Communication link adaptive method, exoskeleton data acquisition system, medium and product

Abstract

The invention provides a communication link adaptive method, an exoskeleton data acquisition system, a medium and a product, and relates to the technical field of communication control. A multi-communication-mode standby link pool containing a ready queue and a sleep queue is constructed, and collected data such as the working state of data collection equipment, communication link heartbeat data, sensor original data and kinematics calculation pose data are cached in real time. Monitoring the communication state of the current working link and the quality of the collected data through a self-adaptive model: when the link fails, switching to an available link with the highest priority of the ready queue, and generating first compensation data based on the cache data; and when the data is abnormal but the link is not invalid, generating second compensation data. And finally, the output data stream is restored by using the corresponding compensation data to form a continuous smooth data stream and send the continuous smooth data stream to the downstream control system, so that data stream fault and distortion caused by disconnection or data abnormality are avoided, and the pose data output to the downstream control system is ensured to have high continuity and smoothness. And the requirement of precise operation in a complex scene is met.

Description

Technical Field

[0001] This application relates to the field of communication control technology, and in particular to a communication link adaptive method, an exoskeleton data acquisition system, a medium, and a product. Background Technology

[0002] In fields such as robotic arm control, remote manipulation of robotic dexterity hands, and human-robot collaboration, exoskeleton data acquisition systems play a crucial role. These systems are typically used to capture the movement and posture of the human upper limbs or hands and map the sensor data in real time to downstream actuators (such as robotic arms or dexterity hands) to replace manual labor in performing delicate tasks in high-risk or extreme environments such as hazardous chemical experiments, nuclear power plant equipment maintenance, and bomb disposal. Due to the hazardous scenarios and precision operations involved, downstream control systems have extremely high requirements for the real-time communication, data continuity, and fault tolerance of the exoskeleton data acquisition system.

[0003] Currently, most exoskeleton data acquisition systems on the market have relatively simple communication architectures. They typically use a fixed "point-to-point" communication mode, such as connecting the acquisition terminal to the host computer only through a wired interface (such as USB, RS232 serial port) or a single wireless frequency band (such as 2.4G custom protocol, Bluetooth or Wi-Fi).

[0004] However, traditional exoskeleton data acquisition devices have significant drawbacks in their communication methods. Existing exoskeleton data acquisition systems are susceptible to factors such as site limitations, electromagnetic interference, and equipment movement in complex communication environments, resulting in poor communication link stability, increased latency, frequent frame drops or data distortion, and an inability to guarantee high continuity and smoothness of pose data received by downstream control systems. Summary of the Invention

[0005] This application provides a communication link adaptive method, an exoskeleton data acquisition system, a medium, and a product to ensure the stability of the communication link and the stability and continuity of data transmission in the exoskeleton data acquisition system under complex operating scenarios.

[0006] In a first aspect, this application provides a communication link adaptive method applied to an exoskeleton data acquisition system comprising a data acquisition terminal and a host computer processing unit. The method is executed on the host computer processing unit and includes: establishing a backup communication link pool containing multiple communication methods, the backup communication link pool including a ready queue and a dormant queue; real-time caching of acquired data within a set time window, the acquired data including at least the working status of the data acquisition device, communication link heartbeat data, raw sensor data, and kinematic pose data; real-time monitoring of the communication status of the currently working communication link and the data quality of the acquired data using an adaptive model; when the communication status is detected as faulty, selecting the highest priority available communication link from the ready queue for switching, and generating first compensation data based on the acquired data within the set time window during the switching period; when the data quality is detected as abnormal, and the communication status is not faulty, generating second compensation data based on the acquired data within the set time window; using the first compensation data or the second compensation data to perform real-time repair on the output data stream to obtain a continuous smooth data stream, and sending the continuous smooth data stream to the downstream control system.

[0007] By adopting the above technical solution, a multi-communication link pool containing ready and dormant queues is first established, providing a redundant resource foundation for link switching. Secondly, multi-dimensional collected data is cached through time windows, reserving effective data sources for subsequent compensation calculations. Thirdly, combined with an adaptive model to monitor communication status and data quality in real time, accurate anomaly prediction is achieved. When a link fails, a high-priority link is preferentially switched from the ready queue, and first compensation data is generated based on the cached data. When only data is abnormal, second compensation data is generated, and finally, the data flow is repaired through the compensation data. These features work together to solve the problem of easy interruption of traditional single links, and avoid data flow disruptions caused by disconnections or data anomalies through data compensation. This ensures that the pose data output to the downstream control system has high continuity and smoothness, meeting the requirements of precision operation in high-risk scenarios.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the step of monitoring the communication status of the current working communication link and the data quality of the collected data in real time through an adaptive model specifically includes: obtaining the communication status of the current communication link, which includes at least external interruption signals and communication delays; performing streaming computation on the collected data using a sliding window algorithm; if the external interruption signal or the communication delay exceeds a set delay threshold, then determining that the communication status is in a failure state; if the timestamps of the collected data are discontinuous, resulting in frame loss, or if the sensor pose data is found to have abnormal frames that deviate from the predicted trajectory using a 3-sigma algorithm, then determining that the data quality is in an abnormal state.

[0009] By adopting the above technical solution, two core communication status indicators—external interruption signals and communication latency—are first acquired to directly pinpoint whether there are physical disconnections or transmission bottlenecks in the link. Then, a sliding window algorithm is used to perform streaming computation on the collected data, enabling real-time dynamic analysis of data quality and avoiding the lag of offline analysis. Simultaneously, a 3-sigma algorithm is introduced to determine whether the pose data deviates from the predicted trajectory, improving the accuracy of abnormal frame identification. This technical feature, through a dual judgment logic of "hard indicator monitoring of communication status + data quality algorithm verification," achieves accurate differentiation between failure states and abnormal data states, avoiding invalid link switching or overcompensation caused by misjudgments. This ensures the targetedness and effectiveness of subsequent link switching and data compensation operations, improving the overall system's operational stability.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the step of selecting the highest priority available communication link from the ready queue for switching specifically includes: prioritizing each communication method in the ready queue at set intervals; when the communication status is detected as invalid, moving the invalid communication link into the dormant queue and marking it as invalid; selecting the communication method with the highest priority from the ready queue as the target communication method; and adjusting the sampling rate and transmission rate of the acquisition terminal according to the protocol characteristics of the target communication method.

[0011] By adopting the above technical solution, the communication methods in the ready queue are prioritized at set intervals, ensuring that priorities match the dynamic changes in the current communication environment. When a link fails, it is moved to the dormant queue and marked as invalid, avoiding repeated calls to the failed link and improving link selection efficiency. The highest priority link is selected as the target link, ensuring optimal communication performance after the switch. At the same time, the sampling rate and transmission rate are adjusted according to the protocol characteristics of the target link, realizing parameter adaptation between the acquisition terminal and the communication link. These features form a complete link switching process of "dynamic priority sorting - failed link isolation - optimal link selection - adaptive parameter adjustment," which not only shortens the response time of link switching but also ensures the transmission efficiency and stability of the communication link after the switch, reducing the impact of the switching process on the continuity of the data flow.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, the step of generating first compensation data or second compensation data based on the acquired data within the set time window specifically includes: calling a vectorized streaming computing engine to calculate and fit the motion trajectory of the exoskeleton device based on the acquired data; when the communication state fails, during the communication link switching period, interpolating the disconnection time period based on the motion trajectory to generate first compensation data to fill the switching gap; when the acquired data is abnormal, if it is a frame loss abnormality, interpolating the sensor pose data of the missing time points based on the motion trajectory to generate second compensation data to fill the frame loss gap; if it is an abnormal frame abnormality, correcting the pose data of the abnormal frame based on the motion trajectory to generate second compensation data.

[0013] By employing the aforementioned technical solution, a vectorized streaming computing engine is invoked to fit the exoskeleton's motion trajectory. The continuity of the trajectory provides a scientific basis for compensation calculations, avoiding the blind nature of compensation data. In the event of link failure, first compensation data is generated based on trajectory interpolation, filling the data stream gap during the switching window. In the event of data anomalies, interpolation and data correction are performed to generate second compensation data for both lost and abnormal frames. This technical feature, through the logic of "trajectory fitting-classification compensation," ensures that the compensation data closely matches the actual motion trajectory, solving the problem of output data distortion when data is disconnected or abnormal in traditional solutions. It guarantees the smoothness of the compensated data stream, ensuring that the pose commands received by the downstream robotic arm or dexterous hand are continuous and accurate, meeting the needs of precision operations in high-risk scenarios.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the step of using the first compensation data or the second compensation data to perform real-time repair on the output data stream to obtain a continuous and smooth data stream specifically includes: if the communication failure state is in progress, after the connection of the target communication method is successfully established, the generated first compensation data is sent to the downstream control system first, and then the data collected in real time through the target communication method is sent; if the data is in an abnormal state, the second compensation data is used to calibrate the corresponding time period data or abnormal frame data in the output data stream to maintain the temporal continuity of the data stream.

[0015] By adopting the above technical solution, after a successful link failure switchover, the first compensation data is sent first, followed by the real-time acquired data, ensuring the sequential continuity of the data stream and preventing action delays in the downstream system due to disordered receiving order. In case of data anomalies, the second compensation data directly replaces the abnormal data, achieving seamless repair of abnormal data without interrupting normal data transmission. This step, through a differentiated repair strategy of "prioritizing retransmission in failure scenarios + calibration in abnormal scenarios," forms a closed loop with the previously mentioned link switching and compensation data generation stages. This ensures that regardless of link failure or data anomalies, the data stream output to the downstream control system remains continuous and smooth, without any breaks or jumps, guaranteeing the stability and accuracy of the robotic arm and dexterous hand operation in high-risk scenarios.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, the multiple communication methods include at least wired communication and wireless communication. The wired communication method includes at least one of serial communication, USB communication and CANFD communication, and the wireless communication method includes at least one of 2.4G communication, Wi-Fi communication and Bluetooth communication.

[0017] By adopting the above technical solution, both wired and wireless communication methods are configured. Wired communication includes serial ports, USB, and CANFD, while wireless communication includes 2.4G, Wi-Fi, and Bluetooth, thus constructing a diverse link resource pool. Different communication methods possess distinct characteristics: wired communication offers strong resistance to electromagnetic interference and stable transmission, while wireless communication overcomes the limitations of site and equipment movement. The combination of multiple communication methods achieves complementary advantages, solving both the site limitations of traditional single wired links and the susceptibility of single wireless links to electromagnetic interference. This provides a rich redundancy pool, ensuring the availability of usable communication links even in complex and high-risk environments, thereby improving the system's environmental adaptability and robustness.

[0018] Secondly, this application provides a multi-communication link adaptive device based on an exoskeleton data acquisition system, comprising: a link management module for establishing a backup communication link pool containing multiple communication methods and managing a ready queue and a dormant queue; a data caching and monitoring module for caching the acquired data within a set time window in real time and monitoring whether the communication status is in a failed state and whether the data quality is in an abnormal state through an adaptive model; an adaptive switching module for switching to the highest priority available communication link when a failed state is detected; a calculation compensation module for calling a streaming computing engine to fit the motion trajectory, generating first compensation data in a failed state and second compensation data in a data abnormal state; and a data stream repair module for repairing the output data stream using the generated compensation data to output a continuous and smooth data stream to the outside.

[0019] By adopting the above technical solution, the system is divided into five functional modules: link management, data caching and monitoring, adaptive switching, computational compensation, and data flow repair. Each module has a clear division of labor and works in concert. The link management module provides redundant link resources for the system, the data caching and monitoring module achieves accurate anomaly identification, the adaptive switching module completes rapid link switching, the computational compensation module generates high-quality compensation data, and the data flow repair module ensures the continuous and smooth output data flow. The modular architecture achieves functional decoupling, which not only improves the maintainability and scalability of the system, but also forms a complete technical closed loop of "link redundancy - anomaly monitoring - link switching - data compensation - data flow repair" through the orderly cooperation of each module. This ensures that the system operates stably in complex and high-risk environments and meets the high-precision control requirements of downstream actuators.

[0020] Thirdly, this application provides an exoskeleton data acquisition system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the exoskeleton data acquisition system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Fourthly, this application provides a computer-readable storage medium including instructions that, when executed on an exoskeleton data acquisition system, cause the exoskeleton data acquisition system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fifthly, this application provides a computer program product, including a computer program that, when run on an exoskeleton data acquisition system, causes the exoskeleton data acquisition system to perform the method described in the first aspect and any possible implementation thereof.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By adopting the above technical solution, the technical problems of single communication link being susceptible to interference, data flow discontinuity after disconnection, and pose command distortion caused by data anomalies are effectively solved in the existing technology by establishing multiple communication link pools and configuring ready / sleep queues, caching multi-dimensional data collection in time windows, monitoring state quality with adaptive models, and classifying and compensating for link failures / data anomalies. In this way, the pose data output to the downstream control system has high continuity and smoothness, which meets the technical effect of the stability requirements of data transmission for the precision operation of robotic arms and dexterous hands in high-risk scenarios.

[0024] 2. By adopting the above technical solution, the technical means of dynamically prioritizing the communication mode of the ready queue at regular intervals, moving failed links into the sleep queue for isolation, selecting the highest priority link for switching, and adjusting the sampling and transmission rate by matching the target link protocol are effectively used. Therefore, the technical problems of slow link switching response, repeated calling of failed links, and low transmission efficiency caused by mismatched link parameters after switching are effectively solved. This achieves the technical effects of fast and accurate link switching, reducing the impact of the switching process on the continuity of data flow, and improving the adaptive capability of the system's communication links.

[0025] 3. By adopting the above technical solution, the differentiated repair techniques of prioritizing the transmission of compensation data after link failure and then transmitting real-time data, and directly replacing abnormal data with compensation data when data is abnormal, effectively solve the technical problems of disordered data flow timing after link switching and untimely repair of abnormal data leading to stuttering or deviation of downstream actuators in the existing technology. This achieves the technical effect of continuous, smooth and uninterrupted output data flow throughout the entire process, ensuring the stability and accuracy of robotic arms and dexterous hands in high-risk scenarios. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating a communication link adaptive method in an embodiment of this application; Figure 2 This is a schematic diagram of a communication link switching scenario in an embodiment of this application; Figure 3 This is a schematic diagram of data transmission and real-time correction and compensation of abnormal data in a wireless communication scenario in the embodiments of this application; Figure 4 This is a module architecture diagram of the multi-communication link adaptive device based on the exoskeleton data acquisition system in the embodiments of this application; Figure 5 This is a schematic diagram of the physical device structure of an exoskeleton data acquisition system in the embodiments of this application. Detailed Implementation

[0027] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to and includes any or all possible combinations of one or more of the listed items.

[0028] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0029] In the field of robotic arm and dexterous hand control, whether in academic research or industrial applications, much of the R&D work aims to replace humans in performing hazardous tasks, particularly in experiments involving hazardous chemicals and equipment monitoring, maintenance, and assembly in dangerous locations. As the upstream component of robotic arm and dexterous hand control, exoskeleton data acquisition devices need to transmit the sensor data they collect to a computer system or the robotic arm / dexterous hand device for processing in real time during data acquisition tasks. Especially for teleoperation and reinforcement learning training, these data acquisition systems have extremely high requirements for the fault tolerance of the communication link, data accuracy, and data quality. Traditionally, the data transmission link between data acquisition devices and computer systems is usually single. If this communication link fails, is interfered with, or is restricted, it often leads to data delays, decreased accuracy, or even data loss, resulting in data acquisition task failure or failure to achieve the expected results. This application can adapt to the device's operating state. Once the current communication fails, is restricted, or the device state changes, it automatically switches to another working communication link, ensuring the timeliness, integrity, and accuracy of the collected data, improving the system's fault tolerance, and guaranteeing the effectiveness and continuity of data acquisition task execution.

[0030] For ease of understanding, the method provided in this implementation is described in process below. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating a communication link adaptive method in an embodiment of this application.

[0031] S101. Establish a backup communication link pool that includes multiple communication methods. The backup communication link pool includes a ready queue and a dormant queue. The ready queue refers to the set of communication links in the backup communication link pool that are in a real-time standby state and can be switched to working links at any time. An example is a Wi-Fi communication link that is currently not occupied and whose signal strength and transmission rate meet the system requirements. The dormant queue refers to the set of communication links in the backup communication link pool that are in a low-power waiting-to-wake state and do not participate in the link switching selection. An example is a 2.4G communication link that currently has weak signal strength or has experienced transmission failure.

[0032] This step is executed during the initialization phase of the exoskeleton data acquisition system and is a prerequisite resource preparation step to ensure adaptive switching of subsequent links. The host computer processing unit first reviews all communication methods supported by the system. Based on the protocol characteristics, transmission rate, and anti-interference capabilities of different communication methods, it creates corresponding communication link instances for each method. Then, it constructs a two-layer queue structure for the backup communication link pool. Communication link instances that have passed initialization and are ready for immediate use are placed in the ready queue. Links in the ready queue must report their connection status, signal strength, and other data to the host computer in real time. Communication link instances that have completed initialization but do not yet meet the usage conditions, or have experienced transmission anomalies requiring investigation, are placed in the dormant queue. Links in the dormant queue are in a low-power state, and the host computer can perform status checks on them at set intervals. When a link in the dormant queue is detected to have recovered its usability, it is transferred to the ready queue. By constructing a backup communication link pool containing both ready and dormant queues, the host computer can achieve fine-grained management of multiple communication links, providing sufficient redundancy resources for subsequent link switching.

[0033] S102. Real-time cache of the acquired data within a set time window, the acquired data including at least the working status of the data acquisition device, communication link heartbeat data, raw sensor data and kinematic pose data; This step is performed during the normal operation of the exoskeleton data acquisition system, spanning the entire data acquisition and transmission process. It serves as the data source guarantee for subsequent data compensation and anomaly monitoring. The host computer first sets a reasonable time window length based on the downstream control system's requirements for data continuity. The time window length must balance the amount of cached data and real-time performance, avoiding excessive cache pressure due to an overly long window or insufficient compensation data due to an overly short window. Then, a real-time caching mechanism is activated. The acquisition terminal transmits the acquired data to the host computer according to the set sampling rate. The host computer continuously writes the received data acquisition device operating status, communication link heartbeat data, raw sensor data, kinematic pose calculations, and corresponding timestamps into the cache space. The communication link heartbeat data is periodically generated by the host computer and the acquisition terminal through a preset communication protocol. Interactive generation: The host computer sends heartbeat request packets to the acquisition terminal at fixed intervals. Upon receiving the packets, the acquisition terminal immediately returns a response packet containing its own hardware status (such as power supply voltage and sensor connection status) and link transmission delay. The host computer extracts key information from the response packet and integrates it with the difference between the request sending time and the response receiving time to form the communication link heartbeat data. Kinematic pose data is generated by the acquisition terminal's sensors collecting raw human motion data in real time (such as acceleration, angular velocity, and joint rotation angles) and synchronously transmitting this raw sensor data to the host computer. The host computer calls a preset kinematic model (such as a robot kinematic model based on the DH parameter method or a Kalman filter fusion algorithm) to perform real-time calculations on the raw sensor data. Through coordinate transformation, data fusion, and other operations, the position, posture, and motion trajectory information of each joint of the exoskeleton device are obtained, ultimately forming the kinematic pose data.

[0034] Simultaneously, a rolling update mechanism for cached data is activated to remove historical data that exceeds the set time window in real time, ensuring that the cache space always contains the latest and most valid collected data. By caching the collected data within the set time window in real time, the host computer can provide complete data source support for the generation of compensation data when the link fails and for the correction calculation when data is abnormal.

[0035] S103. Monitor the communication status of the current working communication link and the data quality of the collected data in real time through an adaptive model; Among them, the adaptive model refers to the intelligent monitoring model built into the host computer that integrates communication status monitoring algorithm and data quality assessment algorithm, which can dynamically adjust the monitoring threshold according to real-time input data.

[0036] This step is performed during the normal operation of the exoskeleton data acquisition system, synchronously with the data caching process of S102, and is the core monitoring link for achieving adaptive link switching and data compensation. The host computer first initializes the adaptive model, importing preset parameters such as communication delay threshold, frame loss judgment rules, and 3-sigma anomaly judgment threshold; then, it inputs the real-time status data of the current working communication link (external interrupt signals, communication delay) and the data acquired by S102 cache into the adaptive model. For communication status monitoring, the adaptive model captures external interruption signals of the working link in real time. If an interruption signal is detected, the link is directly determined to be in a failure state. Simultaneously, the communication delay is calculated; if the communication delay exceeds a preset threshold, the link is also determined to be in a failure state. For data quality assessment, the adaptive model uses a sliding window algorithm to perform streaming computation on the cached collected data. First, it checks the continuity of data timestamps; if a timestamp discontinuity exists, it is determined to be a frame loss anomaly. Then, the 3-sigma algorithm is used to analyze the sensor pose data, calculating the mean and standard deviation of the data. If a frame of data exceeds the range of [mean - 3 × standard deviation, mean + 3 × standard deviation], it is determined to be an abnormal frame deviating from the predicted trajectory. The adaptive model can dynamically adjust monitoring parameters according to changes in the system operating environment. For example, in environments with strong electromagnetic interference, the threshold for determining communication delay can be appropriately increased to avoid misjudgments, thereby achieving accurate and real-time monitoring of communication status and data quality.

[0037] S104. When the communication status is detected to be in a failed state, select the highest priority available communication link from the ready queue for switching, and generate the first compensation data based on the collected data within the set time window during the switching process. This step is executed the instant the adaptive model detects that the current working communication link is in a failed state, and it is a crucial step in ensuring the continuity of data flow when the link is disconnected. Upon detecting the link failure, the host computer immediately removes the failed working communication link from the working state and moves it to the dormant queue, simultaneously marking it as invalid to prevent it from being mistakenly selected as a working link later. Then, it retrieves the list of communication links in the ready queue and selects the first available communication link as the target link according to a preset priority sorting rule (e.g., in high-risk environments, wired communication links with strong anti-interference capabilities are given the highest priority; in scenarios where equipment needs to be moved, wireless communication links with high flexibility are given the highest priority). Next, it initiates the link switching process, sending instructions to the lower-level acquisition terminal based on the protocol characteristics of the target link to adjust the sampling rate and transmission rate of the acquisition terminal to match the transmission capacity of the target link, completing a seamless link switching.

[0038] A diagram illustrating the communication link switching process in this step will be provided later. Figure 2Introduction; The process of generating the first compensation data will be described in detail in step S105.

[0039] S105. When the data quality is detected to be in an abnormal state and the communication status is not invalid, second compensation data is generated based on the collected data within the set time window. This step is performed when the adaptive model detects that the data quality is abnormal, but the communication status is not invalid. It is a supplementary step to ensure the accuracy and continuity of the data stream.

[0040] After the host computer detects a communication failure or abnormal data quality, it first calls the vectorized streaming computing engine. This engine can convert the sensor pose data within a set time window into vector data and perform parallel computing. Through algorithms such as least squares and Bezier curve fitting, it fits the smooth motion trajectory of the exoskeleton device based on the time-series continuous pose data. During the fitting process, obvious noise data is removed to ensure that the motion trajectory is highly consistent with the actual motion state. When a communication failure is detected, the host computer immediately determines the duration of the link switching window. Based on the fitted motion trajectory, it performs interpolation calculations according to the time nodes within the window to generate first compensation data consistent with the trend of the motion trajectory. This data can accurately fill the data stream gaps during the switching period. When an abnormal data quality is detected, the host computer first determines the type of abnormality. If it is a frame loss abnormality, it locates the time point corresponding to the frame loss and performs interpolation calculations on the pose data of that time point based on the motion trajectory to generate second compensation data to fill the gap. If it is an abnormal frame abnormality, it performs correction calculations on the abnormal frame data that deviates from the trajectory based on the continuous trend of the motion trajectory, generating second compensation data that conforms to the motion law to replace the original abnormal frame. Through the above operations, the compensation data can be seamlessly connected with the normal data before and after, ensuring the continuity and smoothness of the output data stream.

[0041] S106. The output data stream is repaired in real time using the first compensation data or the second compensation data to obtain a continuous smooth data stream, and the continuous smooth data stream is sent to the downstream control system.

[0042] This step is performed after the host computer generates the first or second compensation data, and it is the final data output stage connecting the exoskeleton data acquisition system and the downstream actuators. The host computer first determines, based on the current state of the system, whether to use the first or second compensation data for repair.

[0043] In the case of a link switching scenario under communication failure, after the target link connection is successfully established, the host computer will first send the generated first compensation data to the downstream control system in sequence to fill the data flow gap during the link switching window, and then continue to send new data collected in real time through the target link to ensure that the data flow received by the downstream control system is continuous and without gaps.

[0044] In scenarios where data is abnormal but communication remains intact, the host computer directly uses the generated second compensation data to replace missing or abnormal frames in the output data stream without interrupting the normal data transmission process, ensuring the real-time performance and smoothness of the data stream. After data stream repair is complete, the host computer encapsulates the continuous smooth data stream according to the communication protocol and data format required by the downstream control system and sends it to the downstream control system in real time, providing stable and accurate pose data support for the precise operations of robotic arms and dexterous hands in high-risk scenarios.

[0045] In the above embodiment, by constructing a multi-communication link pool with ready / sleep queues, real-time caching of data acquisition within a set time window, and precise monitoring of communication status and data quality through an adaptive model, the system prioritizes switching the highest priority available link and generates first compensation data when a link fails, and generates second compensation data to repair the data stream when data is abnormal. This achieves adaptive switching of communication links and real-time repair of data streams, effectively solving the problems of poor communication stability, high latency, frame loss, or data distortion caused by existing single communication links being susceptible to site limitations and electromagnetic interference, which cannot guarantee the high continuity and smoothness of pose data in downstream control systems. This enables the exoskeleton data acquisition system to operate stably and reliably in high-risk or extreme environments, ensuring that downstream robotic arms, dexterous hands, and other actuators can receive continuous and smooth pose data, meeting the control requirements of precision operations.

[0046] In some embodiments, during processes S101-S106, the multiple communication methods include at least wired and wireless communication methods. The wired communication methods include at least one of serial communication, USB communication, and CANFD communication, while the wireless communication methods include at least one of 2.4G communication, Wi-Fi communication, and Bluetooth communication. By configuring multiple communication methods combining wired and wireless communication, abundant redundancy resources are provided for adaptive link switching, effectively solving the limitation of single communication methods by specific scenarios and ensuring that the system can select the optimal communication link in different high-risk environments.

[0047] Please refer to the following: Figure 2 , Figure 2 This is a schematic diagram of a communication link switching scenario in an embodiment of this application.

[0048] Figure 2 In this context, (a) represents data stream transmission in a normal communication scenario: Figure 2In (a) of the diagram, the DexCap SDK, as the core processing module running in the host computer processing unit of the exoskeleton data acquisition system, is configured with both 2.4G / BT (wireless communication) and USB (wired communication) communication modes. At this time, the USB link is in active mode, and the 2.4G / BT link is in hibernate mode. During data transmission, the DexCap SDK transmits the acquired data to the host computer via the USB link and simultaneously initiates abnormal data monitoring; ultimately, it outputs a continuous pose data stream to support the refined operation of the downstream dexterous hand. The abnormal data monitoring module continuously polls and monitors the heartbeat packets, data frame integrity, or level signals of the USB link. If any of the following trigger events are detected: The physical connection is broken (e.g., the USB cable is accidentally unplugged); Communication timeout (e.g., no valid response received within a preset time); The packet error rate exceeds the threshold (indicating that the wired transmission line is damaged or severely interfered with). The system will determine that the current working link is invalid, thereby automatically triggering a link switching command, changing the system state from [previous state]. Figure 2 (a) jumps to Figure 2 (b) It should be noted that the triggering events of this application include, but are not limited to, the three triggering events mentioned above, and the specific triggering events can be set according to actual requirements.

[0049] Figure 2 (b) in the diagram represents the link switching and data compensation scenario after USB link failure: Figure 2 In (b) of this example, the original working USB link becomes invalid. At this point, the DexCap SDK executes the following process: Link switching: Mark the failed USB link as invalid and move it to the sleep queue, while waking up the 2.4G / BT link that was originally in sleep state and switching it to the current working link; Data compensation: During the window of link switching, the DexCap SDK calls the abnormal data monitoring module to identify data flow breaks and generates compensation data based on the cached data within the set time window; Data stream repair: Data compensation fills the gaps in the data stream during the switching window, ultimately outputting uninterrupted pose data to ensure the continuity of dexterous hand teleoperation.

[0050] Figure 2 (a) and Figure 2(b) fully presents the closed-loop logic of "multi-link redundancy management - anomaly monitoring - link adaptive switching - data compensation - data flow repair" in the technical solution of this application, which solves the problems of easy failure of traditional single link and easy data flow interruption, and ensures the stability of dexterous hand remote operation in high-risk scenarios.

[0051] Please refer to the following: Figure 3 , Figure 3 This is a schematic diagram of data transmission and real-time correction and compensation of abnormal data in a wireless communication scenario in the embodiments of this application.

[0052] exist Figure 3 Figure (a) illustrates the normal data transmission state in wireless portable operation mode. In this scenario, the user wears the DexCap data acquisition glove to perform a data acquisition task (e.g., a long-cycle task lasting approximately one hour). Considering the need for operational flexibility and portability, the device operates on battery power, with 2.4G or Bluetooth (BT) configured as the primary data transmission path (shown as solid lines in the figure), and the USB link in a non-working or auxiliary charging state. The DexCap SDK receives pose data via the wireless link, while simultaneously running an abnormal data monitoring module in the background to output real-time control commands to the dexterous hand.

[0053] exist Figure 3 (b) illustrates the real-time data stream repair status when wireless communication quality is compromised. It is assumed that the operating environment is complex, with electromagnetic interference sources or other unstable factors causing a degradation in the wireless communication link quality. Figure 3 In (b) of the diagram, the 2.4G / BT link is represented by a dashed line, indicating occasional frame drops or data distortion during data transmission. At this time, the abnormal data monitoring module within the DexCap SDK identifies data quality anomalies using an adaptive model (utilizing a sliding window and 3-sigma algorithm). The system then introduces the data into the "correction / compensation data" stage, calling the vectorized streaming computing engine to fit the motion trajectory and quickly generate second compensation data. The DexCap SDK uses this second compensation data to interpolate and fill in dropped frames or correct and replace abnormal frames, thereby shielding the unstable effects of the wireless link and outputting a continuous and smooth data stream to the dexterous hand, ensuring the stability of the high-degree-of-freedom dexterous hand joint movements.

[0054] In some embodiments, this application also includes a multi-communication link adaptive device based on an exoskeleton data acquisition system. (See [link to relevant documentation]). Figure 4 , Figure 4 This is a module architecture diagram of the multi-communication link adaptive device based on the exoskeleton data acquisition system in this application embodiment, including: The link management module is used to establish a backup communication link pool containing multiple communication methods and to manage the ready queue and the dormant queue. The data caching and monitoring module is used to cache the collected data within a set time window in real time, and to monitor whether the communication status is in a failed state and whether the data quality is in an abnormal state through an adaptive model. The adaptive switching module is used to select the highest priority available communication link for switching when a failure state is detected or the data quality index is lower than the preset index threshold. The calculation and compensation module is used to call the streaming computing engine to fit the motion trajectory, generate the first compensation data in the failure state, and generate the second compensation data in the abnormal data state. The data stream repair module is used to repair the output data stream using the generated compensation data, so as to output a continuous and smooth data stream to the outside.

[0055] Specifically, the aforementioned modules work collaboratively within the host computer processing unit, collectively forming the complete functional architecture of the exoskeleton data acquisition system's adaptive communication links. The link management module first performs initialization configuration, establishing a backup communication link pool encompassing wired and wireless communication methods. Qualified links are added to the ready queue, while temporarily unavailable links are added to the dormant queue, and the status information of the links in the queue is updated in real time. The data caching and monitoring module starts synchronously, caching collected data in real time according to a set time window. Simultaneously, it invokes the adaptive model to monitor the communication status of the currently working communication links and the data quality of the cached data. Once an abnormal state is detected, the corresponding processing mechanism is immediately triggered. When a failure state is detected or the data quality index falls below a preset threshold, the adaptive switching module responds immediately, moving the failed link into dormant mode. The queue selects the highest priority available link from the ready queue to complete the switchover, and simultaneously adjusts the sampling rate and transmission rate of the acquisition terminal. The calculation compensation module starts working when the link fails or the data is abnormal. It calls the streaming computing engine to fit the motion trajectory based on the cached data, generates the first compensation data for the link switching gap, and generates the second compensation data for lost or abnormal frames. The data stream repair module finally completes the data stream repair work. In the case of link failure, the first compensation data is sent first and then the real-time data is transmitted. In the case of data abnormality, the second compensation data is directly used to replace the abnormal part. Finally, a continuous and smooth data stream is output to the outside to ensure the stable operation of the downstream control system.

[0056] The exoskeleton data acquisition system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link to relevant documentation]. Figure 5 This is a schematic diagram of the physical device structure of an exoskeleton data acquisition system in an embodiment of this application.

[0057] It should be noted that, Figure 5 The structure of the exoskeleton data acquisition system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0058] like Figure 5 As shown, the exoskeleton data acquisition system includes a Central Processing Unit (CPU) 501, which can perform various appropriate actions and processes based on programs stored in Read-Only Memory (ROM) 502 or programs loaded from storage section 508 into Random Access Memory (RAM) 503, such as performing the methods described in the above embodiments. The RAM 503 also stores various programs and data required for system operation. The CPU 501, ROM 502, and RAM 503 are interconnected via a bus 504. An Input / Output (I / O) interface 505 is also connected to the bus 504.

[0059] The following components are connected to I / O interface 505: input section 506 including audio input devices, push-button switches, etc.; output section 507 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 508 including a hard disk, etc.; and communication section 509 including a network interface card such as a LAN (Local Area Network) card, modem, etc. Communication section 509 performs communication processing via a network such as the Internet. Drive 510 is also connected to I / O interface 505 as needed. Removable media 511, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 510 as needed so that computer programs read from them can be installed into storage section 508 as needed.

[0060] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 509, and / or installed from removable medium 511. When the computer program is executed by central processing unit (CPU) 501, it performs the various functions defined in the present invention.

[0061] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0062] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0063] Specifically, the exoskeleton data acquisition system of this embodiment includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the communication link adaptive method provided in the above embodiment.

[0064] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the exoskeleton data acquisition system described in the above embodiments; or it may exist independently and not assembled into the exoskeleton data acquisition system. The storage medium carries one or more computer programs that, when executed by a processor of the exoskeleton data acquisition system, cause the exoskeleton data acquisition system to implement the communication link adaptation method provided in the above embodiments.

[0065] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0066] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0067] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A communication link adaptive method, applied to an exoskeleton data acquisition system comprising a data acquisition terminal and a host computer processing unit, characterized in that, The method is executed in the host computer processing unit, and the method includes: Establish a backup communication link pool that includes multiple communication methods, the backup communication link pool including a ready queue and a dormant queue; Real-time caching of collected data within a set time window, the collected data including at least the working status of the data acquisition device, heartbeat data of the communication link, raw sensor data, and kinematic pose data; The adaptive model is used to monitor the communication status of the current working communication link and the data quality of the collected data in real time. When the communication status is detected to be in a failed state, the highest priority available communication link is selected from the ready queue for switching, and first compensation data is generated based on the collected data within the set time window during the switching process. When the data quality is detected to be in an abnormal state and the communication status is not invalid, second compensation data is generated based on the collected data within the set time window. The output data stream is repaired in real time using the first compensation data or the second compensation data to obtain a continuous smooth data stream, and the continuous smooth data stream is sent to the downstream control system.

2. The method according to claim 1, characterized in that, The step of monitoring the communication status of the current working communication link and the data quality of the collected data in real time through an adaptive model specifically includes: Obtain the current communication status of the communication link, which includes at least external interruption signals and communication delays; The collected data is processed using a sliding window algorithm; If the external interruption signal is detected or the communication delay exceeds the set delay threshold, the communication state is determined to be in a failed state. If the timestamps of the collected data are found to be discontinuous, resulting in frame loss, or if the sensor pose data is found to have abnormal frames that deviate from the predicted trajectory using the 3-sigma algorithm, then the data quality is determined to be in an abnormal state.

3. The method according to claim 1, characterized in that, The step of selecting the highest priority available communication link from the ready queue for switching specifically includes: Priority sorting is performed on each communication method in the ready queue at set intervals; When the communication status is detected as invalid, the invalid communication link is moved into the dormant queue and marked as invalid. Select the communication method with the highest priority from the ready queue as the target communication method; Adjust the sampling rate and transmission rate of the acquisition terminal according to the protocol characteristics of the target communication method.

4. The method according to claim 1, characterized in that, The step of generating first compensation data or second compensation data based on the collected data within the set time window specifically includes: The vectorized streaming computing engine is invoked to calculate and fit the motion trajectory of the exoskeleton device based on the collected data. When the communication state fails, during the communication link switching period, the disconnection time period is interpolated based on the motion trajectory to generate first compensation data to fill the switching gap. When the collected data is abnormal, if it is a frame loss, then the sensor pose data at the missing time points is interpolated based on the motion trajectory to generate second compensation data to fill the frame loss gap. If the abnormal frame is abnormal, then based on the motion trajectory, the pose data of the abnormal frame is corrected and a second compensation data is generated.

5. The method according to claim 1, characterized in that, The step of using the first compensation data or the second compensation data to perform real-time repair on the output data stream to obtain a continuous and smooth data stream specifically includes: If the communication fails, after the connection of the target communication method is successfully established, the first compensation data generated will be sent to the downstream control system first, and then the data collected in real time through the target communication method will be sent. If the data is in an abnormal state, the corresponding time period data or abnormal frame data in the output data stream is calibrated using the second compensation data to maintain the temporal continuity of the data stream.

6. The method according to claim 1, characterized in that, The multiple communication methods include at least wired communication and wireless communication. The wired communication methods include at least one of serial communication, USB communication and CANFD communication. The wireless communication methods include at least one of 2.4G communication, Wi-Fi communication and Bluetooth communication.

7. A multi-communication link adaptive device based on an exoskeleton data acquisition system, characterized in that, include: The link management module is used to establish a backup communication link pool containing multiple communication methods and to manage the ready queue and the dormant queue. The data caching and monitoring module is used to cache the collected data within a set time window in real time, and to monitor whether the communication status is in a failed state and whether the data quality is in an abnormal state through an adaptive model. The adaptive switching module is used to select the highest priority available communication link for switching when a failure state is detected or the data quality index is lower than the preset index threshold. The calculation and compensation module is used to call the streaming computing engine to fit the motion trajectory, generate the first compensation data in the failure state, and generate the second compensation data in the abnormal data state. The data stream repair module is used to repair the output data stream using the generated compensation data, so as to output a continuous and smooth data stream to the outside.

8. An exoskeleton data acquisition system, characterized in that, The exoskeleton data acquisition system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the exoskeleton data acquisition system to perform the method as described in any one of claims 1-6.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the exoskeleton data acquisition system, the exoskeleton data acquisition system performs the method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is run on the exoskeleton data acquisition system, the exoskeleton data acquisition system performs the method as described in any one of claims 1-6.

Images