A robot intelligent self-diagnosis method, device, equipment and medium

Abstract

The present disclosure provides a robot intelligent self-diagnosis method, device, equipment and medium, and relates to the technical field of computers, which comprises the following steps: collecting multi-dimensional data of a robot body in real time, and preprocessing the multi-dimensional data to obtain preprocessed data; wherein the multi-dimensional data comprises at least one of running data and multi-modal perception information of an external environment; based on the preprocessed data, adopting a hybrid inference mechanism to perform fault diagnosis and obtaining a hierarchical diagnosis result; based on the hierarchical diagnosis result, presenting state data of the robot to a user by using a natural language interaction mode; wherein the state data comprises at least one of a health state and fault information. By fusing time series data, state parameters, images, sounds and other multi-modal information, and using a hybrid inference combining a large model and a knowledge graph for diagnosis, the fault recognition accuracy, comprehensiveness and self-adaptive ability are significantly improved.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to a method, apparatus, device, and medium for intelligent self-diagnosis of robots. Background Technology

[0002] With the widespread application of robotics technology, it plays an increasingly important role in industrial automation, services, and special operations. Ensuring the stable and reliable operation of robots is key to maximizing their effectiveness; therefore, robot fault diagnosis and health management technologies are crucial. While existing robot fault diagnosis technologies can assist in ensuring equipment operation to some extent, they still have the following shortcomings: methods based on rule bases or expert systems rely on manually set rules, have limited coverage, and are difficult to adapt to complex, changing, and unknown fault scenarios; diagnostic methods based on single algorithm models lack generalization ability, are often limited to specific working conditions, and lack unified diagnostic capabilities across tasks and platforms; cloud-based centralized processing solutions have high network requirements, and real-time performance and reliability are difficult to guarantee in weak or offline environments, while also posing risks of data transmission delays and privacy security; locally embedded diagnostic modules are limited by computing power and storage capacity, and cannot efficiently process complex multi-source data, resulting in limited diagnostic depth and intelligence; furthermore, in many practical scenarios, solutions based on manual inspection and remote monitoring are still widely used, relying on technicians to periodically check or remotely analyze operating logs to determine faults. This approach is highly subjective, inefficient in diagnosis, and heavily reliant on human experience, failing to meet the urgent need for real-time, high-precision diagnosis in continuous, autonomous robot operation environments.

[0003] In summary, there is an urgent need for an intelligent self-diagnosis method that can achieve real-time, accurate, adaptive, and co-evolutionary capabilities to support the autonomous operation and maintenance of robots in complex environments. Summary of the Invention

[0004] This disclosure provides a method, apparatus, device, and medium for intelligent self-diagnosis of robots. The main objective is to address the technical problems of existing robot fault diagnosis technologies, such as poor adaptability and generalization ability, and difficulty in adapting to complex dynamic environments.

[0005] According to a first aspect of this disclosure, a robot intelligent self-diagnosis method is provided, comprising:

[0006] Multidimensional data of the robot body is collected in real time, and the multidimensional data is preprocessed to obtain preprocessed data; wherein, the multidimensional data includes at least one of operation data and multimodal perception information of the external environment; Based on the preprocessed data, a hybrid reasoning mechanism is used for fault diagnosis to obtain graded diagnosis results; Based on the graded diagnostic results, the robot's status data is presented to the user using natural language interaction; wherein, the status data includes at least one of health status and fault information.

[0007] Preferably, the real-time acquisition of multidimensional data of the robot body, and the preprocessing of the multidimensional data to obtain preprocessed data, includes: Based on the robot's operating site, real-time data collection is performed on the robot's operating data and multimodal perception information of the external environment. By using data preprocessing methods, the operational data of the robot body and the multimodal perception information of the external environment are cleaned and formatted to obtain preprocessed data.

[0008] Preferably, the step of using a hybrid inference mechanism to perform fault diagnosis based on the preprocessed data to obtain a graded diagnosis result includes: The preprocessed data is used to extract features and recognize patterns using a large model to generate preliminary diagnostic results. The preliminary diagnostic results are used to perform causal inference and logical verification using a knowledge graph, and the faults are classified to obtain a graded diagnostic result.

[0009] Preferably, presenting the robot's status data to the user using natural language interaction based on the graded diagnostic results includes: By using speech recognition and natural language understanding technologies, the system can analyze user questions input via voice. Based on the tiered diagnostic results, a natural language response corresponding to the user's question is generated; The natural language response is output via speech synthesis and may be supplemented with image or text information for multimodal presentation.

[0010] Preferably, it further includes: if the graded diagnostic results indicate that manual intervention is required, generating a repair guidance plan and guiding the user to perform the operation through voice prompts.

[0011] Preferably, the step of generating a repair guidance plan and guiding the user to perform the operation through voice prompts includes: The diagnostic conclusions and preset treatment plans are transformed into step-by-step voice guidance prompts; In the preset operation phase, confirmation is obtained from the user through a task flow control mechanism; The guidance strategy will be dynamically adjusted based on user feedback during the repair process. If the feedback indicates that the repair has failed, an alternative repair plan will be generated and the user will be prompted again via voice until the fault is repaired.

[0012] Preferably, it further includes: The data and results generated throughout the entire process of diagnosis and artificial repair are recorded, and the recorded experience data is consolidated through labeling and structured storage methods. Through an incremental learning mechanism, the diagnostic model and knowledge graph are dynamically updated using accumulated experience data.

[0013] According to a second aspect of this disclosure, a robotic intelligent self-diagnosis device is provided, comprising: The data acquisition module collects multidimensional data of the robot body in real time and preprocesses the multidimensional data to obtain preprocessed data; wherein, the multidimensional data includes at least one of operation data and multimodal perception information of the external environment; The fault diagnosis module, based on the preprocessed data, uses a hybrid reasoning mechanism to perform fault diagnosis and obtain a graded diagnosis result; The interaction module, based on the graded diagnostic results, presents the robot's status data to the user using natural language interaction; wherein, the status data includes at least one of health status and fault information.

[0014] According to a third aspect of this disclosure, an electronic device is provided, comprising: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect above.

[0015] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described in the first aspect above.

[0016] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0017] In the embodiments provided in this disclosure, by deeply integrating time-series data, state parameters, and multimodal information such as images and sounds, and utilizing hybrid reasoning combining large models and knowledge graphs for diagnosis, the limitations of traditional methods relying on a single data source and algorithm are overcome, significantly improving the accuracy, comprehensiveness, and adaptability of fault identification. Natural language voice interaction is introduced as the core human-machine interface, allowing users to directly obtain diagnostic conclusions and repair guidance through voice question-and-answer sessions. Combined with multimodal information presentation, this greatly enhances the intuitiveness of interaction and collaboration efficiency, lowering the professional threshold. After diagnosis, a step-by-step, dynamically adjustable voice repair guidance plan can be automatically generated. Task flow control ensures confirmation of key operations, effectively avoiding misoperation and significantly improving the safety, controllability, and success rate of on-site repair. Through full-process knowledge recording, tagged storage, and incremental learning mechanisms, continuous accumulation, self-evolution, and distributed sharing of diagnostic knowledge are achieved, enabling single machines and even robot groups to possess collaborative diagnosis and continuous optimization capabilities, fundamentally enhancing the system's intelligence and scalability.

[0018] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0019] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein: Figure 1 This is a flowchart illustrating a robot intelligent self-diagnosis method provided in an embodiment of the present disclosure; Figure 2 This is a schematic diagram of the structure of a robot intelligent self-diagnosis device provided in an embodiment of this disclosure. Detailed Implementation

[0020] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0021] The following description, with reference to the accompanying drawings, details a robot intelligent self-diagnosis method, apparatus, device, and medium according to embodiments of the present disclosure.

[0022] Figure 1 This is a flowchart illustrating a robot intelligent self-diagnosis method provided in an embodiment of this disclosure. Figure 1 As shown, the method includes the following steps: Step 101: Collect multidimensional data of the robot body in real time, and preprocess the multidimensional data to obtain preprocessed data; wherein, the multidimensional data includes at least one of operation data and multimodal perception information of the external environment; Step 102: Based on the preprocessed data, a hybrid reasoning mechanism is used to perform fault diagnosis and obtain a graded diagnosis result; Step 103: Based on the graded diagnosis results, present the robot's status data to the user using natural language interaction; wherein the status data includes at least one of health status and fault information.

[0023] Based on the above embodiments, this embodiment will provide a detailed description of step 101: In one embodiment, based on the robot's operating environment, real-time operational data of the robot body and multimodal perception information of the external environment are collected; through a data preprocessing method, the operational data of the robot body and the multimodal perception information of the external environment are cleaned and formatted to obtain preprocessed data.

[0024] Specifically, the perception agent collects multi-source information about the robot and its external environment in real time. This information includes log data, control command streams, operating status parameters of key components, and multimodal information such as images and sounds transmitted from the work site. After preprocessing the above data, it is transmitted to the diagnostic agent.

[0025] Based on the above embodiments, this embodiment will provide a detailed description of step 102: In one embodiment, a large model is used to extract features and recognize patterns from the preprocessed data to generate preliminary diagnostic results; a knowledge graph is used to perform causal inference and logical verification on the preliminary diagnostic results, and the faults are classified to obtain graded diagnostic results, wherein the graded diagnostic results include warning level, fault level and severe level.

[0026] Specifically, the diagnostic agent employs a hybrid reasoning mechanism based on a combination of large models and knowledge graphs. It performs feature fusion and causal inference on the collected time-series data and environmental perception information, which can accurately identify potential fault modes and output faults in a graded manner, including warning level, fault level and severe level, so as to facilitate subsequent handling and intervention.

[0027] Based on the above embodiments, this embodiment will provide a detailed description of step 103: In one embodiment, speech recognition and natural language understanding technologies are used to parse the user's question input via voice; combined with the hierarchical diagnostic results, a natural language response corresponding to the user's question is generated; the natural language response is output via speech synthesis, and optionally supplemented with image or text information for multimodal presentation.

[0028] Specifically, after the diagnostic results are generated, the voice interaction agent is responsible for interacting with the user in natural language. It uses speech recognition and natural language understanding technologies to interpret the user's question intent and, combined with the diagnostic results, generates a natural language response, which is then synthesized and returned to the user. This allows the user to directly obtain the robot's health status and the cause of any malfunction through voice interaction. Simultaneously, to ensure intuitive interaction, the system also supports multimodal interaction methods supplemented by images or log content, achieving human-computer communication that is voice-driven and supplemented by images and text.

[0029] If the tiered diagnostic results indicate that manual intervention is required, a repair guidance plan is generated, and the user is guided to perform the operation through voice prompts. In one embodiment, the diagnostic conclusions and preset treatment plans are transformed into step-by-step voice guidance prompts; in the preset operation steps, confirmation is made with the user through a task flow control mechanism; the guidance strategy is dynamically adjusted based on the user's feedback during the repair process; if the repair is determined to have failed based on the feedback, an alternative repair plan is regenerated and the user is prompted again via voice until the fault is repaired.

[0030] Specifically, when the diagnostic results indicate that manual intervention is required, the manual repair guidance agent translates the diagnostic conclusions and solutions into step-by-step voice prompts. Combined with a task flow control mechanism, the agent confirms with the user at key stages to avoid erroneous operations. During the manual repair process, the system can dynamically adjust the guidance strategy based on user feedback. If repair failure is detected, an alternative solution is regenerated and prompted to the user again, ensuring the safety and effectiveness of fault handling.

[0031] This embodiment also provides data recording during the fault diagnosis and repair process, and updates the diagnostic model and knowledge graph; In one embodiment, the data and results generated throughout the entire process of diagnosis and artificial repair are recorded, and the recorded experience data is accumulated through labeling and structured storage methods; the diagnostic model and knowledge graph are dynamically updated using the accumulated experience data through an incremental learning mechanism.

[0032] Specifically, the knowledge base agent records and structurally stores the entire diagnostic process and manual repair operations. It continuously updates the diagnostic model and knowledge graph through labeling and incremental learning mechanisms, thereby achieving experience accumulation and system self-evolution. The knowledge base can be shared among multiple robots, enabling the swarm of robots to possess distributed learning and collaborative diagnostic capabilities. To adapt to different network environments, this invention supports a collaborative mode of edge computing and cloud computing: when network conditions are good, the cloud agent can be invoked to complete complex model calculations and deep analysis; under weak network or network outage conditions, the edge agent can still independently run the core diagnostic functions, ensuring the system's real-time performance and robustness.

[0033] This embodiment provides a robot intelligent self-diagnosis method. By deeply integrating time-series data, state parameters, and multimodal information such as images and sounds, and utilizing hybrid reasoning combining large models and knowledge graphs for diagnosis, it overcomes the limitations of traditional methods that rely on a single data source and algorithm, significantly improving the accuracy, comprehensiveness, and adaptability of fault identification. Natural language voice interaction is introduced as the core human-machine interface, allowing users to directly obtain diagnostic conclusions and repair guidance through voice question-and-answer sessions. Combined with multimodal information presentation, this greatly enhances the intuitiveness of interaction and collaboration efficiency, lowering the professional threshold. After diagnosis, a step-by-step, dynamically adjustable voice repair guidance plan is automatically generated. Task flow control ensures confirmation of key operations, effectively avoiding misoperation and significantly improving the safety, controllability, and success rate of on-site repair. Through full-process knowledge recording, tagged storage, and incremental learning mechanisms, this method achieves continuous accumulation, self-evolution, and distributed sharing of diagnostic knowledge, enabling single robots and even robot groups to possess collaborative diagnosis and continuous optimization capabilities, fundamentally enhancing the system's intelligence and scalability.

[0034] Based on the above embodiments, this embodiment describes each intelligent agent in the robot intelligent self-diagnosis method as follows: This embodiment uses a robot performing handling tasks in a complex warehouse environment as an example.

[0035] Multi-source information perception and processing: The perception agent continuously collects real-time operational data from the robot, including joint motor current and speed logs from the controller, position feedback streams from the encoder, working scene images from the onboard camera, and mechanical operation sounds from the microphone. This multi-source heterogeneous data is preprocessed in a unified manner, including timestamp alignment, noise filtering, and standardization, forming a structured data stream that integrates temporal state and multimodal perception information, and is sent to the diagnostic agent in real time.

[0036] Hybrid intelligent reasoning and fault diagnosis: Upon receiving the aforementioned data stream, the diagnostic agent initiates a hybrid inference process. First, its built-in large model (e.g., processing images and time-series signals through its branches) performs deep feature extraction and correlation analysis on the fused data. The model initially outputs a probabilistic warning: "Left drive wheel assembly abnormal, confidence level 92%." This preliminary conclusion is then fed into the knowledge graph inference engine. The knowledge graph stores entity relationships of the robot's mechanical structure (e.g., "left drive wheel assembly" includes "wheel hub bearing," "drive motor," etc.) and causal chains of faults (e.g., "poor lubrication" can lead to "bearing wear," which in turn causes "abnormal noise" and "yaw"). Based on the hypothesis of "abnormal drive wheel assembly," the engine searches the graph to verify whether other current data supports a complete causal path: for example, whether the characteristic frequencies of bearing wear are indeed present in the current sound spectrum, and whether the image shows dirt accumulation in the wheel assembly area (potentially corresponding to poor lubrication). Through logical reasoning and evidence matching, the engine generated the final diagnostic conclusion: "The root cause is premature wear of the left drive wheel assembly bearing due to poor lubrication, fault level: fault level", and marked the key evidence.

[0037] Natural Language Interaction and Result Presentation: At this point, the on-site maintenance personnel asked the robot via voice, "Report the current health status." After recognizing the command, the voice interaction agent requested the latest conclusion from the diagnostic agent. It then responded in natural language: "The left drive wheel assembly bearing shows early wear; timely cleaning and grease replenishment are recommended. Here is a spectrum of abnormal sounds and an image of the suspected contamination area for your reference." This response was broadcast via speech synthesis, and simultaneously, the robot's display screen showed a highlighted spectrum analysis graph and a close-up image of the wheel assembly captured by the camera.

[0038] Dynamic manual repair guidance: The system determines that the fault requires manual intervention. The manual repair guidance agent is activated, generating step-by-step voice instructions based on standard maintenance procedures and current diagnostic details from the knowledge graph: "Step 1: Move the robot to the maintenance area and turn off the power… Step 3: Locate the grease filler port on the left drive wheel side." During the critical step of "cleaning the bearing," the system pauses and provides voice confirmation via task flow control: "Please confirm that the old grease has been removed. Continue?" After the maintenance personnel answer "continue," the guidance continues. If the personnel report "part not found" at a certain step, the guidance agent will dynamically adjust and generate alternative guidance: "Please check under the chassis cover, or refer to the exploded view shown on my screen."

[0039] Knowledge Accumulation and System Evolution: The complete closed-loop data from perception, diagnosis, interaction to repair completion, including sensor data, intermediate inference results, and maintenance personnel dialogue and operational feedback, was recorded by the knowledge base agent. This automatically tagged the case with labels such as "walking mechanism," "bearing wear," and "poor lubrication," and stored in a structured manner. Subsequently, through an incremental learning mechanism, the successfully validated feature association of "abnormal sound - image stains - bearing wear" was used to fine-tune the large model, enhancing its sensitivity to similar early wear patterns. Simultaneously, the new experience of "the location of the lubrication port on the left drive wheel assembly of a certain robot model" was updated as a factual relationship in the shared knowledge graph for other robots of the same model to learn from. When network conditions were good, an anonymized summary of this case was uploaded to the cloud and integrated into the group knowledge base; when the network was interrupted, all learning and updates were completed on the robot's local edge computing module, ensuring functional continuity and real-time performance.

[0040] The perception agent runs as a software service within the robot's embedded computer. It includes multiple data acquisition drivers and preprocessing pipelines, performing real-time synchronization, filtering, and initial feature extraction on multi-source data to form standard data packets.

[0041] The diagnostic agent employs a microservice architecture, with its core modules deployed on edge servers. The large model responsible for feature fusion and pattern recognition is loaded onto the GPU, while the knowledge graph query and inference engine, responsible for logical verification and causal inference, resides in memory. The two exchange data via an internal high-speed interface.

[0042] Voice interaction agent and human repair guidance agent: These are a set of collaborative services deployed on the robot itself. They invoke local speech recognition, synthesis modules, and display control interfaces, and maintain communication with the diagnostic agent.

[0043] The Knowledge Base Agent is a distributed service. Its edge instances are deployed on edge servers for high-frequency, real-time knowledge access and local updates. Its cloud-based master instance is responsible for global knowledge integration, version management, and training of complex models. The two synchronize periodically through a secure channel.

[0044] This embodiment provides a robot intelligent self-diagnosis method. Through a multi-agent collaborative architecture, it achieves real-time acquisition and intelligent analysis of multi-source data, integrating logs, status parameters, and multimodal information such as images and sounds. This overcomes the limitations of traditional diagnostic methods that rely on a single data source and produce incomplete results, significantly improving the accuracy and comprehensiveness of fault identification. A self-diagnosis mechanism based on large models and knowledge graph reasoning is introduced, possessing adaptive learning capabilities. It can dynamically adjust diagnostic strategies according to different task scenarios, ensuring the real-time nature and robustness of diagnostic results, making it particularly suitable for complex and changing operating environments. A natural language human-computer interaction channel is constructed through a voice interaction agent, allowing users to directly obtain diagnostic conclusions and processing suggestions via voice. Combined with images, logs, and other information, multimodal presentation is achieved, enhancing the intuitiveness and ease of use of the interaction and improving human-computer collaboration efficiency. After fault confirmation, a step-by-step manual repair guidance plan can be automatically generated, guiding the user through the operation step by step with voice prompts. Dynamic adjustments are made in conjunction with task flow control, lowering the threshold for manual operation, effectively avoiding misoperation, and improving the controllability and safety of the repair process. By using a knowledge base agent to record and consolidate the entire diagnostic and repair process, supporting distributed sharing and continuous learning, the system gains self-evolution capabilities, enabling knowledge sharing and collaborative diagnosis among a group of robots, and further improving the system's intelligence and scalability.

[0045] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0046] According to embodiments of this disclosure, this disclosure also provides a robotic intelligent self-diagnosis device. For example, Figure 2 This is a schematic diagram of the structure of a robot intelligent self-diagnosis device provided in an embodiment of this disclosure. The robot intelligent self-diagnosis device 200 includes: The data acquisition module 210 collects the robot's operating data and multimodal perception information of the external environment in real time, and preprocesses the collected data to obtain preprocessed data. The diagnostic module 220, based on the preprocessed data, uses a hybrid reasoning mechanism to perform fault diagnosis and outputs a graded diagnostic result; The interaction module 230, based on the graded diagnosis results, presents the robot's health status and fault information to the user using natural language interaction.

[0047] This embodiment provides a robot intelligent self-diagnosis device for implementing the aforementioned robot intelligent self-diagnosis method. Therefore, the specific implementation of the robot intelligent self-diagnosis device can be found in the embodiment section of the robot intelligent self-diagnosis method above. For example, the data acquisition module 210, the diagnosis module 220, and the interaction module 230 are respectively used to implement steps 101, 102, and 103 in the aforementioned robot intelligent self-diagnosis method. Therefore, the specific implementation can be referred to the description of the corresponding embodiments, which will not be repeated here.

[0048] Embodiments of this disclosure also provide an electronic device including a memory and a processor, the memory storing a computer program and the processor being configured to run the computer program to perform the steps in any of the above method embodiments.

[0049] Embodiments of this disclosure also provide a computer-readable storage medium storing a computer program configured to perform the steps in any of the above method embodiments when executed.

[0050] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0051] Embodiments of this disclosure also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0052] Embodiments of this disclosure also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0053] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0054] The target detection method provided in this disclosure has been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this disclosure. The descriptions of the embodiments above are only for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this disclosure without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this disclosure.

Claims

1. A method for intelligent self-diagnosis of a robot, characterized in that, include: Multidimensional data of the robot body is collected in real time, and the multidimensional data is preprocessed to obtain preprocessed data; wherein, the multidimensional data includes at least one of operation data and multimodal perception information of the external environment; Based on the preprocessed data, a hybrid reasoning mechanism is used for fault diagnosis to obtain graded diagnosis results; Based on the graded diagnostic results, the robot's status data is presented to the user using natural language interaction; wherein, the status data includes at least one of health status and fault information.

2. The robot intelligent self-diagnosis method according to claim 1, characterized in that, The real-time acquisition of multidimensional data of the robot body, and the preprocessing of the multidimensional data to obtain preprocessed data include: Based on the robot's operating site, real-time data collection is performed on the robot's operating data and multimodal perception information of the external environment. By using data preprocessing methods, the operating data of the robot body and the multimodal perception information of the external environment are cleaned and formatted to obtain preprocessed data.

3. The robot intelligent self-diagnosis method according to claim 2, characterized in that, The fault diagnosis based on the preprocessed data, using a hybrid inference mechanism, yields a graded diagnosis result including: The preprocessed data is used to extract features and recognize patterns using a large model to generate preliminary diagnostic results. The preliminary diagnostic results are used to perform causal inference and logical verification using a knowledge graph, and the faults are classified to obtain a graded diagnostic result.

4. The robot intelligent self-diagnosis method according to claim 3, characterized in that, The process of presenting the robot's status data to the user using natural language interaction based on the graded diagnostic results includes: By using speech recognition and natural language understanding technologies, the system can analyze user questions input via voice. Based on the tiered diagnostic results, a natural language response corresponding to the user's question is generated; The natural language response is output via speech synthesis and may be supplemented with image or text information for multimodal presentation.

5. The robot intelligent self-diagnosis method according to claim 4, characterized in that, Also includes: If the graded diagnostic results indicate that manual intervention is required, a repair guidance plan is generated, and the user is guided to perform the operation through voice prompts according to the repair guidance plan.

6. The robot intelligent self-diagnosis method according to claim 5, characterized in that, The process of generating a repair guidance plan and guiding the user to perform operations via voice prompts includes: The diagnostic conclusions and preset treatment plans are transformed into step-by-step voice guidance prompts; In the preset operation phase, confirmation is obtained from the user through a task flow control mechanism; The guidance strategy will be dynamically adjusted based on user feedback during the repair process. If the feedback indicates that the repair has failed, an alternative repair plan will be generated and the user will be prompted again via voice until the fault is repaired.

7. The robot intelligent self-diagnosis method according to claim 1, characterized in that, Also includes: The data and results generated throughout the entire process of diagnosis and artificial repair are recorded, and the recorded experience data is consolidated through labeling and structured storage methods. Through an incremental learning mechanism, the diagnostic model and knowledge graph are dynamically updated using accumulated experience data.

8. A robotic intelligent self-diagnosis device, characterized in that, include: The data acquisition module collects multidimensional data of the robot body in real time and preprocesses the multidimensional data to obtain preprocessed data; wherein, the multidimensional data includes at least one of operation data and multimodal perception information of the external environment; The fault diagnosis module, based on the preprocessed data, uses a hybrid reasoning mechanism to perform fault diagnosis and obtain a graded diagnosis result; The interaction module, based on the graded diagnostic results, presents the robot's status data to the user using natural language interaction; wherein, the status data includes at least one of health status and fault information.

9. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method described in any one of claims 1-7.

Images