processor to obtain status information of the target person, and sending the status information to the target device; In response to the comprehensive complexity of the target text information being greater than or equal to the first threshold, transmitting the target text information and its corresponding identification information to the second processor based on the target strategy, wherein the identification information is used to instruct the second processor to extract features from the target text information to obtain status information of the target person, and send the status information to the target device; The first processor is a processor set inside the medical robot, and the second processor is a processor set in the cloud.

[0005] A second aspect of the embodiments of the present disclosure provides a robot communication device, including: A complexity calculation module, used to determine the comprehensive complexity of target text information, wherein the target text information is text information obtained by converting the audio information of the target person; A first communication processing module, configured to extract features of the target text information based on the first processor to obtain status information of the target person in response to the comprehensive complexity of the target text information being less than a first threshold, and send the status information to the target device; A second communication processing module, configured to transmit the target text information and its corresponding identification information to a second processor based on a target strategy in response to the comprehensive complexity of the target text information being greater than or equal to a first threshold, wherein the identification information is used to instruct the second processor to perform feature extraction on the target text information to obtain status information of a target person, and send the status information to a target device; The first processor is a processor set inside the medical robot, and the second processor is a processor set in the cloud.

[0006] According to a third aspect of the embodiments of the present disclosure, a medical 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 communication method when executing the computer program.

[0007] The beneficial effects of the robot communication method and device, and the medical robot provided by the embodiments of the present disclosure are: the embodiments of the present disclosure achieve efficient use of resources by determining the comprehensive complexity of the target text information and adopting different processing strategies. For texts with low complexity, they are processed by the first processor inside the medical robot, which reduces the time and resource consumption of communicating with the cloud, can quickly obtain the target person's status information and send it to the target device, and improves the response speed. For texts with high complexity, with the help of the powerful computing power and rich data resources of the second processor in the cloud, more accurate and in-depth analysis can be performed, ensuring the accuracy of status information extraction in complex situations. Secondly, the task allocation mechanism of the embodiments of the present disclosure enhances the adaptability and flexibility of the system, can cope with voice interactions of different difficulties, improves the service quality and reliability of the medical robot, and enables family members and caregivers to understand the status of the target person in a timely and accurate manner, and better provide medical and elderly care services. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] 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.

[0009] Figure 1 A schematic diagram of a flow chart of a robot communication method provided by an embodiment of the present disclosure; Figure 2 A structural block diagram of a robot communication device provided by an embodiment of the present disclosure; Figure 3 A schematic block diagram of a medical care robot provided in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION

[0010] 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.

[0011] 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.

[0012] Please refer to Figure 1 , Figure 1 A schematic flow chart of a robot communication method provided in an embodiment of the present disclosure, the method being applied to a medical robot, comprising: S101: Determine the comprehensive complexity of target text information, where the target text information is text information obtained by converting audio information of a target person.

[0013] A medical and nursing robot is an intelligent robot that integrates medical and elderly care service functions. It can assist medical staff and elderly care service personnel in medical and elderly care scenarios, providing users with comprehensive and personalized health management and life care services.

[0014] Target persons refer to those who interact with the medical robot through voice. They can be people who are in nursing homes, receiving home-based elderly care services, or undergoing rehabilitation in medical institutions, such as the elderly, the disabled, patients with chronic diseases, etc.; or they can be people who want to learn about medical knowledge and health advice.

[0015] The comprehensive complexity of the target text information is a comprehensive indicator to measure the complexity of a text converted from the target person's audio information. The comprehensive indicator may include but is not limited to the following dimensions: Lexical complexity, which indicates the difficulty and professionalism of the vocabulary used in the target text information; Grammatical complexity, which indicates the structural complexity of sentences in the target text information; Semantic complexity refers to the degree of abstraction of the meaning expressed in the target text information, the complexity of the logical relationship, and the number of fields involved.

[0016] In this embodiment, the medical robot can obtain audio information sent by the target person through an audio collection device such as a microphone, and convert the audio information into text form based on speech recognition technology to obtain target text information.

[0017] To determine the comprehensive complexity of the target text information, we can consider it comprehensively from multiple dimensions. For example, if the target text information contains a large number of professional medical terms, complex sentence structures, and knowledge from multiple different fields, the comprehensive complexity of the target text information will be high; on the contrary, if the target text information is just a simple expression of daily language, the complexity will be low.

[0018] S102: In response to the comprehensive complexity of the target text information being less than a first threshold, extracting features from the target text information based on a first processor to obtain state information of the target person, and sending the state information to a target device. The first processor is a processor provided inside the medical robot.

[0019] The first threshold is a pre-set value, which represents a limit of the comprehensive complexity; the comprehensive complexity of the target text information is compared with the first threshold, and the processing strategy to be adopted is determined according to the comparison result.

[0020] In this embodiment, when the comprehensive complexity of the target text information is less than the first threshold, it indicates that the target text information is relatively simple, and the first processor provided inside the medical robot is capable of processing it. The first processor can perform feature extraction on the target text information, which can specifically include identifying keywords or phrases related to the target person's status in the target text information, thereby obtaining the target person's status information.

[0021] After completing feature extraction, the first processor obtains the status information of the target person. Subsequently, the status information is sent to the target device through the communication module of the medical robot according to the preset communication protocol (such as Wi-Fi, Bluetooth, 4G / 5G, etc.). The target device can be a terminal device such as a mobile phone, tablet computer, smart bracelet, etc. of a family member or caregiver, so that they can understand the status of the target person in a timely manner.

[0022] For example, if the target text message is "I feel a little headache today", the first processor can determine that the current state of the target person is a headache through feature extraction. Finally, the first processor sends the extracted state information to the target device so that family members, caregivers, etc. can understand the state of the target person in a timely manner.

[0023] S103: In response to the comprehensive complexity of the target text information being greater than or equal to the first threshold, the target text information and its corresponding identification information are transmitted to the second processor based on the target strategy, the identification information is used to instruct the second processor to extract features from the target text information to obtain the status information of the target person, and send the status information to the target device. The second processor is a processor set in the cloud.

[0024] In this embodiment, if the comprehensive complexity of the target text information is greater than or equal to the first threshold, it indicates that the target text information is relatively complex and exceeds the processing capacity of the first processor. At this time, the medical robot can transmit the target text information and its corresponding identification information to the second processor set in the cloud based on the target strategy. The target strategy can comprehensively consider various factors, such as network conditions (bandwidth, delay, stability), data volume, transmission cost, etc., to select the optimal transmission path and method. The identification information contains the key attributes of the target text information, such as the source of the text, generation time, category, etc., which is used to instruct the second processor to process the target text information based on the identification information.

[0025] After the second processor receives the target text information and identification information, it extracts features of the target text information according to the instructions of the identification information through computing power and rich data resources that are superior to those of the first processor. The cloud can have more advanced natural language processing models, large-scale knowledge bases, and professional medical knowledge graphs, which can conduct in-depth analysis and understanding of complex texts, thereby more accurately extracting the status information of the target person. After the second processor completes feature extraction and obtains the status information of the target person, it also sends the status information to the target device through network communication. During the sending process, the same or compatible communication protocol as the first processor will be followed to ensure that the status information can reach the target device accurately.

[0026] For example, in a warm nursing home, a medical robot is deployed. Elderly A said anxiously to the nurse in the room: "I have been very anxious recently, and I always suffer from insomnia at night. I have a history of heart disease before. Could this be the recurrence of the old disease?" After the medical robot obtains the audio and converts it into text, it analyzes and finds that the text contains professional vocabulary such as "anxiety", "insomnia", and "heart disease history". The sentence structure is relatively complex and involves medical knowledge, and the comprehensive complexity is greater than the first threshold. Based on the target strategy that considers factors such as network bandwidth and stability, the medical robot transmits the text and identification information to the second processor in the cloud through the 4G network. With advanced natural language processing models and medical knowledge graphs, the cloud accurately judges that Grandma Li may have a recurrence of heart disease, and sends this status information to the doctor's tablet in time. The doctor quickly rushed to Grandma Li's room for examination and treatment, avoiding the deterioration of her condition.

[0027] It can be concluded from the above that this embodiment achieves efficient use of resources by determining the comprehensive complexity of the target text information and adopting different processing strategies. For texts with low complexity, they are processed by the first processor inside the medical robot, which reduces the time and resource consumption of communicating with the cloud, can quickly obtain the target person's status information and send it to the target device, and improves the response speed. For texts with high complexity, with the help of the powerful computing power and rich data resources of the second processor in the cloud, more accurate and in-depth analysis can be carried out, ensuring the accuracy of status information extraction in complex situations. Secondly, the task allocation mechanism of this embodiment enhances the adaptability and flexibility of the system, can cope with voice interactions of different difficulties, improves the service quality and reliability of the medical robot, enables family members and caregivers to understand the status of the target person in a timely and accurate manner, and better provide medical and elderly care services.

[0028] In one embodiment of the present disclosure, determining the comprehensive complexity of target text information includes: Extracting features from the audio information of the target person to obtain target audio feature information; Perform semantic analysis on the target text information to obtain target semantic information; Determine the communication type between the target person and the medical robot based on the target text information; assigning first weights to the target audio feature information and the target semantic information respectively based on the communication type; The first weight is weightedly fused to obtain the comprehensive complexity of the target text information.

[0029] In this embodiment, the audio information contains clues about the target person's expression state, and different audio features reflect the complexity of the text information. For example, a fast speaking speed means that the target person wants to express more and more urgently, and the text complexity may be high.

[0030] Audio processing technology can be used to extract speech rate, intonation, volume, pause frequency, etc. from audio information. For example, the speech rate can be determined by calculating the time interval between adjacent speech frames in the audio; intonation information can be obtained by analyzing the frequency and amplitude changes of the audio. These extracted features are combined to form the target audio feature information.

[0031] In this embodiment, semantic analysis can help understand the meaning expressed by the target text information, the relationship between concepts, and the logical structure.

[0032] The target text information can be subjected to part-of-speech tagging, syntactic analysis, named entity recognition, semantic role labeling, etc. For example, part-of-speech tagging can help us understand the distribution of different parts of speech in the text. Complex texts may contain more professional terms and complex vocabulary. Syntactic analysis can help us determine the structural complexity of sentences, such as whether there is a nested structure. The analysis results are integrated to obtain the target semantic information.

[0033] In this embodiment, different communication types may have different complexity characteristics. For example, medical consultation-type communication involves professional medical knowledge and complex symptom descriptions, and therefore has a relatively high complexity; whereas daily greeting-type communication is usually relatively simple.

[0034] The target text information can be classified by preset classification rules. The preset rules can be defined based on keywords, sentence patterns, etc. in the text. For example, if the text contains keywords such as "symptoms" and "treatment methods", it means medical consultation communication.

[0035] In this embodiment, the audio features and semantic information have different influences on the text complexity in different communication types. For example, in emergency communication, the audio features can better reflect the complexity of the text; while in academic communication, semantic information is more important. Therefore, by assigning different weights to the audio feature information and semantic information according to the communication type, these two factors can be considered more reasonably.

[0036] Based on experience, corresponding weight allocation rules can be set in advance for different communication types. For example, for medical consultation communication, a higher weight is assigned to the target semantic information, because professional medical knowledge and symptom description are the key to this type of communication; and for emotional expression communication, a higher weight is assigned to the target audio feature information, because the tone and voice in the audio can better reflect the complexity of the emotion. According to the determined communication type, the corresponding weight is selected from the preset rules and assigned to the target audio feature information and the target semantic information respectively.

[0037] In this embodiment, the target audio feature information and the target semantic information can be combined by weighted fusion to form a comprehensive complexity index, which can comprehensively consider the factors of audio and semantics, thereby more comprehensively and accurately reflecting the complexity of the target text information.

[0038] The target audio feature information and the target semantic information may be multiplied by their corresponding first weights respectively, and then the products are added to obtain the comprehensive complexity of the target text information.

[0039] From the above, it can be concluded that this embodiment comprehensively considers the complexity of the text from both the audio and text semantics aspects by extracting target audio feature information and performing semantic analysis. Combining the weights assigned by communication type can more accurately reflect the impact of various factors on complexity in different communication scenarios. The comprehensive complexity obtained by weighted fusion is more in line with the actual situation, improving the efficiency and accuracy of medical robots in processing voice information.

[0040] In one embodiment of the present disclosure, weighted fusion is performed on the first weight to obtain the comprehensive complexity of the target text information, including: Perform weighted fusion on the first weight based on the first formula to obtain the comprehensive complexity of the target text information; The first formula is:

[0041] in, Represents the comprehensive complexity of the target text information, The value mapped to the communication type between the target person and the medical robot. Indicates the maximum value of the mapped value among all communication types. Represents the target audio feature information, represents the first weight corresponding to the target audio feature information, represents the first weight corresponding to the target semantic information, Represents the target semantic information, represents the nonlinear adjustment index, Represents the information timeliness factor.

[0042] In this embodiment, It indicates the impact of communication type on comprehensive complexity. Different communication types (such as daily greetings, medical consultations, emergency assistance, etc.) correspond to different values. is the maximum value of the mapped values ​​in all communication types, which is used to normalize the impact of communication types. The importance and complexity of target audio feature information and target semantic information vary in different communication types. For example, in the communication type of emergency help, the complexity and importance of audio features and semantic information will increase. This factor can amplify or reduce the impact of audio features and semantic information on the overall complexity depending on the specific communication type.

[0043] and The target audio feature information and the target semantic information are weighted summed respectively. and Reflects the relative importance of audio features and semantic information in the calculation of comprehensive complexity, which is pre-assigned according to different communication types. Nonlinear Adjustment Index The introduction of breaks the traditional linear weighting method. When , the influence of high eigenvalues ​​on the comprehensive complexity can be strengthened, that is, the larger the eigenvalue, the greater its contribution to the comprehensive complexity; when When , the influence of high eigenvalues ​​can be weakened, making the complexity calculation smoother and avoiding excessive influence of individual high eigenvalues ​​on the comprehensive results.

[0044] In the medical and nursing scenario, the timeliness of information is very important. The latest information can often more accurately reflect the current status of the target person. When it is close to 1), the calculation result of the comprehensive complexity will be relatively reduced, because the latest information is more reliable, reducing the uncertainty caused by outdated information; on the contrary, when the information is outdated, When it is close to 0, the calculation result of the comprehensive complexity will be relatively higher, and the target text information needs to be processed more carefully.

[0045] From the above, it can be concluded that this embodiment comprehensively considers the communication type, audio and semantic information and information timeliness. The nonlinear adjustment index enhances the flexibility of the influence of the eigenvalue, can more accurately reflect the complexity of the text, and provide a more reliable basis for the subsequent processing of the medical robot, thereby improving the pertinence and efficiency of the service.

[0046] In one embodiment of the present disclosure, it also includes: determining a threshold adjustment step size based on a deviation between the comprehensive complexity and a standard threshold and a resource status of the first processor; The standard threshold is adjusted based on the threshold adjustment step to obtain a first threshold.

[0047] In this embodiment, the standard threshold is a pre-set ideal reference value for measuring the complexity of the text in the scenario where the medical robot processes the target text information. It is used to determine whether the target text information should be processed by the first processor inside the medical robot or transmitted to the second processor in the cloud for processing. However, in actual use, since the complexity of the target text information will continue to change and the resource status of the first processor is unstable, it is difficult for the standard threshold to accurately adapt to all situations, so there will be a certain error. Therefore, it is necessary to dynamically adjust it in combination with the actual comprehensive complexity and its deviation, as well as the resource status of the first processor, in order to obtain a first threshold that better meets the actual processing needs.

[0048] In this embodiment, the calculated comprehensive complexity With standard threshold Make a comparison and calculate the difference:

[0049] A positive deviation indicates that the comprehensive complexity is higher than the standard threshold, which means that the current text information may be more complex than expected; a negative deviation indicates that the comprehensive complexity is lower than the standard threshold, that is, the text information is relatively simple.

[0050] The resource status of the first processor may include multiple aspects such as CPU usage and memory occupancy. The resource status of the first processor can intuitively reflect the current load and processing capacity of the first processor. For the convenience of calculation, the resource status of the first processor can be quantified as a specific indicator R. For example, the resource utilization rate can be expressed as a percentage, with a value range of The larger the value, the tighter the resources of the first processor and the more limited its processing capacity.

[0051] In this embodiment, the step size can be adjusted by weighted calculation of the threshold, which is expressed as:

[0052] in, represents the threshold adjustment step size, Represents the weight coefficient corresponding to the deviation between the comprehensive complexity and the standard threshold, Indicates the weight coefficient corresponding to the resource state of the first processor.

[0053] The standard threshold is adjusted accordingly according to the determined threshold adjustment step. If the threshold adjustment step is positive, it means that the first threshold needs to be increased so that more complex text information can be assigned to the second processor in the cloud for processing, thereby reducing the burden on the first processor; if the threshold adjustment step is negative, the first threshold needs to be lowered so that the first processor can process more relatively simple text information and make full use of its resources.

[0054] It can be concluded from the above that this embodiment dynamically determines the threshold adjustment step and adjusts the standard threshold by comprehensively considering the deviation between the comprehensive complexity and the standard threshold and the resource status of the first processor, thereby obtaining a first threshold that is more in line with the actual situation. This dynamic adjustment mechanism enables the medical robot to flexibly allocate processing tasks according to real-time conditions, avoids the limitations of fixed thresholds in complex and changeable actual scenarios, and improves the processing efficiency and adaptability of the system.

[0055] In one embodiment of the present disclosure, the first processor extracts features from target text information to obtain status information of a target person, and sends the status information to a target device, including: Extract keywords from target text information; Match the keywords with the nodes in the knowledge graph to obtain the status information of the target person; Sends status information to the target device.

[0056] In this embodiment, keywords are words or phrases that have core meanings in the text and can summarize the main content of the text. By extracting keywords, complex text information can be simplified and focused on key content, providing a basis for subsequent analysis and processing.

[0057] Keywords in the target text information can be extracted based on the term frequency-inverse document frequency (TF-IDF) algorithm. The importance of each word is evaluated by calculating its frequency of occurrence in the text and its inverse document frequency in the entire corpus. Words with high frequency and low frequency in other documents are more likely to be selected as keywords.

[0058] In this embodiment, the knowledge graph is a structured semantic network that can represent various entities (such as diseases, symptoms, treatment methods, etc.) and their relationships (such as causal relationships, treatment relationships, etc.) in the form of a graph. Each entity corresponds to a node in the knowledge graph, and the relationship corresponds to the edge between the nodes. By matching the extracted keywords with the nodes in the knowledge graph, the rich knowledge and associated information in the knowledge graph can be used to infer the status information of the target person.

[0059] Search the knowledge graph for nodes that match the extracted keywords. If the keyword directly corresponds to a node in the knowledge graph, you can directly obtain the information represented by the node and other related node information.

[0060] For example, if the keyword is "cough", after finding the "cough" node in the knowledge graph, you can further obtain information about cough-related diseases (such as colds, pneumonia, etc.), possible causes, and corresponding treatment recommendations. If the keyword does not completely match the node in the knowledge graph, but has a semantic association, it can be associated and inferred through the semantic reasoning mechanism of the knowledge graph. For example, if the keyword is "uncomfortable throat", you can further find related disease and treatment information through the association of "uncomfortable throat" with symptom nodes such as "sore throat" and "dry throat" in the knowledge graph, so as to obtain more accurate status information of the target person.

[0061] After obtaining the status information of the target person, it is necessary to transmit this information to the relevant target device in a timely manner so that family members, caregivers, etc. can understand the situation of the target person in a timely manner and take corresponding measures.

[0062] From the above, it can be concluded that this embodiment can quickly focus on the key content of the target text by extracting keywords. By matching keywords with knowledge graph nodes, the status of the target person can be accurately inferred with the help of the knowledge graph to enrich the information. Finally, the status information is sent to the target device so that relevant personnel can understand the situation in a timely manner, which helps to take timely response measures and improve the efficiency and quality of medical and nursing services.

[0063] In one embodiment of the present disclosure, sending status information to a target device includes: Determine a target network path based on the network status; determining a target transmission rate of the state information based on the target network path, the importance level corresponding to the state information, and the resource state of the first processor; Status information is sent to the target device based on the target transfer rate.

[0064] In this embodiment, different network paths have different performance indicators, such as bandwidth, delay, packet loss rate, etc. The quality of the network status directly affects the efficiency and reliability of information transmission. Therefore, it is necessary to monitor the network status in real time and select the network path with the best performance as the target network path to ensure that the status information can be transmitted quickly and stably.

[0065] In this embodiment, the medical robot can include multiple network connection methods, such as Wi-Fi, 4G / 5G, etc. By sending test data packets on different network interfaces, the network performance indicators such as bandwidth, delay, and packet loss rate are collected. The quality of different network paths is evaluated. For example, an evaluation function can be set to comprehensively consider factors such as bandwidth, delay, and packet loss rate, calculate the score of each network path, and select the network path with the highest score as the target network path.

[0066] In this embodiment, the target network path determines the upper limit capacity of information transmission, the importance level of the status information determines the priority of transmission, and the resource status of the first processor affects the transmission rate that it can support.

[0067] The importance of status information is classified in advance, for example, into three levels: high, medium, and low. Important information such as urgent medical conditions is classified as high; general daily situation reports are classified as medium; and less important chat information is classified as low.

[0068] According to the performance indicators of the target network path, the maximum transmission rate that it can support is determined. At the same time, according to the resource status of the first processor (such as CPU usage, memory occupancy, etc.), the available transmission resources that it can provide are evaluated. Combined with the importance level of the state information, the target transmission rate is determined according to certain rules. For example, for high-level state information, if the resources of the first processor allow, try to use the maximum transmission rate of the target network path; for medium-level state information, appropriately reduce the transmission rate; for low-level state information, the transmission rate can be further reduced.

[0069] In this embodiment, the first processor adjusts the sending frequency and data packet size of the status information according to the target transmission rate. By controlling the sending frequency and data packet size, the status information is transmitted at the target transmission rate. For example, if the target transmission rate is , the sending time interval of each data packet can be calculated and data packet size S, and sends status information in sequence according to this time interval and data packet size.

[0070] From the above, it can be concluded that this embodiment dynamically determines the target network path and target transmission rate by comprehensively considering the network status, the importance level of the status information and the resource status of the first processor, thereby achieving efficient and reliable transmission of status information and improving the performance and service quality of the medical robot communication.

[0071] In one embodiment of the present disclosure, the identification information includes the importance level corresponding to the target text information; The second processor extracts features of the target text information based on the importance level corresponding to the target text information.

[0072] In this embodiment, the medical robot can be used in a nursing home. There are multiple medical robots in the nursing home. The second processor may receive multiple target text information at the same time. In order to process this information efficiently and reasonably, the second processor can extract features of the target text information according to the importance level, give priority to processing high-importance information, and ensure that key information can be analyzed and processed in a timely manner.

[0073] In this embodiment, after the medical robot obtains the audio information of the target person and converts it into target text information, it can determine its importance level according to the text content. Different types of text information have different impacts on the health and safety of the target person, so they need to be graded. The importance level can help the second processor clarify the priority of each target text information so as to reasonably allocate processing resources.

[0074] The medical robot can preset rules to determine the importance level. For example, target text information containing emergency medical assistance (such as "I have a heart attack"), serious safety issues (such as "fire"), etc. can be classified as high importance; content involving daily health consultation (such as "I have a little insomnia recently"), general needs (such as "I want a glass of water"), etc. are classified as medium importance; small talk and insignificant words (such as "The weather is nice today") are classified as low importance. After the level is determined, it is sent to the second processor together with the target text information as identification information.

[0075] In this embodiment, the second processor can receive the target text information and its identification information from each medical robot through the network interface. After receiving the information, it is stored in a queue to be processed and waits for subsequent processing. The second processor prioritizes the information stored in the queue to be processed according to the importance level corresponding to the target text information. Feature extraction is performed on target text information with a high importance level first to ensure that urgent and critical information can be processed and analyzed in a timely manner, providing a basis for taking corresponding measures later.

[0076] The second processor can put the target text information with high importance level in the front of the queue based on the priority queue algorithm. When extracting features, they are processed in the queue order. For target text information with high importance level, the second processor can call more advanced processing resources and algorithms to ensure accurate and rapid extraction of key features.

[0077] For example, we can use natural language processing models and medical knowledge bases to conduct in-depth analysis of texts related to medical help and extract key information such as symptoms and medical history. For target text information of medium and low importance levels, we arrange the processing order and resource allocation according to resource conditions, and gradually complete the feature extraction of other information while ensuring the processing of high-importance information.

[0078] It can be concluded from the above that this embodiment sets the importance level for the target text information and allows the second processor to extract features based on it, thereby realizing efficient processing and reasonable scheduling of information when multiple medical care robots send information at the same time. It ensures that key information can be analyzed and processed in a timely manner, and improves the response speed and quality of medical care services in nursing homes.

[0079] In one embodiment of the present disclosure, the identification information further includes: a judgment strategy corresponding to the target text information; The judgment strategy is used to indicate that the second processor simultaneously receives multiple identification information with the same importance level, performs weighted calculation on the weight corresponding to the complexity of each dimension in the comprehensive complexity, and obtains multiple target importance levels.

[0080] In actual applications, there are multiple medical robots that convert the audio information of the target person into target text information and send it to the second processor with identification information. When there are multiple target text information with the same importance level, the original importance level division cannot meet the needs of reasonable sorting and processing.

[0081] In order to ensure reasonable processing of this information, multiple target text information with the same importance level can be re-ranked. In this embodiment, with the help of the judgment strategy in the identification information, the complexity of each dimension of the comprehensive complexity and its corresponding weight can be weighted to obtain a new target importance level, and then these target text information can be re-ranked to achieve more reasonable feature extraction and processing.

[0082] In this embodiment, when processing target text information, the medical robot can not only determine the importance level, but also generate a judgment strategy according to preset rules and include it in the identification information.

[0083] After receiving the identification information, the second processor may enable the judgment strategy when encountering multiple situations with the same importance level.

[0084] The comprehensive complexity of the target text information consists of multiple dimensions, such as lexical complexity, grammatical complexity, and semantic complexity. Each dimension contributes differently to the overall complexity, and therefore has a different corresponding weight. The evaluation method and corresponding weight of the complexity of each dimension can be pre-set.

[0085] For example, lexical complexity can be evaluated based on the proportion of professional terms and uncommon words in the text; grammatical complexity can be measured by the structural complexity of the sentence (such as the number of nesting levels, sentence length, etc.); semantic complexity can be considered based on the abstract degree of the meaning expressed by the text, the complexity of the logical relationship, etc. The weight can be set based on actual conditions and experience. For example, in medical scenarios, semantic complexity may have a higher weight.

[0086] In this embodiment, according to the judgment strategy, the complexity of each dimension in the comprehensive complexity of each target text information and its corresponding weight are weighted. The weighted calculation can comprehensively consider the influence of each dimension and more accurately reflect the actual importance of the text information. In this way, multiple target text information with the same importance level can be further distinguished.

[0087] Let the vocabulary complexity be , the weight is ; The grammatical complexity is , the weight is ; The semantic complexity is , the weight is The calculation formula of target importance level I can be expressed as:

[0088] Among them, H represents the health risk coefficient of the target person. It can be evaluated based on the target person's medical history, recent health examination results, and real-time monitoring data (such as heart rate, blood pressure, etc.). In the medical and nursing scenario, if the target person has a high health risk, the importance of the text information related to him (such as symptom description, demand expression, etc.) will increase significantly. By introducing the health risk coefficient, the target importance level can better reflect this actual situation, making the importance level assessment more in line with the scenario requirements.

[0089] U represents the urgency coefficient of the target text information. It can be judged based on factors such as keywords and tone in the text. For example, the urgency coefficient value of texts containing urgent keywords such as "help" and "sudden pain" can be set higher; while the urgency coefficient value of general daily communication texts is lower. In nursing homes, timely processing of emergency information is important to protect the life safety and health of the elderly. After the introduction of the urgency coefficient, the target importance level can highlight the importance of emergency information, ensure that such information can be processed first, and improve the response efficiency and security of the system.

[0090] Such calculation is performed for each target text information to obtain multiple target importance levels.

[0091] In this embodiment, after obtaining multiple target importance levels, they are compared and the target text information is sorted from high to low. In this way, the processing order can be clarified, and feature extraction is performed on target text information with high target importance levels first, ensuring that more critical and complex information can be processed in a timely manner.

[0092] The second processor can use a sorting algorithm to sort the multiple target importance levels. After the sorting is completed, the target text information is sequentially extracted. For text information with a high target importance level, more computing resources and more complex processing algorithms can be allocated to ensure accurate extraction of key features; for text information with a lower level, processing is reasonably arranged according to resource conditions.

[0093] From the above, it can be concluded that this embodiment uses the judgment strategy in the identification information to perform weighted calculation and comparison on each dimension of comprehensive complexity, thereby achieving reordering and reasonable processing of multiple target text information with the same importance level, and improving the efficiency and accuracy of the second processor in processing information.

[0094] Corresponding to the robot communication method of the above embodiment, Figure 2 This is a block diagram of a robot communication 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 communication device 20 includes: a complexity calculation module 21, a first communication processing module 22, and a second communication processing module 23.

[0095] The complexity calculation module 21 is used to determine the comprehensive complexity of the target text information, where the target text information is the text information obtained by converting the audio information of the target person; A first communication processing module 22 is used for, in response to the comprehensive complexity of the target text information being less than a first threshold, extracting features from the target text information based on the first processor to obtain the status information of the target person, and sending the status information to the target device; A second communication processing module 23 is used for transmitting the target text information and its corresponding identification information to the second processor based on the target strategy in response to the comprehensive complexity of the target text information being greater than or equal to the first threshold, wherein the identification information is used to instruct the second processor to extract features from the target text information to obtain the status information of the target person, and send the status information to the target device; The first processor is a processor set inside the medical robot, and the second processor is a processor set in the cloud.

[0096] In one embodiment of the present disclosure, the complexity calculation module 21 is specifically used for: Extracting features from the audio information of the target person to obtain target audio feature information; Perform semantic analysis on the target text information to obtain target semantic information; Determine the communication type between the target person and the medical robot based on the target text information; assigning first weights to the target audio feature information and the target semantic information respectively based on the communication type; The first weight is weightedly fused to obtain the comprehensive complexity of the target text information.

[0097] In one embodiment of the present disclosure, the complexity calculation module 21 is further configured to: Perform weighted fusion on the first weight based on the first formula to obtain the comprehensive complexity of the target text information; The first formula is:

[0098] in, Represents the comprehensive complexity of the target text information, The value mapped to the communication type between the target person and the medical robot. Indicates the maximum value of the mapped value among all communication types. Represents the target audio feature information, represents the first weight corresponding to the target audio feature information, represents the first weight corresponding to the target semantic information, Represents the target semantic information, represents the nonlinear adjustment index, Represents the information timeliness factor.

[0099] In one embodiment of the present disclosure, the robot communication device 20 further includes: a threshold determination module; the threshold determination module is specifically used to: determining a threshold adjustment step size based on a deviation between the comprehensive complexity and a standard threshold and a resource status of the first processor; The standard threshold is adjusted based on the threshold adjustment step to obtain a first threshold.

[0100] In one embodiment of the present disclosure, the first communication processing module 22 is specifically used to: Extract keywords from target text information; Match the keywords with the nodes in the knowledge graph to obtain the status information of the target person; Sends status information to the target device.

[0101] In one embodiment of the present disclosure, the first communication processing module 22 is further configured to: Determine a target network path based on the network status; determining a target transmission rate of the state information based on the target network path, the importance level corresponding to the state information, and the resource state of the first processor; Status information is sent to the target device based on the target transfer rate.

[0102] In one embodiment of the present disclosure, the identification information includes the importance level corresponding to the target text information; The second processor extracts features of the target text information based on the importance level corresponding to the target text information.

[0103] In one embodiment of the present disclosure, the identification information further includes: a judgment strategy corresponding to the target text information; The judgment strategy is used to indicate that the second processor simultaneously receives multiple identification information with the same importance level, performs weighted calculation on the weight corresponding to the complexity of each dimension in the comprehensive complexity, and obtains multiple target importance levels.

[0104] See also Figure 3 , Figure 3 This is a schematic block diagram of a medical robot provided by an embodiment of the present disclosure. Figure 3The medical 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 in the above-mentioned device embodiments, such as Figure 2 The functions of modules 21 to 23 are shown.

[0105] 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.

[0106] 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.

[0107] 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.

[0108] 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 communication method provided in the embodiments of the present disclosure, and can also execute the implementation methods of the medical care robot described in the embodiments of the present disclosure, which will not be repeated here.

[0109] 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.

[0110] The computer-readable storage medium may be an internal storage unit of the medical robot of any of the aforementioned embodiments, such as a hard disk or memory of the medical robot. The computer-readable storage medium may also be an external storage device of the medical 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 medical robot. Furthermore, the computer-readable storage medium may also include both an internal storage unit and an external storage device of the medical robot. The computer-readable storage medium is used to store computer programs and other programs and data required by the medical robot. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0111] 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.

[0112] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the medical robot and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0113] In the several embodiments provided in the present application, it should be understood that the disclosed medical care 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. For example, 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.

[0114] 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.

[0115] 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.

[0116] 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 communication method, applied to a medical robot, characterized in that: include: Determining the comprehensive complexity of target text information, wherein the target text information is text information obtained by converting audio information of a target person; In response to the comprehensive complexity of the target text information being less than a first threshold, extracting features from the target text information based on the first processor to obtain status information of the target person, and sending the status information to the target device; In response to the comprehensive complexity of the target text information being greater than or equal to the first threshold, transmitting the target text information and its corresponding identification information to the second processor based on the target strategy, wherein the identification information is used to instruct the second processor to perform feature extraction on the target text information to obtain the status information of the target person, and send the status information to the target device; The first processor is a processor set inside the medical robot, and the second processor is a processor set in the cloud.

2. The robot communication method according to claim 1, characterized in that: Determining the comprehensive complexity of the target text information includes: Extracting features from the audio information of the target person to obtain target audio feature information; Performing semantic analysis on the target text information to obtain target semantic information; Determine the communication type between the target person and the medical robot based on the target text information; Assigning first weights to the target audio feature information and the target semantic information based on the communication type; The first weights are weighted and fused to obtain the comprehensive complexity of the target text information.

3. The robot communication method according to claim 2, characterized in that: The step of performing weighted fusion on the first weights to obtain the comprehensive complexity of the target text information includes: Performing weighted fusion on the first weights based on the first formula to obtain the comprehensive complexity of the target text information; The first formula is: in, Represents the comprehensive complexity of the target text information, The value mapped to the communication type between the target person and the medical robot. Indicates the maximum value of the mapped value among all communication types. Represents the target audio feature information, represents the first weight corresponding to the target audio feature information, represents the first weight corresponding to the target semantic information, Represents the target semantic information, represents the nonlinear regulation index, Represents the information timeliness factor.

4. The robot communication method according to claim 1, characterized in that: Also includes: Determine a threshold adjustment step size based on a deviation between the comprehensive complexity and a standard threshold and a resource state of the first processor; The standard threshold is adjusted based on the threshold adjustment step to obtain the first threshold.

5. The robot communication method according to claim 1, characterized in that: The step of extracting features from the target text information based on the first processor to obtain status information of the target person, and sending the status information to the target device, includes: Extracting keywords from the target text information; Match the keywords with the nodes in the knowledge graph to obtain the status information of the target person; The status information is sent to the target device.

6. The robot communication method according to claim 5, characterized in that: The sending the status information to the target device includes: Determine a target network path based on the network status; determining a target transmission rate of the status information based on the target network path, the importance level corresponding to the status information, and the resource state of the first processor; The status information is sent to a target device based on the target transmission rate.

7. The robot communication method according to claim 1, characterized in that: The identification information includes the importance level corresponding to the target text information; The second processor extracts features of the target text information based on the importance level corresponding to the target text information.

8. The robot communication method according to claim 7, characterized in that: The identification information also includes: a judgment strategy corresponding to the target text information; The judgment strategy is used to indicate that the second processor simultaneously receives multiple identification information with the same importance level, performs weighted calculation on the weight corresponding to the complexity of each dimension in the comprehensive complexity, and obtains multiple target importance levels.

9. A robot communication device, applied to a medical robot, characterized in that: include: A complexity calculation module, used to determine the comprehensive complexity of target text information, wherein the target text information is text information obtained by converting the audio information of the target person; A first communication processing module, configured to extract features of the target text information based on the first processor to obtain status information of the target person in response to the comprehensive complexity of the target text information being less than a first threshold, and send the status information to the target device; A second communication processing module, configured to transmit the target text information and its corresponding identification information to a second processor based on a target strategy in response to the comprehensive complexity of the target text information being greater than or equal to a first threshold, wherein the identification information is used to instruct the second processor to perform feature extraction on the target text information to obtain status information of a target person, and send the status information to a target device; The first processor is a processor set inside the medical robot, and the second processor is a processor set in the cloud.

10. A medical 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 8 are implemented.

Images