Bionic ant colony robot additive construction process and system based on calibrated dice space constraint

Abstract

The invention discloses a bionic ant colony robot additive construction technology and system based on calibration dice space constraint, and relates to the technical field of intelligent construction, the bionic ant colony robot additive construction technology comprises the following calibration dice additive construction core steps: step 1, system initialization and global positioning, step 2, calibration dice setting and number recording, step 3, visual identification and local fine calibration, and step 4, calibration dice additive construction. According to the method, a calibration dice component, SLAM + UWB + visual identification multi-source fusion positioning technology is introduced into a bionic ant colony building robot system, and a closed-loop control mechanism of'simultaneous printing, identification and calibration 'is combined, so that the system realizes high-precision, autonomous and continuous in-situ building in a complex and unstructured environment; compared with existing building printing equipment, the building printing equipment has the remarkable advantages in the aspects of positioning precision, building stability, environmental adaptability, multi-machine cooperation efficiency, structural integration and the like.

Description

Technical Field

[0001] This invention relates to the field of intelligent construction technology, specifically to a biomimetic ant colony robot additive construction process and system based on calibration dice space constraints. Background Technology

[0002] With the deep integration of construction robots and 3D printing technology, biomimetic multi-robot collaborative construction has gradually become an important research direction in the field of intelligent construction. At present, construction printing equipment is a fixed track, gantry or single-machine mobile structure. Construction printing robots generally adopt SLAM or UWB-based positioning solutions. In complex construction, multiple robots need to work together to complete the task, which requires high global control of the robots and strict local positioning. Existing construction robots have high requirements for the working environment and low stability when used in complex environments. Traditional construction printing robots have a decrease in printing accuracy over time and require global calibration before construction. They lack self-correction and dynamic feedback capabilities, making it difficult to maintain continuous accuracy. When building in a group, the local positioning errors of each robot cannot be unified, which can easily lead to path conflicts, construction misalignment and collaborative instability. In light of the aforementioned technical problems, it is necessary to propose a biomimetic ant colony robot additive manufacturing process and system based on calibration dice space constraints that can continuously provide visual landmarks, achieve real-time coordinate correction, and be integrated with structural materials during the printing and construction process, in order to overcome the deficiencies in the prior art. Summary of the Invention

[0003] This invention provides a biomimetic ant colony robot additive manufacturing process and system based on calibration dice space constraints. It can effectively solve the problems mentioned in the background art, such as the high requirements of the working environment of existing construction robots, the low stability in complex environments, the decrease in printing accuracy of traditional construction printing robots over time, the need for global calibration before construction, the lack of self-correction and dynamic feedback capabilities, the difficulty in maintaining continuous accuracy, and the inability to unify the local positioning errors of each robot during collaborative construction, which can easily lead to path conflicts, construction misalignment and collaborative instability.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a biomimetic ant colony robot additive manufacturing process based on calibration dice spatial constraints, comprising the following core steps of calibration dice additive manufacturing: Step 1: System initialization and global location; Step 2: Setting up and recording the calibration dice; Step 3: Visual recognition and local fine-tuning; Step 4: Construction Path Planning and Execution; Step 5: Dynamic calibration and error correction; Step 6: Regional progression and structural integration; The detailed steps for step 3 are as follows: Before construction, the robot scans images of the work area, and the AprilTag recognition module extracts corner features and calculates the spatial pose of each calibration die. According to theoretical position With observation location deviation : , If the deviation exceeds the threshold =2mm, If the value is 0.5, then perform attitude correction and update the robot coordinates. : in Here are the translation and rotation gain matrices; This step eliminates SLAM drift and establishes a locally fine-calibrated coordinate system.

[0005] According to the above technical solution, after the robot is started in step 1, the SLAM module of the bionic construction robot constructs an environmental point cloud map in real time using LiDAR and IMU data to obtain the relative pose. Simultaneously, the UWB module and the UWB anchor points set around the site perform distance measurement and registration to obtain absolute positioning coordinates. The integrated positioning module dynamically calculates the confidence coefficient based on environmental conditions. Represents SLAM weights and confidence coefficients. The UWB weights are represented, and a weighted fusion algorithm is used to calculate the global initial pose. : in, ,and , Based on the adaptive adjustment of signal quality, the system establishes a unified world coordinate system, which serves as the benchmark for subsequent calibration die placement and printing path planning.

[0006] According to the above technical solution, in step 2, the robot receives the task area information assigned by the group scheduling platform, uses the deployment arm to hold a regular hexahedral calibration die, identifies its AprilTag code through the end-effector camera to confirm the direction and ID number, and then accurately places the calibration die at the target position, with the position coordinates marked as follows: =( , , ).

[0007] According to the above technical solution, in step 4, the group scheduling platform slices the input 3D building model into layers, generates G-code printing path data, the robot's central processing unit parses the path instructions, and uses inverse kinematics to solve the nozzle tip angle. : in, The target point of the path; Extrusion speed of the nozzle Determined by material flow rate Q, spray width w, and layer height h: in, As an empirical proportional coefficient, the high-degree-of-freedom nozzle at the end of the actuator arm adjusts its attitude according to the real-time trajectory.

[0008] According to the above technical solution, in step 5, during the printing process, the visual recognition module detects the visible quantity and recognition confidence S of AprilTag in real time. At a certain moment, the visible quantity... At that time, the system automatically reduces the movement speed and triggers a scan for correction; If local path drift is detected, the path is corrected based on the feedback. : Simultaneously, the vision module detects the occlusion rate r of the AprilTag pattern, and when it exceeds a threshold... When the value is 0.9, the printing coverage of that surface is considered complete.

[0009] According to the above technical solution, in step 6, when the AprilTag face of all the calibration dice in the area is covered and the system detects that the duration of the corresponding signal disappearance τ>1.0s, it is determined that the printing of the area is completed. The scheduling platform instructs the robot to move to the next area to perform the construction task. After the construction is completed, the calibration dice and the structural material naturally merge to form an internal skeleton, which serves as both a positioning residual unit and a force support unit.

[0010] According to the above technical solution, the step of additive manufacturing of calibration dice is realized by a construction system. The construction system includes the bionic construction robot mentioned in step 1, calibration dice components, fusion positioning module, group scheduling platform, and supply and detection module. The various modules of the construction system form a closed-loop construction system through data link, motion control link and material transmission link. The biomimetic construction robot is the main body of the construction. It adopts a six-legged biomimetic motor structure. Each mechanical leg consists of four degrees of freedom: hip, knee, ankle, and claw. The robot's body is equipped with a lidar, vision camera, IMU, UWB module, central processing unit, and communication module to realize environmental perception, path planning, pose calculation, and real-time motion control.

[0011] According to the above technical solution, the calibration die component is a calibration element with a regular hexahedral structure and a side length of 30–50 mm. Each face is printed with an AprilTag visual code for visual recognition and pose calculation. The outer surface of the die is coated with a matte reflective coating, and a UWB passive reflection module is embedded inside to provide auxiliary positioning when the light degrades. A coarsening coupling layer is provided on the outer layer. The calibration die is automatically placed and identified by a robot during the construction process. After being covered by materials, it is integrated with the structure to form a stable internal positioning unit.

[0012] According to the above technical solution, the fusion positioning module consists of a SLAM module, a UWB positioning module, and a visual recognition module. The three modules provide relative pose, absolute coordinates, and local fine-tuning data, respectively. The positioning module uses extended Kalman filtering and factor graph optimization algorithms for dynamic weighted fusion to output high-precision real-time pose.

[0013] According to the above technical solution, the group scheduling platform is a multi-robot control and information synchronization center, responsible for task division, path allocation, obstacle avoidance scheduling and data fusion. It includes a communication base station, a task allocation module, a path optimization module and a payload computer. Each robot periodically uploads status data through a wireless network. The group scheduling platform performs global scheduling based on the ant colony optimization algorithm. The supply and testing module is located at the edge of the construction site and includes a material supply station, a charging station, and a testing station.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Regarding positioning and construction accuracy, this invention achieves real-time pose calculation at the centimeter to millimeter level by dynamically weighted fusion of the overall positioning module (SLAM, UWB, and visual recognition). Traditional building printing robots rely solely on LiDAR SLAM or inertial navigation, which can easily lead to cumulative drift and misalignment between printing layers when there is insufficient lighting, dust, fog, or reflection interference. This system establishes a continuously visible visual landmark network by arranging hexahedral calibration dice components with AprilTag encoding in the work area. Combined with the absolute reference provided by UWB anchor points, it achieves dual calibration of global and local values. This technology eliminates the cumulative error of long-term SLAM drift and can automatically correct trajectory deviations during printing, ensuring that the component layer height and stacking path are consistent with the theoretical model, thus significantly improving printing accuracy.

[0015] 2. This invention employs a six-legged motorized structure and adjustable suction claw design, similar to a biomimetic construction robot, enabling the robot to move and stand stably on complex terrains such as slopes, gravel, and mud. Through a multi-sensor perception mechanism using lidar and visual cameras, the system can continuously maintain its positioning function in complex environments such as lighting, dust, and reflection interference. When the visual recognition signal attenuates, the fusion positioning module automatically increases the UWB weight ratio to maintain continuous positioning output. This mechanism originates from the adaptive weight scheduling algorithm of the fusion positioning module and the omnidirectional visual feature design of the hexahedral calibration die component, enabling the system to work reliably in scenarios without GPS or with low visibility, exhibiting strong robustness and autonomy.

[0016] 3. This invention achieves online error monitoring and self-calibration during the printing process through a "dynamic calibration and error correction" mechanism. The visual recognition module detects the recognition confidence and visible quantity of AprilTag in real time. When the signal decreases or the path drifts, the payload computer automatically decelerates and recalibrates its attitude, thereby realizing closed-loop control of "printing-recognition-correction". Unlike the traditional method of "one-time calibration before start-up and fixed coordinates throughout", this method can dynamically compensate for errors during printing, prevent the accumulation of inter-layer errors, and ensure the dimensional stability of long-term continuous printing. This advantage comes directly from the closed-loop control logic design of the payload computer and the fusion positioning module.

[0017] 4. This invention achieves distributed construction by multiple robots through a group scheduling platform. The task allocation module and trajectory optimization module within the platform adopt the ant colony optimization algorithm to schedule multiple biomimetic construction robots to work synchronously in different areas of the same building model in real time. The coordinates are kept consistent through UWB and visual recognition data. The task division and path conflict resolution mechanism of the group scheduling platform of this system can achieve parallel tasks under multi-machine cooperation, avoid path conflicts and repetitive operations, and improve the overall construction efficiency.

[0018] 5. The calibration die component of this invention is covered with a roughened coupling layer, which forms a high-strength chemical / mechanical bond with the main building body after the printing material is deposited, becoming part of the structure. Traditional calibration objects are often external auxiliary parts that need to be removed after printing, which can easily cause holes or structural weakening. The calibration die of this invention does not need to be removed after printing and can be permanently retained inside the building. It serves as a structural skeleton to strengthen local strength and as a positioning element for later expansion or quality inspection, realizing the integration of calibration and construction. This advantage comes from the material design and surface treatment structure of the calibration die component.

[0019] In summary, this invention introduces calibration dice components and multi-source fusion positioning technology combining SLAM, UWB, and visual recognition into a biomimetic ant colony construction robot system. Combined with a closed-loop control mechanism of "printing, recognizing, and calibrating simultaneously," the system achieves high-precision, autonomous, and continuous in-situ construction in complex, unstructured environments. Compared to existing construction printing equipment, this invention offers significant advantages in positioning accuracy, construction stability, environmental adaptability, multi-machine collaborative efficiency, and structural integration. Therefore, this invention not only surpasses existing technologies in accuracy, efficiency, and robustness but also proposes innovative solutions for structural integration and sustainable positioning. It enables highly reliable in-situ construction in scenarios such as disaster relief, underground construction, and rapid construction in extreme environments, demonstrating significant engineering application and promotion value. Attached Figure Description

[0020] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0021] In the attached diagram: Figure 1 This is a schematic diagram of the construction system structure of the present invention; Figure 2 This is a diagram illustrating the core steps of additive manufacturing with calibration dice in Embodiment 1 of the present invention; Figure 3 This is a flowchart of the additive manufacturing process for the calibration die of this invention; Figure 4 This is a front view of the biomimetic construction robot of the present invention; Figure 5 This is a side view of the biomimetic construction robot of the present invention; Figure 6 This is a schematic diagram of the working of the bionic construction robot of the present invention; Figure 7 This is a schematic diagram of the construction components of the biomimetic construction robot of the present invention; Figure 8 This is a schematic diagram of the calibration die structure of the present invention; The following components are labeled in the diagram: 1. Walking system; 2. Sensing system; 3. Control system; 4. 3D printing material storage compartment; 5. Calibration die placement compartment; 6. 3D printing nozzle; 7. Clawing and calibrating die mechanism; 8. Vision camera; 9. Two-position five-way solenoid valve; 10. LiDAR; 11. Programmable computer controller; 12. Smart battery pack; 13. Flexible suction cup; 14. Double-acting cylinder; 15. Air pump; 16. Charging plug. Detailed Implementation

[0022] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0023] Example 1: like Figure 1-3 As shown, this invention provides a technical solution: a biomimetic ant colony robot additive manufacturing process based on calibration dice spatial constraints, comprising the following core steps of calibration dice additive manufacturing: Step 1: System initialization and global location; After the robot is activated, its SLAM module uses LiDAR and IMU data to build a real-time point cloud map of the environment and obtain relative pose. Simultaneously, the UWB module and the UWB anchor points set around the site perform distance measurement and registration to obtain absolute positioning coordinates. The integrated positioning module dynamically calculates the confidence coefficient based on environmental conditions. Represents SLAM weights and confidence coefficients. The UWB weights are represented, and a weighted fusion algorithm is used to calculate the global initial pose. : in, ,and , Based on the adaptive adjustment of signal quality, the system establishes a unified world coordinate system as the benchmark for subsequent calibration die placement and printing path planning; Step 2: Setting up and recording the calibration dice; The robot receives task area information assigned by the group scheduling platform, uses its deployment arm to hold a hexahedral calibration die, and identifies its AprilTag code through its end-effector camera to confirm the direction and ID number. It then precisely places the calibration die at the target location, with the coordinates marked as follows: =( , , ); The placement strategy adopts the principle of multi-view visibility optimization, while ensuring the number of calibration dice visible at any printing position. This allows the vision system to maintain stable calibration even when there is partial occlusion or pose change. Step 3: Visual recognition and local fine-tuning; Before construction, the robot scans images of the work area, and the AprilTag recognition module extracts corner features and calculates the spatial pose of each calibration die. According to theoretical position With observation location deviation : , If the deviation exceeds the threshold =2mm, If the value is 0.5, then perform attitude correction and update the robot coordinates. : in Here are the translation and rotation gain matrices; This step eliminates SLAM drift and establishes a local fine-calibration coordinate system, ensuring that the printed trajectory is strictly aligned with the calibration datum. Step 4: Construction Path Planning and Execution; The group scheduling platform slices the input 3D building model into layers, generates G-code printing path data, and the robot's central processing unit parses the path instructions and uses inverse kinematics to solve for the nozzle tip angle. : in, The target point of the path; Extrusion speed of the nozzle Determined by material flow rate Q, spray width w, and layer height h: in, Using an empirical proportionality coefficient, the high-degree-of-freedom nozzle at the end of the actuator arm adjusts its attitude according to the real-time trajectory to achieve uniform deposition on complex surfaces; Step 5: Dynamic calibration and error correction; During the printing process, the visual recognition module continuously monitors the number of visible AprilTags and their recognition confidence level S. At any given moment, the number of visible AprilTags... At that time, the system automatically reduces the movement speed and triggers a scan for correction; If local path drift is detected, the path is corrected based on the feedback. : Simultaneously, the vision module detects the occlusion rate r of the AprilTag pattern, and when it exceeds a threshold... When the value is 0.9, the printing coverage of that surface is considered complete. The above process forms a continuous "print-recognition-correction" closed loop, enabling the robot to maintain trajectory accuracy under environmental disturbances and positioning drift. Step 6: Regional progression and structural integration; When the AprilTag faces of all the calibration dice in the area are covered and the system detects that the duration of the corresponding signal disappearance τ > 1.0s, it is determined that the printing of the area is completed, and the scheduling platform instructs the robot to move to the next area to perform the construction task; After construction, the calibration die naturally integrates with the structural materials to form an internal skeleton, serving as both a positioning element and a load-bearing support unit, thereby enhancing the overall structural strength and traceability for future expansions.

[0024] To address the problems of insufficient positioning accuracy, single calibration method, inability to integrate structures, and unstable group collaboration in existing building printing robots in complex or unstructured environments, a calibration die construction system for biomimetic ant colony printing construction robots is proposed. The system uses a hexahedral visual calibration die designed to be integrated with the printed structure. Combined with fusion algorithms of SLAM, UWB and visual recognition, it enables construction robots to achieve autonomous positioning, dynamic calibration and closed-loop construction control in environments without external positioning signals. The system can simultaneously support high-precision printing by a single robot and collaborative operation by multiple robots, achieving full-process guidance and control of "building, recognizing and calibrating at the same time".

[0025] like Figure 1 and Figure 2 As shown, the steps of additive manufacturing of calibration dice are realized through a construction system. The construction system includes the bionic construction robot of step 1, calibration dice components, fusion positioning module, group scheduling platform, and supply and detection module. The various modules of the construction system form a closed-loop construction system through data link, motion control link and material transmission link. The biomimetic construction robot is the main body of the construction. It adopts a six-legged biomimetic motor structure. Each mechanical leg consists of four degrees of freedom: hip, knee, ankle, and claw. The adjustable suction claws at the ends of the mechanical legs provide stable support on uneven ground such as slopes and gravel. The body of the biomimetic construction robot is equipped with LiDAR, vision camera, IMU, UWB module, central processing unit and communication module to realize environmental perception, path planning, pose calculation and real-time motion control. The calibration die assembly is a calibration component with a regular hexahedral structure and a side length of 30–50 mm. Each face is printed with an AprilTag visual code for visual recognition and pose calculation. The outer surface of the die is coated with a matte reflective coating to enhance recognition robustness. An embedded UWB passive reflection module is used to provide auxiliary positioning when the light degrades. The outer layer has a coarsened coupling layer to ensure a high-strength bond with the printing material. The calibration die is automatically placed and identified by a robot during the construction process. After being covered by the material, it merges with the structure to form a stable internal positioning unit. The fusion localization module consists of a SLAM module, a UWB localization module, and a visual recognition module. These three modules provide relative pose, absolute coordinates, and local calibration data, respectively. The localization module uses Extended Kalman Filter (EKF) and Graph Optimization (Graph Optimization) algorithms for dynamic weighted fusion, outputting high-precision real-time pose. The core components of the SLAM-UWB-vision three-source fusion and dynamic calibration closed loop are as follows: Figure 3 As shown; The swarm scheduling platform serves as the control and information synchronization center for multiple robots. It is responsible for task division, path allocation, obstacle avoidance scheduling, and data fusion. It includes a communication base station, a task allocation module, a path optimization module, and a payload computer. Each robot periodically uploads status data through a wireless network. The swarm scheduling platform performs global scheduling based on the ant colony optimization algorithm. The supply and testing modules are located at the edge of the construction site and include a material supply station, a charging station, and a testing station.

[0026] Example 2: The construction system is a distributed building printing device with a physical structure. It consists of a biomimetic construction robot, a calibration die component, a fusion positioning module, a group scheduling platform, a supply and testing module, and a formed building area. The overall structure of the biomimetic construction robot is as follows Figure 4 and Figure 5 As shown, the device consists of a body platform, a six-legged bionic walking mechanism, dual actuators, a fusion positioning module, a sensing unit, an energy module, and a high-degree-of-freedom nozzle. The six-legged bionic walking mechanism provides a stable posture platform to support the high-precision printing actuators. The high-degree-of-freedom nozzle is mounted at the front end of the printing actuator in the dual actuators, and its spraying unit can achieve continuous real-time adjustment of the material deposition direction and attitude. The numbering correspondence of each component is as follows: Figure 7 As shown, The walking system 1 is equipped with two sets, each set having three flexible suction cups 13. The two sets of flexible suction cups 13 are connected to the sensing system 2 respectively. The sensing system 2 is connected to the 3D printing material storage chamber 4 and the calibration die placement chamber 5 at both ends. The sensing system 2 is equipped with a lidar 10 and two vision cameras 8 for data signal acquisition. The lidar 10 also contains a 3D printing nozzle 6 and a die-grabbing and calibration mechanism 7. The lidar 10 packages and transmits the data to the control system 3. The control system 3 contains two sets of double-acting cylinders 14, a two-position five-way solenoid valve 9, and an air pump 15. The control system 3 also contains a programmable computer controller 11 and a smart battery pack 12, which is connected to a charging plug 16 for power delivery. The control system 3 is used to adjust the posture of the overall hexapod bionic walking mechanism, while the control system 3 also controls the start and stop of the walking system 1.

[0027] The robot platform is an integral load-bearing frame made of lightweight, high-strength materials, used to install control circuits, energy modules, and communication equipment. The bottom of the robot is equipped with a six-legged bionic walking mechanism, with six bionic mechanical legs distributed symmetrically. Each mechanical leg consists of a hip joint, a knee joint, and an ankle joint, all of which are controlled by a servo drive and torque sensing system. The joints are connected by rigid linkages and ball joints, enabling multi-degree-of-freedom movement. The feet are equipped with anti-slip and wear-resistant pads to ensure the robot maintains a stable posture on uneven terrain such as gravel and concrete. The six-legged mechanism uses a three-legged support gait control algorithm to keep all three legs grounded at all times, thus forming a stable support surface. The load computer adjusts the gait rhythm and support leg distribution in real time based on the pose feedback from the fusion positioning module to reduce robot vibration and ensure the trajectory accuracy of the printing execution arm in complex terrain. The sensing unit is mounted on the top of the machine body and at the front end of the actuator arm. It includes a lidar, a vision camera, an inertial measurement unit (IMU), and an ultra-wideband (UWB) module. The lidar is used to scan the environment to generate a point cloud map. The IMU is used to measure acceleration and angular velocity information. The UWB module performs wireless ranging with anchor points deployed around the construction site to obtain absolute coordinates. The vision camera is used to identify AprilTag tags on the calibration die surface to extract relative pose information. All sensing data is transmitted to the fusion positioning module for processing. The fusion positioning module is installed inside the aircraft and contains a SLAM submodule, a UWB positioning submodule, and a visual recognition submodule. This module uses a confidence-based weighted fusion algorithm to fuse the relative positioning of the LiDAR and IMU, the absolute positioning of the UWB, and the local fine-tuning results of the visual recognition, and outputs a high-precision pose with six degrees of freedom. This pose data is transmitted to the onboard computer via a bus for navigation, gait planning, and printing control. The fusion positioning module can dynamically adjust the weights of each submodule in environments with signal obstruction, changes in lighting, and dust and fog interference, thereby maintaining the robustness and accuracy of the overall positioning. The payload computer is mounted on the top of the machine platform and is the core control unit of the whole machine. It integrates a central processing unit, motion control module, task scheduling module and communication module. The central processing unit is responsible for path planning, attitude analysis, and generation of printing task instructions. The motion control module executes real-time motion control of the hexapod walking mechanism, the delivery arm, and the printing arm. The task scheduling module allocates paths and performs local obstacle avoidance based on the building segment tasks issued by the group scheduling platform. The communication module maintains data interaction with the group scheduling platform through a wireless network. The payload computer is connected to the fusion positioning module, servo driver, and sensor network through a high-speed bus to form a closed loop of task analysis, motion control, and feedback. The biomimetic construction robot is equipped with two execution devices at the front: a delivery execution arm and a printing execution arm. The delivery execution arm is a six-degree-of-freedom robotic arm with a gripping mechanism and a recognition camera at its end. The gripping mechanism picks up a regular hexahedral calibration die from the storage bin and accurately places it in the predetermined position. The gripping mechanism has a built-in torque sensor to ensure the stability of the placement process and prevent surface damage. The camera is responsible for confirming the AprilTag number and orientation, realizing orientation correction and number recording. The printing execution arm is a multi-joint robotic arm with a high-degree-of-freedom nozzle connected to its end. The high-degree-of-freedom nozzle consists of multiple independently controlled spraying units, each connected to an independent material channel, which can realize multi-material and multi-angle printing. The nozzle posture is controlled in real time by the load computer to ensure the printing accuracy and uniformity of different angles and curved surfaces. During the printing process, the nozzle can achieve ±90° pitch, 360° rotation and posture reset in space, thereby meeting the printing needs of complex architectural forms.

[0028] The calibration die assembly is a key structure in the system, such as... Figure 8 As shown, each calibration die has a regular hexahedral structure with a side length of 30-50mm. Each of the six faces is printed with an AprilTag code. The outer shell of the die is made of heat-resistant polymer or composite material, and the surface is treated with a matte anti-reflective finish to enhance recognition stability. Before printing begins, the calibration die is placed at a designated coordinate position by the delivery arm. During the printing process, it provides local visual reference information for robot posture calibration. After printing, the calibration die is covered by building materials and integrated with the structure to form an internal positioning skeleton and support nodes.

[0029] The group scheduling platform is a host computer control system that connects to multiple biomimetic construction robots via a wireless communication network. The platform consists of a task allocation module, a path planning module, and a status monitoring module. The task allocation module automatically partitions the 3D building model and assigns it to different robots. The path planning module generates a global trajectory and performs collision avoidance optimization. The status monitoring module receives the pose, energy consumption, and construction progress of each robot in real time, coordinates the group's work sequence, and achieves synchronous construction and dynamic scheduling.

[0030] The supply and testing module is located at the edge of the construction site and includes a material supply station, a charging station, and a testing station. The material supply station delivers building materials to the print head of the printing arm through a flexible material conveying pipe. The charging station replenishes energy through wireless charging and quick battery replacement. The testing station is equipped with laser scanning and visual inspection devices to detect the dimensional deviations and surface quality of the formed building and feeds the test results back to the group scheduling platform for accuracy correction.

[0031] The sensing unit is connected to the fusion positioning module via a data bus. The fusion positioning module communicates bidirectionally with the payload computer. The payload computer controls the hexapod walking mechanism, the delivery execution arm, and the printing execution arm via a servo bus. The clamping mechanism at the end of the delivery execution arm is mechanically connected to the calibration die assembly. The high-degree-of-freedom nozzle of the printing execution arm is connected to the feeding station via a flexible material conveying pipe. The group scheduling platform interacts with the payload computer via a wireless communication module. The supply and detection module shares task and detection information with the scheduling platform.

[0032] The hexapod bionic walking mechanism provides high stability and terrain adaptability, enabling the robot to walk autonomously in trackless environments while maintaining stable body posture. The integrated positioning module achieves multi-source fusion positioning of SLAM, UWB, and vision, improving construction accuracy and environmental robustness. The deployment of the execution arm completes the placement and numbering of calibration dice, providing high-confidence landmarks for the vision system. The printing execution arm and high-degree-of-freedom nozzle complete the precise deposition of construction materials, enabling the printing of complex geometric structures. The group scheduling platform is responsible for the task division and coordination of multiple robots. The supply and detection modules realize continuous operation and closed-loop quality detection. The entire system forms a closed-loop process of "autonomous positioning - calibration placement - path planning - printing construction - accuracy detection". It has a compact structure, coordinated functions, high precision, high stability, and strong environmental adaptability, and is suitable for post-disaster reconstruction, underground space construction, and unmanned building scenarios.

[0033] like Figure 6 As shown, the system is suitable for in-situ construction scenarios such as large curvature surfaces, high-altitude exterior walls, and enclosed spaces. Multiple robots can simultaneously perform tasks such as placing calibration dice, path calibration, and material deposition on complex curved surfaces, demonstrating a mobility and spatial adaptability that traditional gantry printers cannot achieve.

[0034] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A biomimetic ant colony robot additive manufacturing process based on calibration dice space constraints, characterized in that: The core steps of additive manufacturing with calibration dice include the following: Step 1: System initialization and global location; Step 2: Setting up and recording the calibration dice; Step 3: Visual recognition and local fine-tuning; Step 4: Construction Path Planning and Execution; Step 5: Dynamic calibration and error correction; Step 6: Regional progression and structural integration; The detailed steps for step 3 are as follows: Before construction, the robot scans images of the work area. The AprilTag recognition module extracts corner features and calculates the spatial pose of each calibration die. According to theoretical position and observation location deviation : ， If the deviation exceeds the threshold =2mm, If the value is 0.5, then perform attitude correction and update the robot coordinates. : in Here are the translation and rotation gain matrices; This step eliminates SLAM drift and establishes a locally fine-calibrated coordinate system.

2. The additive manufacturing process for a biomimetic ant colony robot based on calibration dice spatial constraints according to claim 1, characterized in that: After the robot is started in step 1, the SLAM module of the biomimetic construction robot constructs an environmental point cloud map in real time using LiDAR and IMU data to obtain the relative pose. Simultaneously, the UWB module and the UWB anchor points set around the site perform distance measurement and registration to obtain absolute positioning coordinates. The integrated positioning module dynamically calculates the confidence coefficient based on environmental conditions. Represents SLAM weights and confidence coefficients. The UWB weights are represented, and a weighted fusion algorithm is used to calculate the global initial pose. : in, ,and , Based on the adaptive adjustment of signal quality, the system establishes a unified world coordinate system, which serves as the benchmark for subsequent calibration die placement and printing path planning.

3. The additive manufacturing process for a biomimetic ant colony robot based on calibration dice space constraints according to claim 1, characterized in that: In step 2, the robot receives the task area information assigned by the group scheduling platform, uses its deployment arm to hold a hexahedral calibration die, identifies its AprilTag code through its end-effector camera to confirm the direction and ID number, and then precisely places the calibration die at the target location, with the position coordinates marked as follows: =( , , ).

4. The additive manufacturing process for a biomimetic ant colony robot based on calibration dice spatial constraints according to claim 1, characterized in that: In step 4, the group scheduling platform slices the input 3D building model into layers, generates G-code printing path data, and the robot's central processing unit parses the path instructions and uses inverse kinematics to solve for the nozzle tip angle. : in, The target point of the path; Extrusion speed of the nozzle Determined by material flow rate Q, spray width w, and layer height h: in, As an empirical proportional coefficient, the high-degree-of-freedom nozzle at the end of the actuator arm adjusts its attitude according to the real-time trajectory.

5. The additive manufacturing process for a biomimetic ant colony robot based on calibration dice spatial constraints according to claim 1, characterized in that: In step 5, during the printing process, the visual recognition module continuously monitors the visible quantity and recognition confidence level S of AprilTags. At a certain moment, the visible quantity... At that time, the system automatically reduces the movement speed and triggers a scan for correction; If local path drift is detected, the path is corrected based on feedback. : Simultaneously, the vision module detects the occlusion rate r of the AprilTag pattern, and when it exceeds a threshold... When the value is 0.9, the printing coverage of that surface is considered complete.

6. The additive manufacturing process for a biomimetic ant colony robot based on calibration dice space constraints according to claim 1, characterized in that: In step 6, when the AprilTag faces of all calibration dice in the area are covered and the system detects that the duration of the corresponding signal disappearance τ>1.0s, it is determined that the printing of the area is completed. The scheduling platform instructs the robot to move to the next area to perform the construction task. After the construction is completed, the calibration dice and the structural material naturally merge to form an internal skeleton, which serves as both a positioning residual unit and a force support unit.

7. A biomimetic ant colony robot additive manufacturing system based on calibration dice space constraints, the system of the biomimetic ant colony robot additive manufacturing process based on calibration dice space constraints according to any one of claims 1-6, characterized in that: The steps of additive manufacturing with calibration dice are implemented through a construction system, which includes the bionic construction robot described in step 1, calibration dice components, fusion positioning module, group scheduling platform, and supply and detection module. The various modules of the construction system form a closed-loop construction system through data links, motion control links, and material transmission links. The biomimetic construction robot is the main body of the construction. It adopts a six-legged biomimetic motor structure. Each mechanical leg consists of four degrees of freedom: hip, knee, ankle, and claw. The body of the biomimetic construction robot is equipped with a lidar, vision camera, IMU, UWB module, central processing unit and communication module to realize environmental perception, path planning, pose calculation and real-time motion control.

8. The biomimetic ant colony robot additive manufacturing system based on calibration dice space constraints according to claim 7, characterized in that: The calibration die assembly is a calibration component with a regular hexahedral structure and a side length of 30–50 mm. Each face is printed with an AprilTag visual code for visual recognition and pose calculation. The outer surface of the die is coated with a matte reflective coating, and a UWB passive reflection module is embedded inside to provide auxiliary positioning when the light degrades. A coarsening coupling layer is provided on the outer layer. The calibration die is automatically placed and identified by a robot during the construction process. After being covered by materials, it is integrated with the structure to form a stable internal positioning unit.

9. The biomimetic ant colony robot additive manufacturing system based on calibration dice space constraints according to claim 8, characterized in that: The fusion positioning module consists of a SLAM module, a UWB positioning module, and a visual recognition module. The three modules provide relative pose, absolute coordinates, and local fine-tuning data, respectively. The positioning module uses extended Kalman filtering and factor graph optimization algorithms for dynamic weighted fusion to output high-precision real-time pose.

10. The biomimetic ant colony robot additive manufacturing system based on calibration dice space constraints according to claim 8, characterized in that: The group scheduling platform is a multi-robot control and information synchronization center, responsible for task division, path allocation, obstacle avoidance scheduling and data fusion. It includes a communication base station, a task allocation module, a path optimization module and a payload computer. Each robot periodically uploads status data through a wireless network. The group scheduling platform performs global scheduling based on the ant colony optimization algorithm. The supply and testing module is located at the edge of the construction site and includes a material supply station, a charging station, and a testing station.

Images