Intelligent lifting and reloading system and method for high-density and fast-response small launch vehicles

Abstract

The present invention discloses an intelligent hoisting and reloading system and method for high-density and fast-response small launch vehicles. The system includes a real-time sensing unit: acquiring multi-source sensor data through intelligent identification and positioning of small targets at the lifting point, intelligent identification and centering of the transport vehicle bracket, centerline centering, relative posture measurement of the manipulator system and the rocket body, relative posture measurement of the manipulator system and the transport vehicle, analysis of the force situation of the manipulator system, and determination of the current state of the manipulator system; an intelligent decision-making and support unit: real-time analysis and decision-making of multi-source sensor data, and issuing tasks and instructions; a mechanism behavior control unit: receiving tasks and instructions and performing corresponding operations to complete the hoisting and reloading process of positioning-grasping-carrying-releasing-returning to the origin. The advantages are: the present invention improves the hoisting and reloading efficiency, reduces the intensity of manual operation, has the characteristics of safety, intelligence, efficiency, convenience, etc., and better adapts to the hoisting and reloading test needs of the field area.

Description

Technical Field

[0001] The present invention relates to the field of aerospace test and launch technology, and in particular to an intelligent lifting and transferring system and method for a high-density, fast-response small launch vehicle. Background Art

[0002] At present, the vast majority of launch sites in relevant countries still use manually assisted operations to perform rocket body hoisting and docking. Operators are required to observe the working environment, locate target hoisting points, install workpiece hoists, perform hoisting and transfer, identify the position of the bracket and accurately place the rocket body, and monitor the process status and safety. There are problems such as low efficiency of transfer operations, high labor intensity, and difficulty in standardizing operating procedures. Truly intelligent hoisting and transfer has not yet been achieved, which seriously restricts the test launch efficiency of small launch vehicles in high-density, ultra-intensity and complex operating environments. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent lifting and transferring system and method for high-density and fast-response small launch vehicles, thereby solving the above-mentioned problems existing in the prior art.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0005] An intelligent lifting and loading system for high-density and fast-response small launch vehicles, comprising:

[0006] Real-time perception unit: through intelligent identification and positioning of small targets at hanging points, intelligent identification and centering of the transport vehicle bracket, centerline centering, relative posture measurement between the manipulator system and the arrow body, relative posture measurement between the manipulator system and the transport vehicle, force analysis of the manipulator system, and determination of the current state of the manipulator system, it obtains multi-source sensor data, fuses the multi-source sensor data and transmits it to the intelligent decision-making and support unit;

[0007] Intelligent decision-making and support unit: used to conduct real-time analysis and decision-making on multi-source sensor data, determine the current task status and issue tasks and instructions;

[0008] Mechanism behavior control unit: used to receive tasks and instructions issued by the intelligent decision-making and support unit and perform corresponding operations to complete the hoisting and transfer process of positioning-grasping-handling-releasing-returning to the origin;

[0009] The mechanism behavior control unit includes an overhead crane system, a bridge crane system, a transport vehicle, a manipulator system and an integrated control system; the overhead crane system is used for horizontal movement to realize handling operations; the bridge crane system includes a lifting mechanism, a posture adjustment mechanism and an anti-sway mechanism; the lifting mechanism is used to realize the transportation, lifting and release of the arrow body; the posture adjustment mechanism can rotate the arrow body in the vertical direction to realize the centering of the axis of the arrow body and the transport vehicle; the anti-sway mechanism is used to prevent the shaking and swinging of the arrow body during the lifting process; the transport vehicle is used to carry and transport the arrow body; the manipulator system includes a mechanical arm, a lifting plate installed on the mechanical arm And a visual sensor, a strapping mechanism and a clamping mechanism installed on the lifting plate; the visual sensor is used to obtain image data in real time; the strapping mechanism is used to realize the strap feeding, strapping and locking operations of the strapping; the clamping mechanism is used to realize the grabbing and releasing operations of the arrow body; the integrated control system is used to receive decisions and instructions issued by the intelligent decision-making and support unit, and control the actions of the overhead crane system, the bridge crane system and the manipulator system to complete the lifting and transfer operations of the arrow body; the manipulator system is equipped with a visual sensor and a clamping mechanism for grabbing the arrow body; the arrow body is provided with a cooperative target and a grasped mechanism.

[0010] Preferably, the intelligent identification and positioning of the hanging point micro-target is to achieve high-precision identification and positioning of the cooperative target by using image processing and machine vision technology; specifically, it includes the following contents:

[0011] A1. Target design: Use geometric shapes and high-contrast colors recognized by visual sensors, introduce asymmetric images and specific codes to design target images;

[0012] A2. Target recognition: Use the trained deep learning model convolutional neural network to recognize and process the designed target image to ensure accurate recognition of the target image under various lighting and background conditions;

[0013] A3. Mark point center positioning: After edge detection using the Sobel operator, the Hough transform is applied to calculate the center coordinates of the marker point;

[0014] A4. Pose measurement: Capture the 3D spatial position of the target through the visual sensor and estimate the spatial pose using the 3D reconstruction algorithm and multi-view geometry theory;

[0015] A5. Hand-eye calibration: The iterative least squares method is used to accurately measure the relative position and direction between the vision sensor and the manipulator system to ensure that the movement of the manipulator system is synchronized with the measurement data of the vision sensor to improve positioning accuracy.

[0016] Preferably, the intelligent identification and centering of the transport vehicle bracket is to perform high-precision identification and accurate positioning of the transport vehicle bracket through image processing and deep learning algorithms;

[0017] Specifically, the visual sensor collects image data of the transport vehicle bracket in real time; uses the deep learning model convolutional neural network to extract features and identify targets from the image data; uses geometric calculation and pattern matching technology to determine the precise position and direction of the transport vehicle bracket; and adjusts the action of the bridge crane system in real time according to the precise position and direction to ensure accurate centering of the transport vehicle bracket.

[0018] Centerline alignment is to precisely align the reference line of the manipulator system with the centerline of the rocket body and the centerline of the rocket body with the reference line of the transport vehicle bracket;

[0019] Specifically, laser scanning and optical sensors are used to accurately capture the positions of reference lines and center lines. Through edge detection and image segmentation, image data is analyzed and acquired to determine the exact alignment points. The machine vision system combines artificial intelligence algorithms to automatically adjust the positions of the robotic system and transport vehicle to ensure the alignment of the corresponding lines. The alignment status is continuously monitored and adjusted through a feedback control loop to ensure the accuracy and repeatability of the alignment process.

[0020] Preferably, the relative posture measurement between the manipulator system and the arrow body is that the visual sensor captures the image of the cooperative target, and uses a visual method to measure the relative posture between the cooperative target and the visual sensor in real time;

[0021] Specifically, the cooperative target is identified and located through image processing technology, and then the spatial transformation from the visual sensor to the cooperative target is calculated, and the obtained relative posture information is converted into the relative posture between the clamping mechanism and the captured mechanism on the arrow body;

[0022] The relative position and posture of the manipulator system and the transport vehicle are measured by using the position and posture of the end effector of the manipulator system relative to the transport vehicle, capturing the image of the cooperative target through the visual sensor, and using the visual method to measure the relative position and posture between the cooperative target and the visual sensor in real time;

[0023] Specifically, the cooperative target is located through the visual system to identify its specific position and posture in the image; the spatial transformation from the visual sensor to the cooperative target is calculated using the visual sensor calibration parameters and three-dimensional vision technology; these data are then converted into the relative posture between the clamping mechanism and the clamping point of the transport vehicle.

[0024] Preferably, the force analysis of the manipulator system is an analysis of the tension, pressure, friction or shear force exerted on the belt wrapping mechanism and / or the clamping mechanism of the manipulator system when the manipulator system performs a task;

[0025] Specifically, by integrating force sensors and torque sensors, the force conditions of the strapping mechanism and / or clamping mechanism are monitored in real time; based on the feedback closed-loop control strategy, the tightness of the strapping mechanism and the action of the clamping mechanism are automatically adjusted according to the sensor data;

[0026] The current state of the robot system is determined to include:

[0027] Condition monitoring and evaluation: Obtain real-time data through the position, speed, and force / torque sensors of the manipulator system, and use the visual system and lidar to perceive environmental information, identify the position, shape, and size of objects, analyze the current status of the manipulator system and the surrounding environment, including the progress of completed tasks and any potential problems or anomalies, and evaluate the status of the manipulator system;

[0028] Task planning and decision-making: Based on the current state of the robot system and the preset target task, determine the next task and task sequence to be performed, calculate the best path to transfer from the current state to the next task, and generate the action sequence and control instructions required to complete the task;

[0029] Action execution: Send the control command sequence to the manipulator system, which moves to a specific position, executes the arrow clamping action, belt feeding, belt binding and locking operations, and continuously monitors the status of the manipulator system and environmental changes, and makes real-time adjustments when necessary;

[0030] Feedback and iteration: After the robot system completes the action, it evaluates whether the execution results are consistent with the expected goals, as well as any new situations and problems that arise. It makes necessary adjustments based on the evaluation results to optimize subsequent task planning and action execution.

[0031] Preferably, multi-source sensor data fusion includes the following:

[0032] B1. Use different types of sensors to collect observation target data of corresponding components;

[0033] B2. Use deep learning methods to extract features from sensor output data;

[0034] B3. Perform pattern recognition processing on the extracted feature vectors, using support vector machines and adaptive neural networks for classification and regression analysis to learn and identify complex patterns and associations from the features;

[0035] B4. Use data association technology and Kalman filtering technology to align and group the data after pattern recognition in time and space;

[0036] B5. Use multimodal fusion algorithm to integrate and optimize the aligned and grouped data.

[0037] Preferably, the specific working process of the intelligent decision-making and support unit is to first call the built-in model library and method library to analyze the multi-source sensor data; determine the current working stage of the manipulator system, decide the next stage task and the corresponding instruction sequence; send the next stage task and the corresponding instruction sequence to the mechanism behavior control unit.

[0038] Preferably, the posture adjustment mechanism includes a rotating mechanism and a swinging mechanism; the rotating mechanism and the swinging mechanism are driven by motors respectively to realize the rotation of the arrow body in a plane and the swinging of the arrow body in a vertical plane, thereby ensuring that the arrow body is in a suitable position and direction.

[0039] Preferably, the anti-sway mechanism uses an acceleration sensor and a gyroscope to detect the swing of the hoisted arrow body in real time, and adjusts the hook movement of the lifting mechanism by calculating vibration suppression instructions to reduce or eliminate the swing of the arrow body during movement.

[0040] The present invention also aims to provide a smart hoisting and reloading method for a high-density, fast-response small launch vehicle based on the system, comprising the following steps:

[0041] S1. Positioning: including coarse positioning and fine positioning; the coarse positioning is to complete the overall modeling of the target body by relying on the visual sensors on the four corners of the lifting plate of the manipulator system, detect the axis deviation of the manipulator system of the target body in real time during the descent of the manipulator system, and realize visual servo alignment by controlling the movement of the overhead crane system and the attitude adjustment mechanism, and end the coarse positioning when the manipulator system reaches the top of the target body;

[0042] The precise positioning is that the manipulator system continues to descend under the guidance of the visual sensors at the four corners of the lifting plate, and at the same time, with the assistance of the visual sensor guidance, continuously adjusts the horizontal direction until the image position of the rear lifting point coincides with the expected position, indicating that the manipulator system has reached the ideal position and the positioning operation is ended;

[0043] S2, grabbing: The strapping guide frame in the strapping mechanism extends out to guide the strapping to pass through the arrow body to complete the strapping action; when the strapping is detected to be in place, the strapping fixing pin completes the locking action, and finally the strapping guide frame disengages from contact and retracts in the reverse direction; the overhead crane system controls the manipulator system to continue to descend until the pressure sensor at the support position on the manipulator system detects that it is in place and the visual sensor detects that the sub-mouth is basically in the middle position, and then the manipulator system descends to the accurate position, indicating that the manipulator system has completed the axial positioning and radial positioning with the arrow body, and the grabbing is completed;

[0044] S3, handling: Under the guidance of the intelligent decision-making and support unit, the bridge crane system of the mechanism behavior control unit starts to perform the transfer operation; the anti-sway mechanism ensures that the vertical direction accuracy is controlled within the preset vertical accuracy, and the overhead crane system ensures that the horizontal movement accuracy is controlled within the preset horizontal accuracy, and the weight of the arrow body is all applied to the closed bag; during the transfer process, the real-time perception unit of the intelligent lifting and transfer system detects the entire process and the entire environment, and issues safety warnings for abnormal situations;

[0045] S4, release: The transport vehicle bracket is modeled as a whole through the visual sensor on the lifting plate; during the descent of the arrow body, the axis deviation between the arrow body target and the transport vehicle bracket is detected in real time, and the movement of the overhead crane system and the attitude adjustment mechanism is controlled to achieve visual servo alignment; at the same time, under the guidance of the visual sensor, the horizontal direction of the manipulator system is continuously adjusted until the rear lifting point completely falls into the slot of the transport vehicle, indicating that the arrow body has reached the ideal position and the release is completed; the strap guide frame is extended again to guide the strap to complete the retraction to achieve the loosening action;

[0046] S5. Return to origin: After the release phase, the intelligent lifting and transferring system returns to the origin, and each unit returns to the initial position or state, waiting for the next target to enter the warehouse or to issue other tasks or instructions.

[0047] The beneficial effects of the present invention are as follows: the present invention improves the efficiency of hoisting and transferring, reduces the intensity of manual operation, has the characteristics of safety, intelligence, efficiency, convenience, etc., and better adapts to the needs of hoisting and transferring testing in the field. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a structural schematic diagram of a hoisting and transferring system in an embodiment of the present invention;

[0049] Figure 2 It is a principle flow chart of the hoisting and transferring method in an embodiment of the present invention;

[0050] Figure 3 It is a schematic diagram of positioning in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation methods described herein are only used to explain the present invention and are not used to limit the present invention.

[0052] like Figure 1 As shown, in this embodiment, a smart lifting and loading system for a high-density, fast-response small launch vehicle is provided, comprising the following three units:

[0053] 1. Real-time Perception Unit

[0054] Multi-source sensor data is acquired through intelligent identification and positioning of tiny targets at hanging points, intelligent identification and centering of transport vehicle brackets, centerline centering, relative posture measurement of the manipulator system and the arrow body, relative posture measurement of the manipulator system and the transport vehicle, force analysis of the manipulator system and determination of the current state of the manipulator system, and then fused and transmitted to the intelligent decision-making and support unit.

[0055] In this embodiment, the goal of multi-source sensor information fusion is to derive more useful information by combining multi-level and multi-faceted information based on the separated observation information obtained by each sensor. On the one hand, it uses multiple sensors to operate collaboratively, and on the other hand, it comprehensively processes data from other information sources to improve the environmental perception ability and intelligent processing level of the lifting and loading system. Multi-source sensor fusion technology is divided into three levels: data layer fusion, feature layer fusion, and decision layer fusion. According to different data processing methods, there are three types of information fusion architectures: distributed, centralized, and hybrid.

[0056] Multi-source sensor information fusion refers to the process of using computer technology to automatically analyze and comprehensively process data from multiple sensors or information sources under certain criteria to complete the information processing required for decision-making and estimation.

[0057] The principle of multi-source sensor information fusion is like the process of comprehensive information processing in the human brain. It obtains multi-dimensional information from multiple sensors, and performs multi-level and multi-space information complementation and optimization combination processing on various sensor data, making full use of multi-source data for reasonable control and use, and forming a consistent interpretation of the observation environment. The main process is: First, multiple different types of sensors (such as cameras, lidars, inertial measurement units, etc.) collect observation target data such as manipulator systems and transport vehicle brackets. Second, a deep learning method is used to extract features from sensor output data. Specifically, a convolutional neural network (CNN) is used to automatically identify and learn useful features from data collected from different sensors (such as cameras, lidars, inertial measurement units, etc.). These features include but are not limited to shape, edge, angle, motion information, etc. The feature extraction module will be customized according to the needs of the observation target to ensure that the diversity and richness of the data are fully utilized in the algorithm training process, and extract the feature vector Yi representing the observation data. Third, the extracted feature vector Yi is processed for pattern recognition, and support vector machines (SVM) and adaptive neural networks (such as back propagation neural networks) are used for classification and regression analysis. Through these algorithms, complex patterns and associations are learned and identified from features, such as the operating status of the manipulator system or the moving trajectory of the transport vehicle, to complete the description of each sensor about the target. Fourth, data association techniques such as least squares estimation and Kalman filtering are used to align and group data from different sensors in time and space. This process ensures that the observation data of the same target can be effectively associated and integrated, thereby improving the accuracy and explanatory power of the data; fifth, multimodal fusion algorithms such as weighted averaging and multi-layer perceptron (MLP) networks are used to integrate and optimize the outputs of each sensor. These fusion algorithms not only improve the accuracy of target attribute judgment, but also make the lifting and reloading system more stable and reliable when processing data in complex environments. Multi-source sensor information fusion methods are basically divided into two categories: random and artificial intelligence, which can be used for transport vehicle positioning, tracking, navigation, and mechanism behavior control units.

[0058] 1.1. Intelligent identification and positioning of small targets at hanging points uses advanced image processing and machine vision technologies to achieve high-precision identification and positioning of cooperative targets, including key steps such as target design, target identification, landmark center positioning, posture measurement, and hand-eye calibration. In the target design stage, geometric figures and high-contrast colors that are easy for machine vision systems to identify are used, and asymmetric figures and specific codes are deliberately introduced to maintain recognition consistency and reliability under the environment. The target recognition link uses an improved deep learning model, convolutional neural network (CNN), which is specially trained to identify and process designed target images to ensure accurate identification under a variety of lighting and background conditions. The landmark center positioning is determined by high-precision edge detection algorithms and geometric analysis techniques. After edge detection using the Sobel operator, the Hough transform is applied to calculate the center coordinates. Pose measurement combines stereo vision technology and point cloud analysis. The three-dimensional spatial position of the target is captured by a dual-camera system, and the spatial posture is estimated using a three-dimensional reconstruction algorithm and multi-view geometry theory. During the hand-eye calibration process, the iterative least squares method is used to accurately determine the relative position and direction between the camera and the robotic arm to ensure that the robot's movement is synchronized with the measurement data of the visual system and improve positioning accuracy. These steps use specially developed or improved algorithms rather than existing processing procedures to ensure the efficiency and accuracy of the lifting and loading system in complex industrial environments.

[0059] 1.2. Intelligent identification and alignment of the transport vehicle bracket uses modern information technology to obtain image data through cameras or sensors, and comprehensively applies machine vision, deep learning and sensor technology. In complex environments, the hoisting and transfer system uses advanced image processing and deep learning algorithms to perform high-precision identification and accurate positioning of the transport vehicle bracket. The specific implementation process includes: first, the camera or visual sensor collects real-time image data; second, the deep learning model convolutional neural network (CNN) is applied to extract features and identify targets from the image; then, geometric calculations and pattern matching techniques are used to determine the precise position and direction of the bracket; finally, the hoisting and transfer system adjusts the hoisting equipment action in real time according to the recognition results to ensure accurate alignment of the bracket. These technologies not only improve the accuracy of identification and positioning, but also ensure the stability and reliability of operation in complex industrial environments.

[0060] 1.3. Centerline alignment is a key step in the lifting and transfer operation, which involves accurately aligning the reference line of the manipulator system with the centerline of the rocket body and the centerline of the rocket body with the reference line of the transport vehicle bracket. This process uses a variety of technical means to ensure high precision and quality. First, laser scanning and optical sensors are used to accurately capture the position of the reference line and centerline, providing millimeter-level positioning accuracy. Then, the acquired image data is analyzed through real-time image processing technology, including edge detection and image segmentation, to determine the exact alignment point. In addition, the machine vision system combines artificial intelligence algorithms, such as deep learning networks, to automatically adjust the position of the manipulator system and the transport vehicle to ensure that all lines are perfectly aligned. Finally, the lifting and transfer system continuously monitors and adjusts the alignment status through a feedback control loop to ensure the accuracy and repeatability of the entire process. These integrated technical means have greatly improved the efficiency and quality of operations.

[0061] 1.4. The relative position and orientation of the manipulator system and the rocket body are represented by a 4×4 homogeneous transformation matrix, which includes a 3×3 rotation matrix and a 3×1 position vector, which together express the position and orientation of the object in space. The manipulator system is equipped with a fixed visual sensor (hand-eye camera) and a clamping mechanism, while the rocket body is equipped with a cooperative target (an auxiliary measurement mark with a known fixed shape) and a grasped mechanism. The visual sensor captures the image of the cooperative target and uses a visual method to measure the relative position and orientation between the cooperative target and the hand-eye camera in real time. This measurement process first identifies and locates the target through image processing technology, and then calculates the spatial transformation from the camera to the target, which involves parsing the feature points in the target image and applying the camera calibration parameters. The relative position and orientation information obtained is then converted into the relative position and orientation between the clamping mechanism and the grasped mechanism, which is accomplished by applying inverse kinematics and a fixed transformation from the camera to the end of the manipulator system to ensure that the manipulator system can accurately plan and execute the grasping path. This process not only enhances the accuracy of the grasping operation, but also improves the efficiency of the automated operation.

[0062] 1.5. The relative pose of the manipulator system and the transport vehicle is defined by the position and posture of the manipulator system end effector relative to the transport vehicle, and the image of the cooperative target is captured by the visual sensor, and the relative pose between the cooperative target and the hand-eye camera is measured in real time using a visual method. The process begins by locating the cooperative target through the visual system and identifying its specific position and posture in the image. Next, the hoisting and transfer system uses camera calibration parameters and 3D vision technology to calculate the spatial transformation from the camera to the target. These data are then converted into the relative pose between the clamping mechanism and the clamping point of the transport vehicle. Specifically, once the accurate relative pose between the clamping mechanism and the clamping point of the transport vehicle is obtained, this information will be used as the basis for the manipulator system to plan the release path. The hoisting and transfer system generates an operation instruction containing a specific manipulator system motion sequence and path by analyzing this relative pose. This includes the translation and rotation of the manipulator system to ensure that the end effector is accurately aligned with the clamping point of the transport vehicle. Subsequently, the manipulator system action is adjusted in real time to respond to any changes in the actual position and posture of the transport vehicle. This adjustment is performed continuously, ensuring that the movement of the manipulator system and the alignment of the gripping point of the transporter are always kept in an optimal state. This precise movement adjustment is achieved through an advanced feedback control system, which responds to the data collected from the vision sensor and other sensors (such as displacement sensors, force sensors, etc.), so as to ensure the safe and efficient completion of the task. In this way, the manipulator system can work effectively in a dynamic environment, adapt to small changes in the position of the transporter, and ensure the accuracy and reliability of operation.

[0063] 1.6. The force condition of the strapping of the manipulator system is the analysis of the tension, pressure, friction or shear force of the strapping mechanism or clamping mechanism when the manipulator system performs tasks, so as to ensure the stability and safety of the manipulator system in operation and the integrity of the captured arrow body. By integrating force sensors and torque sensors, the force condition on the strapping is monitored in real time for fine-tuning and optimization; the closed-loop control strategy based on feedback automatically adjusts the tightness of the strapping and the action of the manipulator system according to the sensor data.

[0064] 1.7. Determining the current working stage of the manipulator system and deciding the next stage of tasks and the corresponding instruction sequence are the keys to achieving effective control and automation of the manipulator system, involving steps such as state monitoring, task planning, decision making and action execution, including: First, state monitoring and evaluation, obtaining real-time data through sensors such as position, speed, force / torque, etc. on the manipulator system, and using visual systems, lidar and other environmental information to perceive the position, shape and size of objects, analyze the current state of the manipulator system and the surrounding environment, including the progress of completed tasks and any potential problems or anomalies, and evaluate the system state; second, task planning and decision making, based on The current state of the manipulator system and the preset target task determine the next task to be performed, determine the task sequence, calculate the best path to transfer from the current state to the next task, and generate the action sequence and control instructions required to complete the task; third, action execution, sending the control instruction sequence to the manipulator system, the manipulator system moves to a specific position, performs actions such as clamping the arrow body, and continuously monitors the state of the manipulator system and environmental changes, and makes real-time adjustments when necessary; fourth, feedback and iteration, after the action is completed, evaluate whether the execution result is consistent with the expected goal, as well as new situations and problems that arise, make necessary adjustments based on the evaluation results, and optimize subsequent task planning and action execution.

[0065] The whole process requires a high degree of integration and coordination, involving the interaction of multiple subsystems, including but not limited to the manipulator system, sensor system, environmental perception system and decision support system. Through real-time data flow and fast decision-making cycle, the manipulator system can complete complex tasks efficiently and accurately. In engineering practice, factors such as the physical limitations of the manipulator system, work safety and task priority must also be considered to achieve the optimal strategy for task execution.

[0066] 2. Intelligent Decision-making and Support Unit

[0067] Used to conduct real-time analysis and decision-making on multi-source sensor data, determine the current task status and issue tasks and instructions.

[0068] The decision support steps of the intelligent decision-making and support unit are mainly as follows: first, calling the built-in model library and method library to analyze the multi-source sensor information, including the intelligent recognition and positioning of tiny targets at the hanging points, the intelligent recognition and centering of the transport vehicle bracket under complex backgrounds, centerline alignment, the relative position of the manipulator system, the rocket body and the transport vehicle, and the stress condition of the straps of the manipulator system; second, judging the current working stage of the manipulator system, deciding the next stage task and the corresponding instruction sequence; finally, the driving mechanism behavior control unit performs corresponding operations.

[0069] 3. Institutional Behavior Control Unit

[0070] It is used to receive tasks and instructions issued by the intelligent decision-making and support unit and perform corresponding operations to complete the lifting and transfer process of positioning - grasping - transporting - releasing - returning to the origin.

[0071] The mechanism behavior control unit includes the overhead crane system, bridge crane system, transport vehicle, manipulator system and integrated control system, and all components are required to be explosion-proof.

[0072] The overhead crane system is used for horizontal movement to realize the handling operation. The bridge crane system includes a lifting mechanism, a posture adjustment mechanism and an anti-sway mechanism; the lifting mechanism is used to realize the transportation, lifting and release of the arrow body; the posture adjustment mechanism can rotate the arrow body in the vertical direction to realize the centering of the axis of the arrow body and the transport vehicle; the anti-sway mechanism is used to prevent the shaking and swinging of the arrow body during the lifting process. The transport vehicle is used to carry and transport the arrow body. The manipulator system includes a mechanical arm, a lifting plate installed on the mechanical arm, and a visual sensor, a belt wrapping mechanism and a clamping mechanism installed on the lifting plate; the visual sensor is used to obtain image data in real time; the belt wrapping mechanism is used to realize the belt feeding, strapping and locking operations of the belt wrapping; the clamping mechanism is used to realize the grabbing and releasing operations of the arrow body. The integrated control system is used to receive the decisions and instructions issued by the intelligent decision-making and support unit, and control the actions of the overhead crane system, the bridge crane system and the manipulator system to complete the lifting and reloading operations of the arrow body; the manipulator system is equipped with a visual sensor and a clamping mechanism for grabbing the arrow body; the arrow body is provided with a cooperative target and a grasping mechanism.

[0073] The mechanism behavior control unit needs to consider the specific needs, working environment and operating range of the manipulator system, and ensure the high performance and high reliability of the system by integrating advanced sensors, efficient actuators, intelligent control algorithms and a good user interface.

[0074] The bridge-type lifting system comprises a lifting mechanism, a posture adjustment mechanism and an anti-sway mechanism, which cooperate with each other to ensure the efficiency, accuracy and safety of the lifting operation; the lifting mechanism is the core part of the bridge-type lifting system, and is responsible for lifting or lowering the suspended arrow body; the posture adjustment mechanism comprises a rotating mechanism and a swinging mechanism, etc., which are driven by a motor to realize the rotation of the arrow body in a plane or the swing of the arrow body in a vertical plane, ensuring that it is in a suitable position and direction; the anti-sway mechanism uses acceleration sensors, gyroscopes and other equipment to monitor the swing of the suspended arrow body in real time, and calculates the vibration suppression command through the control system to adjust the hook movement, thereby reducing or eliminating the swing of the suspended arrow body during movement, and improving the stability and safety of the operation.

[0075] The manipulator system includes a manipulator arm, an end effector (a visual sensor, a belt wrapping mechanism and a clamping mechanism arranged on the lifting plate), a control system and other components. The visual sensor (visual system) includes four-corner visual sensors and a target visual sensor arranged on the lifting plate.

[0076] The comprehensive control system integrates multiple subsystems such as the overhead crane system, bridge crane system, and manipulator system through advanced control technology, realizes coordinated control and data exchange of each subsystem, and supports remote monitoring and fault diagnosis.

[0077] Embodiment 2

[0078] like Figure 2 As shown, this embodiment provides a smart lifting and reloading method for high-density fast-response small launch vehicles. The lifting and reloading process can be decomposed into five stages: positioning - grabbing - handling - releasing - origin, among which the grabbing and releasing stages are more critical. Taking the transfer of the rocket body from the AGV trolley to the transport vehicle as an example, the smart lifting and reloading methods mainly include:

[0079] 1. Positioning (such as Figure 3 (shown)

[0080] The positioning stage includes coarse positioning and fine positioning; the purpose of coarse positioning is to make the strapping manipulator system basically parallel to the axis of the arrow body, mainly relying on the visual sensors on the four corners of the lifting plate to complete the overall modeling of the arrow body target, and to detect the deviation between the arrow body target and the axis of the manipulator system in real time during the descent of the manipulator system, and to achieve visual servo alignment by controlling the movement of the overhead crane system and the attitude adjustment mechanism, and to end the coarse positioning when the manipulator system reaches the top of the arrow body target; the purpose of fine positioning is, on the one hand, to make the manipulator system and the axis of the arrow body coincide accurately, and on the other hand, to determine the position deviation between the manipulator system and the axis of the arrow body. Following the above coarse positioning, the manipulator system continues to descend under the guidance of the visual sensors at the four corners of the lifting plate, and at the same time, with the guidance of the target visual sensor, it continuously adjusts the horizontal direction until the image position of the rear lifting point coincides with the expected position, indicating that the manipulator system has reached the ideal position and the positioning operation is ended.

[0081] 2. Grasping: The intelligent lifting and reloading system completes the grabbing operation of the target body of the arrow through the manipulator system. Through active control, the strap guide frame extends out to guide the strap through the arrow body to complete the strapping action. A gap is reserved between the strap guide frame and the arrow body to meet the cable avoidance constraints in the design requirements. When the strap is detected to be in place, the strap fixing pin completes the locking action, and finally the strap guide frame disengages from contact and retracts into the gripper body in the reverse direction. After alignment, the visual sensor is retracted. At this time, the manipulator is 50mm away from the arrow body. The overhead crane system controls the manipulator system to continue to descend until the pressure sensor at the support position on the manipulator system detects that it is in place and the visual sensor detects that the sub-mouth (the visual sensor has been retracted at this time) is basically in the middle position. The manipulator system descends to the accurate position, indicating that the manipulator system has completed the axial positioning and radial positioning with the arrow body and the grasping is completed.

[0082] 3. Transfer: Under the guidance of the intelligent decision-making and support unit, the bridge crane system of the mechanism behavior control unit begins to perform the transfer operation. The anti-sway mechanism ensures that the vertical direction accuracy is controlled within 5mm, and the overhead crane system ensures that the horizontal movement accuracy is controlled within 5mm. The weight of the arrow body is all applied to the closed bag. During the transfer process, the real-time perception unit of the intelligent lifting and reloading system detects the entire process and the entire environment, and issues safety warnings for abnormal situations to ensure that the entire transfer stage can not only meet the accuracy requirements well, but also ensure operational safety.

[0083] 4. Release: It includes two steps: placement and loosening. The placement is similar to the precise positioning process. The visual sensor on the lifting plate is used to model the transport vehicle bracket as a whole. During the descent of the arrow body, the axis deviation between the target and the transport vehicle bracket is detected in real time, and the movement of the overhead crane system and the attitude adjustment mechanism is controlled to achieve visual servo alignment. At the same time, under the guidance of the target visual sensor, the horizontal direction of the manipulator system is continuously adjusted until the rear lifting point completely falls into the slot of the transport vehicle, indicating that the arrow body has reached the ideal position and the release stage placement step is completed. The loosening process is the reverse movement of the grasping process. The strap guide frame extends again to guide the strap to complete the retraction to achieve the loosening action.

[0084] 5. Return to origin: After the release phase, the intelligent lifting and transferring system enters the origin phase, and each unit returns to the initial position, waiting for the next target to enter the warehouse or issuing other decision-making tasks or instructions.

[0085] By adopting the above technical solution disclosed in the present invention, the following beneficial effects are obtained:

[0086] The present invention provides an intelligent lifting and transferring system and method for high-density and fast-response small launch vehicles. The present invention improves the lifting and transferring efficiency, reduces the intensity of manual operation, has the characteristics of safety, intelligence, efficiency, convenience, etc., and better adapts to the lifting and transferring testing needs of the site.

[0087] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. An intelligent lifting and transfer system for high-density and fast-response small launch vehicles, characterized by: include, Real-time perception unit: through intelligent identification and positioning of small targets at hanging points, intelligent identification and centering of the transport vehicle bracket, centerline centering, relative posture measurement between the manipulator system and the arrow body, relative posture measurement between the manipulator system and the transport vehicle, force analysis of the manipulator system, and determination of the current state of the manipulator system, it obtains multi-source sensor data, fuses the multi-source sensor data and transmits it to the intelligent decision-making and support unit; Intelligent decision-making and support unit: used to conduct real-time analysis and decision-making on multi-source sensor data, determine the current task status and issue tasks and instructions; Mechanism behavior control unit: used to receive tasks and instructions issued by the intelligent decision-making and support unit and perform corresponding operations to complete the hoisting and transfer process of positioning-grasping-handling-releasing-returning to the origin; The mechanism behavior control unit includes an overhead crane system, a bridge crane system, a transport vehicle, a manipulator system and an integrated control system; the overhead crane system is used for horizontal movement to realize handling operations; the bridge crane system includes a lifting mechanism, a posture adjustment mechanism and an anti-sway mechanism; the lifting mechanism is used to realize the transportation, lifting and release of the arrow body; the posture adjustment mechanism can rotate the arrow body in the vertical direction to realize the centering of the axis of the arrow body and the transport vehicle; the anti-sway mechanism is used to prevent the shaking and swinging of the arrow body during the lifting process; the transport vehicle is used to carry and transport the arrow body; the manipulator system includes a mechanical arm, a lifting plate installed on the mechanical arm And a visual sensor, a strapping mechanism and a clamping mechanism installed on the lifting plate; the visual sensor is used to obtain image data in real time; the strapping mechanism is used to realize the strap feeding, strapping and locking operations of the strapping; the clamping mechanism is used to realize the grabbing and releasing operations of the arrow body; the integrated control system is used to receive decisions and instructions issued by the intelligent decision-making and support unit, and control the actions of the overhead crane system, the bridge crane system and the manipulator system to complete the lifting and transfer operations of the arrow body; the manipulator system is equipped with a visual sensor and a clamping mechanism for grabbing the arrow body; the arrow body is provided with a cooperative target and a grasped mechanism.

2. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: Intelligent identification and positioning of small targets at hanging points is to achieve high-precision identification and positioning of cooperative targets by using image processing and machine vision technology; specifically, it includes the following contents: A1. Target design: Using geometric shapes and high-contrast colors recognized by visual sensors, introducing asymmetric images and specific codes to design target images; A2. Target recognition: Use the trained deep learning model convolutional neural network to recognize and process the designed target image to ensure accurate recognition of the target image under various lighting and background conditions; A3. Mark point center positioning: After edge detection using the Sobel operator, the Hough transform is applied to calculate the coordinates of the center of the marker point; A4. Pose measurement: Capture the 3D spatial position of the target through the visual sensor and estimate the spatial pose using the 3D reconstruction algorithm and multi-view geometry theory; A5. Hand-eye calibration: The iterative least squares method is used to accurately measure the relative position and direction between the vision sensor and the manipulator system to ensure that the movement of the manipulator system is synchronized with the measurement data of the vision sensor to improve positioning accuracy.

3. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: Intelligent identification and centering of transport vehicle brackets: high-precision identification and accurate positioning of transport vehicle brackets through image processing and deep learning algorithms; Specifically, the visual sensor collects image data of the transport vehicle bracket in real time; uses the deep learning model convolutional neural network to extract features and identify targets from the image data; uses geometric calculation and pattern matching technology to determine the precise position and direction of the transport vehicle bracket; and adjusts the action of the bridge crane system in real time according to the precise position and direction to ensure accurate centering of the transport vehicle bracket. Centerline alignment is to precisely align the reference line of the manipulator system with the centerline of the rocket body and the centerline of the rocket body with the reference line of the transport vehicle bracket; Specifically, laser scanning and optical sensors are used to accurately capture the positions of reference lines and center lines. Image data is analyzed and acquired through edge detection and image segmentation to determine the exact alignment points. The machine vision system, combined with artificial intelligence algorithms, automatically adjusts the positions of the manipulator system and transport vehicle to ensure that the corresponding lines are aligned. The alignment status is continuously monitored and adjusted via a feedback control loop, ensuring the accuracy and repeatability of the alignment process.

4. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: The relative posture measurement between the manipulator system and the arrow body is that the visual sensor captures the image of the cooperative target and uses the visual method to measure the relative posture between the cooperative target and the visual sensor in real time; Specifically, the cooperative target is identified and located through image processing technology, and then the spatial transformation from the visual sensor to the cooperative target is calculated, and the obtained relative posture information is converted into the relative posture between the clamping mechanism and the captured mechanism on the arrow body; The relative position and posture of the manipulator system and the transport vehicle are measured by using the position and posture of the end effector of the manipulator system relative to the transport vehicle, capturing the image of the cooperative target through the visual sensor, and using the visual method to measure the relative position and posture between the cooperative target and the visual sensor in real time; Specifically, the cooperative target is located through the visual system to identify its specific position and posture in the image; The spatial transformation from the vision sensor to the cooperative target is calculated using the vision sensor calibration parameters and 3D vision technology; this data is then converted into the relative position between the clamping mechanism and the clamping point of the transport vehicle.

5. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: The force analysis of the manipulator system is the analysis of the tension, pressure, friction or shear force on the belt wrapping mechanism and / or clamping mechanism when the manipulator system performs tasks; Specifically, by integrating force sensors and torque sensors, the force conditions of the strapping mechanism and / or clamping mechanism are monitored in real time; based on the feedback closed-loop control strategy, the tightness of the strapping mechanism and the action of the clamping mechanism are automatically adjusted according to the sensor data; The current state of the robot system is determined to include: Condition monitoring and evaluation: Obtain real-time data through the position, speed, and force / torque sensors of the manipulator system, and use the visual system and lidar to perceive environmental information, identify the position, shape, and size of objects, analyze the current status of the manipulator system and the surrounding environment, including the progress of completed tasks and any potential problems or anomalies, and evaluate the status of the manipulator system; Task planning and decision-making: Based on the current state of the robot system and the preset target task, determine the next task and task sequence to be performed, calculate the best path to transfer from the current state to the next task, and generate the action sequence and control instructions required to complete the task; Action execution: Send the control command sequence to the manipulator system, which moves to a specific position, executes the arrow clamping action, belt feeding, belt binding and locking operations, and continuously monitors the status of the manipulator system and environmental changes, and makes real-time adjustments when necessary; Feedback and iteration: After the robot system completes the action, it evaluates whether the execution results are consistent with the expected goals, as well as any new situations and problems that arise. It makes necessary adjustments based on the evaluation results to optimize subsequent task planning and action execution.

6. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: Multi-source sensor data fusion includes the following: B1. Use different types of sensors to collect observation target data of corresponding components; B2. Use deep learning methods to extract features from sensor output data; B3. Perform pattern recognition processing on the extracted feature vectors, using support vector machines and adaptive neural networks for classification and regression analysis to learn and identify complex patterns and associations from the features; B4. Use data association technology and Kalman filtering technology to align and group the data after pattern recognition in time and space; B5. Use multimodal fusion algorithm to integrate and optimize the aligned and grouped data.

7. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: The specific working process of the intelligent decision-making and support unit is as follows: first, call the built-in model library and method library to analyze the multi-source sensor data; determine the current working stage of the manipulator system, decide the next stage task and the corresponding instruction sequence; send the next stage task and the corresponding instruction sequence to the mechanism behavior control unit.

8. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: The posture adjustment mechanism includes a rotating mechanism and a swinging mechanism; the rotating mechanism and the swinging mechanism are driven by motors respectively to realize the rotation of the arrow body in a plane and the swinging in a vertical plane, so as to ensure that the arrow body is in a suitable position and direction.

9. The intelligent lifting and loading system for high-density and fast-response small launch vehicles according to claim 1 is characterized in that: The anti-sway mechanism uses an acceleration sensor and a gyroscope to detect the swing of the hoisted arrow body in real time, and adjusts the hook movement of the lifting mechanism by calculating the vibration suppression instruction to reduce or eliminate the swing of the arrow body during movement.

10. An intelligent hoisting and reloading method for a high-density, fast-response small launch vehicle based on the system of any one of claims 1 to 9, characterized in that: The following steps are included: S1. Positioning: including coarse positioning and fine positioning; the coarse positioning is to complete the overall modeling of the target body by relying on the visual sensors on the four corners of the lifting plate of the manipulator system, detect the axis deviation of the manipulator system of the target body in real time during the descent of the manipulator system, and realize visual servo alignment by controlling the movement of the overhead crane system and the attitude adjustment mechanism, and end the coarse positioning when the manipulator system reaches the top of the target body; The precise positioning is that the manipulator system continues to descend under the guidance of the visual sensors at the four corners of the lifting plate, and at the same time, with the assistance of the visual sensor guidance, continuously adjusts the horizontal direction until the image position of the rear lifting point coincides with the expected position, indicating that the manipulator system has reached the ideal position and the positioning operation is ended; S2, grabbing: The strapping guide frame in the strapping mechanism extends out to guide the strapping to pass through the arrow body to complete the strapping action; when the strapping is detected to be in place, the strapping fixing pin completes the locking action, and finally the strapping guide frame disengages from contact and retracts in the reverse direction; the overhead crane system controls the manipulator system to continue to descend until the pressure sensor at the support position on the manipulator system detects that it is in place and the visual sensor detects that the sub-mouth is basically in the middle position, and then the manipulator system descends to the accurate position, indicating that the manipulator system has completed the axial positioning and radial positioning with the arrow body, and the grabbing is completed; S3, handling: Under the guidance of the intelligent decision-making and support unit, the bridge crane system of the mechanism behavior control unit starts to perform the transfer operation; the anti-sway mechanism ensures that the vertical direction accuracy is controlled within the preset vertical accuracy, and the overhead crane system ensures that the horizontal movement accuracy is controlled within the preset horizontal accuracy, and the weight of the arrow body is all applied to the closed bag; during the transfer process, the real-time perception unit of the intelligent lifting and transfer system detects the entire process and the entire environment, and issues safety warnings for abnormal situations; S4, release: The transport vehicle bracket is modeled as a whole through the visual sensor on the lifting plate; during the descent of the arrow body, the axis deviation between the arrow body target and the transport vehicle bracket is detected in real time, and the movement of the overhead crane system and the attitude adjustment mechanism is controlled to achieve visual servo alignment; at the same time, under the guidance of the visual sensor, the horizontal direction of the manipulator system is continuously adjusted until the rear lifting point completely falls into the slot of the transport vehicle, indicating that the arrow body has reached the ideal position and the release is completed; the strap guide frame is extended again to guide the strap to complete the retraction to achieve the loosening action; S5. Return to origin: After the release phase, the intelligent lifting and transferring system returns to the origin, and each unit returns to the initial position or state, waiting for the next target to enter the warehouse or to issue other tasks or instructions.

Images