A walking gait control method and system for a quadruped robot

Abstract

The application discloses a walking gait control method and system for a quadruped robot, and relates to the technical field of robot gait control.The method comprises the following steps: synchronously collecting robot body and working tool data and aligning into a linkage chain; collecting environmental characteristics to generate a working allowable range constraint; generating a preliminary gait control instruction and calibrating; collecting extreme environment and component data to identify characteristics, matching a self-reinforcing gait scheme and superimposing compensation parameters, and outputting a gait control instruction.The application realizes dynamic collaborative adaptation of gait control and working tools, working environment and extreme working conditions through the construction of a full-process closed-loop control of data linkage, constraint generation, instruction optimization and extreme adaptation, and can improve the collaboration and environmental robustness of gait control compared with the prior art, thus solving the problems of disconnection between body movement and tool operation, lagging adaptation to environmental changes, and insufficient stability under extreme working conditions in traditional quadruped robot gait control.

Description

Technical Field

[0001] This invention relates to the field of robot gait control technology, and in particular to a walking gait control method and system for quadruped robots. Background Technology

[0002] The walking gait control method and system for quadruped robots are used to plan and coordinate the movement trajectory and timing of the quadruped robot's limbs, dynamically adjust gait parameters according to terrain and task requirements, improve the robot's adaptability in complex environments, and optimize energy consumption and response speed.

[0003] Currently, most gait control methods and systems used for quadruped robots fail to achieve dynamic linkage between robot body movement and tool status. They lack precise quantification mechanisms for the constraint transformation of working environment characteristics and lack adaptive gait adjustment strategies for extreme environments and component performance degradation. This results in insufficient coordination between gait control and tool operation, delayed adaptation to environmental changes, and problems such as decreased stability, deviation in work accuracy, and increased component wear under extreme working conditions.

[0004] Therefore, a walking gait control method and system for quadruped robots is proposed to solve the above problems. Summary of the Invention

[0005] The main objective of this invention is to provide a walking gait control method and system for quadruped robots to solve the problems mentioned in the background above.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: a walking gait control method for a quadruped robot, the method comprising the following steps: Step S1. Synchronously collect robot body status data and work tool status data, and align them according to timestamps to form a linkage data chain; Step S2. Collect the physical characteristics of the working environment and generate the allowable range constraints for tool operation in a programmed manner based on the first preset mapping relationship; The first preset mapping relationship is a mapping relationship between environmental features and tool constraints; Step S3. Receive the linkage data link and the allowable range constraints of the tool operation, extract the core constraint items in a programmatic manner, and call the gait parameter combination according to the second preset mapping relationship to generate preliminary gait control instructions; The second preset mapping relationship is the mapping relationship between core motion constraints and gait parameters; Step S4. Collect gait execution feedback data, compare it with the preset co-adaptation standard in a programmed manner, adjust gait parameters, and output calibrated gait control commands; Step S5. Collect physical characteristics of extreme environments and operational data of robot components, and identify the type of extreme environment, the scope of influence, and the performance degradation characteristics of components based on the third preset mapping relationship; The third preset mapping relationship is a mapping relationship between physical characteristics, component operating data and environmental type, and performance degradation characteristics; Step S6. Receive the recognition result, match the target scheme from the self-reinforcing gait scheme library according to the fourth preset mapping relationship, and superimpose compensation parameters according to the fifth preset mapping relationship to generate optimized gait control instructions; The fourth preset mapping relationship is the mapping relationship between environment type, component attenuation characteristics and self-reinforcing gait strategy; The fifth preset mapping relationship is the mapping relationship between the degree of component performance degradation and the compensation parameter.

[0007] Preferably, step S1 includes: The robot's body state data is generated by collecting the movement trajectory and rotation angle signals of the leg joints through joint sensors and the balance posture and center of gravity distribution offset trend of the torso through trunk inertial sensors. The tool's contact force signal is collected by the pressure sensor at the end of the tool, the tool's posture angle is collected by the joint angle sensor, and the tool's working mode switching signal is read to form the tool's status data. Align the body status data and tool status data by timestamp, add sensor connection status and signal strength identifiers as data validity verification bits, and form a linked data chain.

[0008] Preferably, step S2 includes: The system collects spatial information about the target location and obstacles using a visual sensor, relative distance information between the target and the machine body using a distance sensor, and operational accuracy information using a precision sensor. Based on the first preset mapping relationship, spatial information is transformed into fuselage attitude limits and spatial constraints, relative distance information is transformed into stride limits, and operational accuracy information is transformed into center of gravity movement boundaries, which are then integrated to form the allowable range constraints for tool operation.

[0009] Preferably, steps S3 and S4 include: The core constraint terms are extracted from the allowable range constraints of tool operation, including fuselage stability constraints, field of view continuity constraints, and space limitation constraints; Based on the second preset mapping relationship, the gait parameter combination of low center of gravity and wide support, and synchronous turning stride adaptation is called to adjust the support phase duration, center of gravity position, stride distribution and fuselage attitude, and output preliminary gait control commands. Collect fuselage posture fluctuation data and joint motion offset data as gait execution feedback data, and compare them with preset collaborative adaptation standards; Adjust the support phase stiffness and center of gravity offset parameters, and output the calibrated gait control command.

[0010] Preferably, steps S5 and S6 include: Ambient temperature, air humidity, dust concentration, and ground movement resistance are collected by temperature sensors, humidity sensors, dust sensors, and resistance sensors, respectively, as physical characteristics of extreme environments; The robot collects joint motion resistance and motion smoothness data through joint sensors, posture stability data through torso inertial sensors, and operating current and heat generation temperature through drive unit sensors, and integrates them to form robot component operation data. Based on the third preset mapping relationship, the extreme environment type and impact range are determined, and the robot component operation data are compared with the preset component operation standards to determine the component performance degradation characteristics. Based on the fourth preset mapping relationship, at least one self-reinforcing scheme is matched from the high-frequency small-amplitude swing phase, the high ground clearance support phase, and the low-energy heat dissipation gait. The ratio of the swing phase to the support phase and the sequence of joint force application are adjusted according to the matching self-reinforcing scheme, and the compensation parameters of joint force intensity, gait rhythm and support phase ratio are determined according to the fifth preset mapping relationship. The compensation parameters are superimposed on the self-reinforcing gait command to generate an optimized gait control command.

[0011] Preferably, the system includes a linkage data acquisition module, a work environment constraint conversion module, a gait coordination control command generation module, a coordination deviation calibration module, an extreme environment and component status recognition module, and a gait self-reinforcement and compensation module; The linkage data acquisition module is used to synchronously collect robot body status data and working tool status data, and align them according to timestamps to form a linkage data chain; The work environment constraint conversion module is used to collect the physical characteristics of the work environment and, based on the first preset mapping relationship, programmatically generate the allowable range constraints for tool operation; The gait coordination control instruction generation module is used to receive the linkage data link and the allowable range constraints of the tool operation, programmatically extract the core constraint items, and call the gait parameter combination according to the second preset mapping relationship to generate preliminary gait control instructions; The coordination deviation calibration module is used to collect gait execution feedback data, compare it with the preset coordination adaptation standard in a programmed manner, adjust gait parameters, and output calibrated gait control commands. The extreme environment and component status identification module is used to collect physical characteristics of extreme environments and robot component operation data, and identify the type of extreme environment, the scope of influence, and component performance degradation characteristics based on a third preset mapping relationship; The gait self-reinforcement and compensation module is used to receive the recognition results, match the target scheme from the self-reinforcement gait scheme library according to the fourth preset mapping relationship, and superimpose compensation parameters according to the fifth preset mapping relationship to generate optimized gait control commands.

[0012] Preferably, the linkage data acquisition module is specifically used for: It receives signals transmitted from joint sensors and torso inertial sensors to form body state data; It receives signals from the tool end pressure sensor and the tool joint angle sensor, as well as the tool working mode switching signal, and forms tool status data; Add timestamps and validity check bits to the ontology status data and tool status data, align the two types of data according to the timestamps to form a linked data chain and output it.

[0013] Preferably, the work environment constraint conversion module is specifically used for: It receives signals transmitted from vision sensors, distance sensors, and precision sensors; Based on the first preset mapping relationship, visual and distance signals are converted into spatial constraints, posture constraints and stride limits, and accuracy signals are converted into center of gravity movement boundaries, which are then integrated into the allowable range constraints for the tool's operation.

[0014] Preferably, the gait coordination control command generation module is specifically used for: Receive the linkage data link and verify the validity of the data; Extract fuselage stability constraints, field of view continuity constraints, and spatial limitation constraints from the tool operation allowable range constraints; Based on the second preset mapping relationship, the corresponding gait parameter combination is invoked to adjust the support phase duration, center of gravity position, stride distribution and fuselage attitude, and output preliminary gait control commands.

[0015] Preferably, the gait self-reinforcing and compensation module is specifically used for: Based on the fourth preset mapping relationship, at least one self-reinforcing scheme is matched from the high-frequency small-amplitude swing phase, the high ground clearance support phase, and the low-energy heat dissipation gait. Adjustments are made based on the matching self-reinforcing scheme, and compensation parameters for joint force intensity, gait rhythm, and support ratio are determined based on the fifth preset mapping relationship. The compensation parameters are superimposed on the self-reinforcing gait command to generate an optimized gait control command.

[0016] The present invention has the following beneficial effects: 1. This invention synchronously collects robot body state data and tool state data, adds validity check bits to form a linked data chain, and combines the physical characteristics of the working environment to programmatically generate allowable range constraints for tool operation. After core constraint extraction, gait parameter calling, feedback calibration, and self-reinforcing compensation after extreme environment and component state identification, a closed-loop control system is constructed for the entire process of data linkage, constraint generation, instruction optimization, and extreme adaptation. This achieves dynamic collaborative adaptation of gait control with the working tool, working environment, and extreme working conditions. Compared with existing technologies, it can significantly improve the coordination and environmental robustness of gait control. Therefore, it can solve the problems of disconnect between body movement and tool operation, lag in environmental change adaptation, and insufficient stability under extreme working conditions in traditional quadruped robot gait control.

[0017] 2. This invention sets up validity verification bits for sensor connection status and signal strength in the linkage data chain, transforming the spatial information, relative distance, and operational accuracy of the working environment into specific constraints such as body posture limits, stride limits, and center of gravity movement boundaries according to a preset mapping relationship. By combining the core constraints, the invention accurately calls the gait parameter combination and dynamically adjusts the support phase stiffness and center of gravity offset through feedback data, thereby achieving precise optimization of gait parameters. Compared with existing technologies, this invention can improve the accuracy and reliability of gait control, thus solving the problem of data interference causing command deviation and fuzzy environmental constraint conversion in traditional quadruped robot gait control, which in turn affects the operational accuracy of quadruped robots.

[0018] 3. This invention collects physical characteristics of extreme environments and robot component operation data through multi-dimensional sensors. Based on a preset mapping relationship, it accurately identifies the type of extreme environment, the scope of influence, and the performance degradation characteristics of components. It matches a targeted gait scheme from a self-reinforcing gait scheme library and superimposes compensation parameters adapted to the degree of component performance degradation. This achieves active adaptation of gait and performance degradation compensation in extreme environments. Compared with existing technologies, this invention can significantly enhance the robot's operational continuity and stability in extreme environments. Therefore, it can solve the problem of traditional quadruped robots lacking targeted gait strategies in complex and extreme environments, leading to operational interruptions due to component performance degradation. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the walking gait control method for a quadruped robot according to the present invention; Figure 2 This is a schematic diagram of the walking gait control system architecture for a quadruped robot according to the present invention. Detailed Implementation

[0020] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0021] Example 1, please refer to Figure 1 As shown: A gait control method for a quadruped robot, the method includes the following steps: Step S1. Synchronously collect robot body status data and work tool status data, and align them according to timestamps to form a linkage data chain; Step S2. Collect the physical characteristics of the working environment and generate the allowable range constraints for tool operation in a programmed manner based on the first preset mapping relationship; The first preset mapping relationship is the mapping relationship between environmental characteristics and tool constraints; Step S3. Receive the linkage data link and the allowable range constraints of the tool operation, extract the core constraint items in a programmatic manner, and call the gait parameter combination according to the second preset mapping relationship to generate preliminary gait control instructions; The second preset mapping relationship is the mapping relationship between core motion constraints and gait parameters; Step S4. Collect gait execution feedback data, compare it with the preset co-adaptation standard in a programmed manner, adjust gait parameters, and output calibrated gait control commands; Step S5. Collect physical characteristics of extreme environments and operational data of robot components, and identify the type of extreme environment, the scope of influence, and the performance degradation characteristics of components based on the third preset mapping relationship; The third preset mapping relationship is the mapping relationship between physical characteristics, component operating data and environmental type, and performance degradation characteristics; Step S6. Receive the recognition result, match the target scheme from the self-reinforcing gait scheme library according to the fourth preset mapping relationship, and superimpose compensation parameters according to the fifth preset mapping relationship to generate optimized gait control instructions; The fourth preset mapping relationship is the mapping relationship between environment type, component attenuation characteristics and self-reinforcing gait strategy; The fifth preset mapping relationship is the mapping relationship between the degree of component performance degradation and the compensation parameters.

[0022] Step S1 includes: The robot's body state data is generated by collecting the movement trajectory and rotation angle signals of the leg joints through joint sensors and the balance posture and center of gravity distribution offset trend of the torso through trunk inertial sensors. The tool's contact force signal is collected by the pressure sensor at the end of the tool, the tool's posture angle is collected by the joint angle sensor, and the tool's working mode switching signal is read to form the tool's status data. Align the body status data and tool status data by timestamp, add sensor connection status and signal strength identifiers as data validity verification bits, and form a linked data chain.

[0023] Step S2 includes: The system collects spatial information about the target location and obstacles using a visual sensor, relative distance information between the target and the machine body using a distance sensor, and operational accuracy information using a precision sensor. Based on the first preset mapping relationship, spatial information is transformed into fuselage attitude limits and spatial constraints, relative distance information is transformed into stride limits, and operational accuracy information is transformed into center of gravity movement boundaries, which are then integrated to form the allowable range constraints for tool operation.

[0024] Steps S3 and S4 include: The core constraint terms are extracted from the allowable range constraints of tool operation, including fuselage stability constraints, field of view continuity constraints, and space limitation constraints; Based on the second preset mapping relationship, the gait parameter combination of low center of gravity and wide support phase and synchronous turning stride adaptation is called to adjust the support phase duration, center of gravity position, stride distribution and fuselage attitude, and output the initial gait control command. Collect fuselage posture fluctuation data and joint motion offset data as gait execution feedback data, and compare them with preset collaborative adaptation standards; Adjust the support phase stiffness and center of gravity offset parameters, and output the calibrated gait control command.

[0025] Steps S5 and S6 include: Ambient temperature, air humidity, dust concentration, and ground movement resistance are collected by temperature sensors, humidity sensors, dust sensors, and resistance sensors, respectively, as physical characteristics of extreme environments; The robot collects joint motion resistance and motion smoothness data through joint sensors, posture stability data through torso inertial sensors, and operating current and heat generation temperature through drive unit sensors, and integrates them to form robot component operation data. Based on the third preset mapping relationship, the type and scope of extreme environment are determined, and the performance degradation characteristics of the robot components are determined by comparing the robot component operation data with the preset component operation standards. Based on the fourth preset mapping relationship, at least one self-reinforcing scheme is matched from the high-frequency small-amplitude oscillation phase, the high ground clearance support phase, and the low-energy heat dissipation gait. The basic ratio of the swing phase to the support phase and the sequence of joint force application are adjusted according to the matching self-reinforcing scheme. The compensation parameters of joint force application intensity, gait rhythm and support phase ratio compensation coefficient are determined according to the fifth preset mapping relationship. The compensation parameters are superimposed on the self-reinforcing gait command to generate an optimized gait control command.

[0026] Example 2, please refer to Figure 2As shown: A walking gait control system for a quadruped robot, the system includes a linkage data acquisition module, a working environment constraint conversion module, a gait cooperative control command generation module, a cooperative deviation calibration module, an extreme environment and component state recognition module, and a gait self-reinforcement and compensation module; The linkage data acquisition module is used to synchronously collect robot body status data and working tool status data, and align them according to timestamps to form a linkage data chain; The linked data acquisition module is specifically used for: It receives signals transmitted from joint sensors and torso inertial sensors to form body state data; It receives signals from the tool end pressure sensor and the tool joint angle sensor, as well as the tool working mode switching signal, and forms tool status data; Add timestamps and validity check bits to the ontology status data and tool status data, align the two types of data according to the timestamps to form a linked data chain and output it.

[0027] Furthermore, the linkage data acquisition module is used to synchronously acquire the status data of the robot body and the working tool, and form a linkage data chain by aligning the timestamps. Specifically, the above control method step S1 is implemented, and the workflow is as follows: First, status data is acquired through multiple types of sensors. The joint sensors adopt magnetostrictive displacement sensors, which are deployed on the 6 leg joints. The displacement signal is acquired at a sampling frequency of 120Hz. After signal conditioning, ADC conversion and moving average filtering, it is converted into a joint angle of 0 to 360° through a preset transmission ratio.

[0028] The trunk inertial sensor is installed at the center of the quadruped robot's trunk and collects three-dimensional acceleration and angular velocity at a sampling frequency of 100Hz. The center of gravity offset is calculated after Kalman filtering. The tool end sensors include pressure sensors and angle encoders, which collect force signals, tool joint angles and mode switching signals, respectively. The signals are integrated after temperature compensation and median filtering.

[0029] Data synchronization and encapsulation are then performed. A 17-byte timestamp is generated using the built-in real-time clock and embedded in two types of data frames. The validity check bit contains 2 bytes, including a connection status bit and a signal strength bit. A signal strength ≥128 is considered normal. Synchronization data groups are matched according to the 8ms timestamp threshold and spliced ​​into an 85-byte linkage data chain. The data is then transmitted in batches to the work environment constraint conversion module via the CANFD bus. At the same time, nearly 24 hours of data are backed up in 64MB Flash to ensure data traceability.

[0030] Furthermore, the linkage data acquisition module has self-testing and adaptation functions: each frame of data is checked for frame header, frame tail and length. If the check fails for 10 consecutive frames or ≥3 sensors are disconnected, an alarm is output and the backup sensor is activated. It supports online adjustment of parameters such as sampling frequency and filtering window through the USART interface to adapt to the data acquisition needs of different operating scenarios and ensure the continuity and flexibility of the acquisition process.

[0031] In summary, this embodiment specifically implements step S1 of the above control method. The linkage data acquisition module establishes a complete closed loop of data acquisition, processing, synchronization and output through standardized sensor selection and signal processing procedures. It ensures data reliability by using timestamp alignment and validity verification, and improves robustness through self-testing function. The final generated structured linkage data chain provides accurate and real-time basic data support for subsequent constraint transformation and gait command generation.

[0032] Example 3, please refer to Figure 2 As shown: A walking gait control system for a quadruped robot. The work environment constraint conversion module is used to collect the physical characteristics of the work environment and, based on the first preset mapping relationship, programmatically generate the allowable range constraints for tool operation; The job environment constraint conversion module is specifically used for: It receives signals transmitted from vision sensors, distance sensors, and precision sensors; Based on the first preset mapping relationship, visual and distance signals are converted into spatial constraints, posture constraints and stride limits, and accuracy signals are converted into center of gravity movement boundaries, which are then integrated into the allowable range constraints for tool operation.

[0033] Furthermore, the work environment constraint transformation module is used to collect the physical features of the work environment and programmatically generate the allowable range constraints of the tool's work according to the first preset mapping relationship. Specifically, it implements step S2 of the above control method. The workflow is as follows: First, the physical features of the environment are collected through multiple types of sensors. The vision sensor adopts a ToF depth camera, which is installed on the robot's head gimbal. It collects three-dimensional point cloud data at a frame rate of 30fps. After coordinate transformation and pass-through filtering to remove invalid points, in the robot's body coordinate system, with the center of the torso as the origin, the X-axis moves forward, the Y-axis moves laterally, and the Z-axis moves upward. The target and obstacles are distinguished by the spatial distance grouping method, and the target center coordinates and obstacle bounding box information are extracted.

[0034] The distance sensor uses a laser rangefinder, symmetrically mounted at the front of the torso, to collect relative distance data at a baud rate of 9600bps. The shortest distance between the target and the obstacle is obtained by sliding median filtering. The accuracy sensor uses a capacitive displacement sensor, mounted at the end of the tool, to collect vertical distance data at a sampling frequency of 150Hz. The standard deviation Δh of 50 frames is calculated to reflect the operation accuracy.

[0035] Subsequently, the constraint parameters are transformed according to the first preset mapping relationship. This mapping relationship is stored in the module's Flash memory in the form of a four-dimensional data table and supports online updates: the safety distance of the obstacle bounding box is converted into spatial constraints, attitude constraints are set according to the distance to the obstacle, the upper limit of stride is set according to the target distance Dt, and the center of gravity movement boundary is determined according to Δh; conflicting constraints are processed with the priority of spatial constraints being greater than attitude constraints, and integrated into a JSON format constraint file, which is output to the gait coordination control command generation module through the Ethernet interface, and at the same time, the data of the past 30 days is backed up on a 32GB SD card.

[0036] Furthermore, the work environment constraint conversion module has dynamic update and anomaly handling functions: it updates constraint parameters every 200ms, triggers emergency updates and outputs warnings when environmental characteristics change abruptly; when a sensor has no data for 10 consecutive frames, it automatically switches to a backup sensor to generate simplified constraints; it supports adjusting the mapping relationship threshold through the Web configuration interface to adapt to different work scenarios such as precision assembly and outdoor inspection, ensuring the continuity and flexibility of constraint generation.

[0037] In summary, this embodiment specifically implements step S2 of the above control method. The work environment constraint conversion module establishes a quantitative mapping from environmental features to constraint parameters by clarifying sensor selection and signal processing procedures. It ensures the rationality of constraints through conflict handling and data verification. The dynamic update function improves the adaptability to complex environments. The final generated tool operation allowable range constraint provides clear and safe boundary conditions for subsequent gait command generation, taking into account both work safety and accuracy requirements.

[0038] Example 4, please refer to Figure 2 As shown: A walking gait control system for a quadruped robot, wherein the gait cooperative control command generation module is used to receive the linkage data link and the allowable range constraints of the tool operation, programmatically extract the core constraint items, and call the gait parameter combination according to the second preset mapping relationship to generate preliminary gait control commands; The gait coordination control command generation module is specifically used for: Receive the linkage data link and verify the validity of the data; Extract fuselage stability constraints, field of view continuity constraints, and spatial limitation constraints from the tool operation allowable range constraints; Based on the second preset mapping relationship, the corresponding gait parameter combination is called to adjust the support phase duration, center of gravity position, stride distribution and fuselage attitude, and outputs preliminary gait control commands.

[0039] Furthermore, the gait coordination control command generation module is used to receive the linkage data link and the tool's allowable range constraints, programmatically extract the core constraint items and call the gait parameter combination to generate preliminary gait control commands. Specifically, the workflow of step S3 of the above control method is as follows: First, it receives data from the upstream module and verifies its validity. It receives an 85-byte data link from the linkage data acquisition module through the CANFD bus, verifies the frame header and tail, length and 2-byte validity check bit, and discards and records invalid data after marking it. At the same time, it receives the JSON format constraint file from the work environment constraint conversion module through the Ethernet interface, parses it and extracts the core constraint items, including fuselage stability constraints, field of view continuity constraints, and spatial limitation constraints. The fuselage stability constraint refers to the pitch or roll angle limit, the field of view continuity constraint refers to the tool angle change rate ≤ 5° / s, the spatial limitation constraint refers to the effective range of the work area, and the stride upper limit is an independent constraint item.

[0040] Subsequently, the gait parameter combination is invoked according to the second preset mapping relationship. This mapping relationship is stored in the module's Flash memory in the form of a three-dimensional data table of core constraint items, gait parameter types, and parameter ranges, and supports online updates by the host computer. For fuselage stability constraints, the combination of low center of gravity and wide support is invoked, with the low center of gravity being 0.15~0.18m and the wide support being 0.32~0.36m. The continuous field of vision constraint is adapted to the synchronous steering stride attenuation coefficient, and the spatial limitation constraint adjusts the support phase duration and stride distribution. After integrating the parameters, the instructions are encapsulated according to the actuator protocol and output to the cooperative deviation calibration module through the CANFD bus. At the same time, the instructions are stored in the 32MB Flash backup for nearly 12 hours of instruction logs.

[0041] Furthermore, the gait coordination control command generation module has parameter optimization and scenario adaptation functions: when there is no constraint conflict, the parameters are fine-tuned according to the principle of minimum energy consumption; when there is a conflict, the parameters are adjusted according to the priority of stability > field of view > space. It supports preset parameter templates through the USART interface to adapt to scenarios of precision assembly or rapid movement. Precision assembly refers to small stride and long support phase, and rapid movement refers to large stride and short support phase. The command generation delay is ≤10ms to ensure the real-time performance of gait response.

[0042] In summary, this embodiment specifically implements step S3 of the above control method. The gait coordination control command generation module establishes a precise mapping between constraint items and gait parameters through data validity verification and core constraint extraction. The parameter optimization and scenario adaptation functions improve the flexibility of the commands. The final generated preliminary gait control commands not only meet the operational constraint requirements, but also provide a standardized basis for subsequent deviation calibration, ensuring the stability and adaptability of gait control.

[0043] Example 5, please refer to Figure 2As shown: A walking gait control system for a quadruped robot. The cooperative deviation calibration module is used to collect gait execution feedback data, compare it with the preset cooperative adaptation standard in a programmed manner, adjust the gait parameters, and output the calibrated gait control command.

[0044] Furthermore, the collaborative deviation calibration module is used to collect gait execution feedback data, compare it with the preset collaborative adaptation standard, adjust the parameters, and output the calibrated gait control command. Specifically, step S4 of the above control method is implemented, and the workflow is as follows: First, gait execution feedback data is collected. The joint sensor and torso inertial sensor of the linkage data acquisition module are used to collect joint motion offset data at a sampling frequency of 120Hz and fuselage attitude fluctuation data at a frequency of 100Hz. After removing execution noise by moving average filtering, the joint offset Δθ and fuselage attitude fluctuation value Δφ are extracted to form a feedback data frame. The joint offset Δθ is the difference between the actual angle and the commanded angle, and the fuselage attitude fluctuation value Δφ is the pitch or roll angle fluctuation amplitude. The feedback data frame includes a frame header 0xDD, offset data, fluctuation data, and checksum.

[0045] Subsequently, data comparison and parameter adjustment are performed. The preset collaborative adaptation standard is stored in the module's Flash memory in the form of a data table of parameter types and allowable thresholds, such as the allowable threshold for joint offset ≤0.5° and the allowable threshold for attitude fluctuation ≤1°. The feedback data is compared with the standard. If Δθ exceeds the threshold, the support phase stiffness is adjusted proportionally, with an adjustment range of 1.2 to 2.0 N·m / rad. If Δφ exceeds the threshold, the center of gravity offset is finely adjusted, with an adjustment range of ±0.005m. The adjusted parameters are integrated, and the calibration command is repackaged according to the actuator protocol and output to the execution unit through the CANFD bus. At the same time, the calibration log of nearly 8 hours is stored in the 32MB Flash backup.

[0046] Furthermore, the collaborative deviation calibration module has deviation warning and adaptation optimization functions: when the feedback data exceeds the threshold for 5 consecutive frames, it outputs a warning signal to the main control unit; it supports online adjustment of the collaborative adaptation standard threshold through the USART interface to adapt to different operational accuracy requirements, with the threshold tightened to 0.3° for precision operations and relaxed to 0.8° for outdoor operations, and the calibration delay ≤8ms, ensuring real-time gait correction.

[0047] In summary, this embodiment specifically implements step S4 of the above control method. The collaborative deviation calibration module establishes a linkage between instruction execution and parameter adjustment through a closed-loop feedback mechanism. The deviation warning function improves control safety, the adaptation and optimization capability enhances scenario compatibility, and the final output calibrated gait control command corrects the execution deviation, ensuring the gait stability and accuracy of the robot in complex operations.

[0048] Example 6, please refer to Figure 2As shown: A walking gait control system for a quadruped robot, wherein an extreme environment and component state recognition module is used to collect physical characteristics of extreme environments and operational data of robot components, and to identify the type of extreme environment, the range of influence, and the performance degradation characteristics of components based on a third preset mapping relationship.

[0049] Furthermore, the extreme environment and component status identification module is used to collect physical characteristics of extreme environments and component operation data. Based on the third preset mapping relationship, it identifies the environment type and performance degradation characteristics, specifically implementing step S5 of the above control method. The workflow is as follows: First, two types of core data are collected. The physical characteristics of the extreme environment are collected through multiple sensors: temperature sensor, humidity sensor, dust sensor, and resistance sensor, all of which collect data at a sampling frequency of 50Hz. The robot component operation data are collected through joint sensors to collect motion resistance and smoothness, torso inertial sensors to collect posture stability, and drive unit sensors to collect operating current and heating temperature. After filtering and integration, a data matrix is ​​formed.

[0050] Subsequently, identification is completed based on the third preset mapping relationship. This mapping relationship is stored in the module's Flash memory in the form of a three-dimensional data table of environmental parameter ranges, component operating thresholds, and identification results, and supports online updates: for example, a temperature ≥45℃ is identified as a high-temperature environment, humidity ≥85%RH is identified as a high-humidity environment, and dust concentration ≥500μg / m³ is identified as a dusty environment. By comparing the component operating data with preset standards, such as a joint movement resistance exceeding the standard value by 20%, it is identified as a slight degradation, and the performance degradation characteristics and affected components are determined. The performance degradation characteristics include slight, moderate, and severe. The identification results are encapsulated in JSON format, including environmental type, scope of influence, and degradation level, and output to the self-reinforcing and compensation module through the Ethernet interface, while also storing the data for the past 7 days in a 64MB Flash backup.

[0051] Furthermore, the extreme environment and component status identification module has dynamic identification and fault tolerance functions: it updates data and identification results every 100ms, triggers emergency identification when environmental parameters change abruptly (a change in environmental parameters is defined as a single change of ≥30%), and completes the identification based on the data of other sensors when a certain environmental sensor fails. It also supports expanding the types of extreme environments through the Web configuration interface, including low temperature and high resistance ground, thereby improving the module's adaptability.

[0052] In summary, this embodiment specifically implements step S5 of the above control method. The extreme environment and component state recognition module establishes the correlation between the environment and component state through multi-dimensional data collection. The accurate mapping rules ensure the recognition accuracy. The dynamic recognition and fault tolerance functions enhance the adaptability to complex scenarios. The final output recognition result provides a key basis for subsequent gait self-reinforcement and compensation, ensuring the reliable operation of the robot in extreme environments.

[0053] Example 7, please refer to Figure 2 As shown: A walking gait control system for a quadruped robot. The gait self-reinforcement and compensation module is used to receive the recognition results, match the target scheme from the self-reinforcement gait scheme library according to the fourth preset mapping relationship, and superimpose compensation parameters according to the fifth preset mapping relationship to generate optimized gait control commands.

[0054] The gait self-reinforcement and compensation module is specifically used for: Based on the fourth preset mapping relationship, at least one self-reinforcing scheme is matched from the high-frequency small-amplitude oscillation phase, the high ground clearance support phase, and the low-energy heat dissipation gait. Adjustments are made based on the matched self-reinforcing scheme, and compensation parameters for joint force intensity, gait rhythm, and support ratio are determined based on the fifth preset mapping relationship. The compensation parameters are superimposed on the self-reinforcing gait command to generate an optimized gait control command.

[0055] Furthermore, the gait self-reinforcement and compensation module is used to receive the extreme environment and component status recognition results, match the self-reinforcement scheme and superimpose compensation parameters to generate optimized gait control commands. Specifically, step S6 of the above control method is implemented, and the workflow is as follows: First, the upstream recognition results are received. JSON format data from the extreme environment and component status recognition module is received through the Ethernet interface, including environment type, influence range and attenuation level. After verifying the integrity of the data, the core information is parsed. Fourth, the preset mapping relationship is stored in the module's Flash memory in the form of a three-dimensional data table of environment type, attenuation characteristics and self-reinforcement scheme, which supports online updates. For example, high temperature environment is matched with low energy consumption heat dissipation gait, high resistance ground is matched with high ground clearance support phase, and dusty and mild attenuation are matched with high frequency small amplitude oscillation phase.

[0056] Subsequently, compensation parameters were determined and optimized. The fifth preset mapping relationship was used to construct a data table based on the degree of attenuation, the type of compensation parameter, and the adjustment range, and stored in the same address Flash partition: for mild attenuation, the joint force intensity was adjusted by 1.1 to 1.3 times and the gait rhythm was accelerated by 10% to 15%; for moderate attenuation, the force intensity was adjusted by 1.3 to 1.5 times and the support phase ratio was increased to 60% to 70%; for severe attenuation, the force intensity was adjusted by 1.5 to 1.8 times and adapted to a slow rhythm and high stability gait. The swing phase / support phase ratio and joint force sequence were adjusted according to the matched self-reinforcing scheme. The compensation parameters were superimposed on the scheme instructions according to the weights, the optimization instructions were encapsulated according to the actuator protocol, and output to the execution unit through the CANFD bus. At the same time, the optimization log of nearly 12 hours was stored in 64MB Flash backup.

[0057] Furthermore, the gait self-reinforcement and compensation module has the functions of scheme combination and dynamic adaptation: when multiple extreme environments are superimposed, the combination scheme is matched according to the priority of environmental impact; it supports expanding the self-reinforcement scheme library through the Web configuration interface and adjusting the compensation parameter weights online; for example, in high temperature and high humidity environments, a low-energy heat dissipation gait combined with a high ground clearance support phase is adopted; when the recognition result is abnormal, the basic optimization scheme is activated to ensure that the command output is not interrupted, the optimization delay is ≤15ms, and it adapts to the real-time control requirements in extreme environments.

[0058] In summary, this embodiment specifically implements step S6 of the above control method. The gait self-reinforcement and compensation module establishes a precise correlation between the environment and component states and gait optimization through a dual preset mapping relationship. The scheme combination and dynamic adaptation functions enhance the compatibility of complex scenarios. The finally generated optimized gait control commands not only enhance the robot's adaptability to extreme environments but also compensate for the performance degradation of components, ensuring the reliability and long-term effectiveness of gait control.

[0059] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for gait control in a quadruped robot, characterized in that, The method includes the following steps: Step S1. Synchronously collect robot body status data and work tool status data, and align them according to timestamps to form a linkage data chain; Step S2. Collect the physical characteristics of the working environment and generate the allowable range constraints for tool operation based on the first preset mapping relationship; The first preset mapping relationship is a mapping relationship between the physical characteristics of the working environment and the allowable range constraints of tool operation; Step S3. Receive the linkage data link and the allowable range constraints of the tool operation, extract the constraint items, and call the gait parameter combination according to the second preset mapping relationship to generate the preliminary gait control command; The second preset mapping relationship is the mapping relationship between constraint terms and gait parameters; Step S4. Collect gait execution feedback data, compare it with the preset co-adaptation standard, adjust gait parameters, and output the calibrated gait control command; Step S5. Collect physical characteristics of extreme environments and operational data of robot components, and identify the type of extreme environment, the scope of influence, and the performance degradation characteristics of components based on the third preset mapping relationship; The third preset mapping relationship is a mapping relationship between the physical characteristics of extreme environments, the operating data of robot components, and the types of extreme environments and the performance degradation characteristics of components; Step S6. Receive the recognition result, match the target self-reinforcing gait scheme from the preset self-reinforcing gait scheme library according to the fourth preset mapping relationship, and superimpose compensation parameters according to the fifth preset mapping relationship to generate optimized gait control instructions; The fourth preset mapping relationship is the mapping relationship between extreme environment type, component performance degradation characteristics and self-reinforcing gait scheme; The fifth preset mapping relationship is the mapping relationship between the degree of component performance degradation and the compensation parameter.

2. The walking gait control method for a quadruped robot according to claim 1, characterized in that, Step S1 includes: The robot's body state data is generated by collecting the movement trajectory and rotation angle signals of the leg joints through joint sensors and the balance posture and center of gravity distribution offset trend of the torso through trunk inertial sensors. The tool's contact force signal is collected by the pressure sensor at the end of the tool, the tool's posture angle is collected by the joint angle sensor, and the tool's working mode switching signal is read to form the tool's status data. Align the robot's physical status data with the operational tool's status data by timestamp, and add sensor connection status and signal strength identifiers as data validity verification bits to form a linked data chain.

3. The walking gait control method for a quadruped robot according to claim 1, characterized in that, Step S2 includes: The system collects spatial information about the target location and obstacles using a visual sensor, relative distance information between the target and the machine body using a distance sensor, and operational accuracy information using a precision sensor. Based on the first preset mapping relationship, the target position and obstacle spatial information are transformed into body posture restrictions and spatial constraints, the relative distance information is transformed into stride upper limit, and the operation accuracy information is transformed into center of gravity movement boundary, which are integrated to form the allowable range constraint of the tool operation.

4. The walking gait control method for a quadruped robot according to claim 1, characterized in that, Steps S3 and S4 include: The constraint terms are extracted from the allowable range constraints of the tool's operation, including fuselage stability constraints, field of view continuity constraints, and spatial constraints; Based on the second preset mapping relationship, the gait parameter combination of low center of gravity and wide support phase and synchronous turning stride adaptation is called to adjust the support phase duration, center of gravity position, stride distribution and fuselage attitude, and output the initial gait control command. Collect fuselage posture fluctuation data and joint motion offset data as gait execution feedback data, and compare them with preset collaborative adaptation standards; Adjust the support phase stiffness and center of gravity offset parameters, and output the calibrated gait control command.

5. The walking gait control method for a quadruped robot according to claim 1, characterized in that, Steps S5 and S6 include: Ambient temperature, air humidity, dust concentration, and ground movement resistance are collected by temperature sensors, humidity sensors, dust sensors, and resistance sensors, respectively, as physical characteristics of extreme environments; The robot collects joint motion resistance and motion smoothness data through joint sensors, posture stability data through torso inertial sensors, and operating current and heat generation temperature through drive unit sensors, and integrates them to form robot component operation data. Based on the third preset mapping relationship, the extreme environment type and impact range are determined, and the robot component operation data are compared with the preset component operation standards to determine the component performance degradation characteristics. Based on the fourth preset mapping relationship, at least one self-reinforcing gait scheme is matched from the preset self-reinforcing gait scheme library, including high-frequency small-amplitude swing phase gait, high ground clearance support phase gait, and low-energy heat dissipation gait. Based on the matched self-reinforcing gait scheme, generate self-reinforcing gait commands, adjust the ratio of swing phase to support phase and the joint force sequence, and determine the compensation parameters of joint force intensity, gait rhythm and support phase ratio based on the fifth preset mapping relationship. The compensation parameters are superimposed on the self-reinforcing gait command to generate an optimized gait control command.

6. A gait control system for a quadruped robot, applicable to the gait control method for a quadruped robot as described in any one of claims 1-5, characterized in that, The system includes a linkage data acquisition module, a work environment constraint conversion module, a gait coordination control command generation module, a coordination deviation calibration module, an extreme environment and component status recognition module, and a gait self-reinforcement and compensation module. The linkage data acquisition module is used to synchronously collect robot body status data and working tool status data, and align them according to timestamps to form a linkage data chain; The work environment constraint conversion module is used to collect the physical characteristics of the work environment and generate the allowable range constraints for tool operation based on the first preset mapping relationship; The gait coordination control command generation module is used to receive the linkage data link and the allowable range constraints of the tool operation, extract the constraint items, and generate preliminary gait control commands by calling the gait parameter combination according to the second preset mapping relationship. The coordination deviation calibration module is used to collect gait execution feedback data, compare it with the preset coordination adaptation standard, adjust gait parameters, and output calibrated gait control commands. The extreme environment and component status identification module is used to collect physical characteristics of extreme environments and robot component operation data, and identify the type of extreme environment, the scope of influence, and component performance degradation characteristics based on a third preset mapping relationship; The gait self-reinforcement and compensation module is used to receive the recognition results, match the target self-reinforcement gait scheme from the preset self-reinforcement gait scheme library according to the fourth preset mapping relationship, and superimpose compensation parameters according to the fifth preset mapping relationship to generate optimized gait control commands.

7. A walking gait control system for a quadruped robot according to claim 6, characterized in that, The linked data acquisition module is specifically used for: It receives signals transmitted from joint sensors and torso inertial sensors to form robot body state data; It receives signals from the tool end pressure sensor and the tool joint angle sensor, as well as the tool working mode switching signal, and forms tool status data. Add timestamps and validity check bits to the robot body status data and the working tool status data, align the two types of data according to the timestamps to form a linked data chain and output it.

8. A walking gait control system for a quadruped robot according to claim 6, characterized in that, The working environment constraint conversion module is specifically used for: It receives signals transmitted from vision sensors, distance sensors, and precision sensors; Based on the first preset mapping relationship, the target position and obstacle spatial information collected by the vision sensor are converted into body posture restrictions and spatial constraints, the relative distance information between the target and the body collected by the distance sensor is converted into the upper limit of stride, and the accuracy signal is converted into the center of gravity movement boundary, which are integrated into the allowable range constraint of the tool operation.

9. A walking gait control system for a quadruped robot according to claim 6, characterized in that, The gait coordination control command generation module is specifically used for: Receive the linkage data link and verify the validity of the data; Extract fuselage stability constraints, field of view continuity constraints, and spatial constraints from the tool's permissible operating range constraints; Based on the second preset mapping relationship, the corresponding gait parameter combination is invoked to adjust the support phase duration, center of gravity position, stride distribution and fuselage attitude, and output preliminary gait control commands.

10. A walking gait control system for a quadruped robot according to claim 6, characterized in that, The gait self-reinforcement and compensation module is specifically used for: Based on the fourth preset mapping relationship, at least one self-reinforcing gait scheme is matched from the preset self-reinforcing gait scheme library, including high-frequency small-amplitude swing phase gait, high ground clearance support phase gait, and low-energy heat dissipation gait. A self-reinforcing gait command is generated based on the matched self-reinforcing gait scheme, the ratio of the swing phase to the support phase and the sequence of joint force application are adjusted, and the compensation parameters for joint force intensity, gait rhythm and support phase ratio are determined based on the fifth preset mapping relationship. The compensation parameters are superimposed on the self-reinforcing gait command to generate an optimized gait control command.

Images