CN117214908B

CN117214908B - Positioning control method and system based on intelligent cable cutting machine

Info

Publication number: CN117214908B
Application number: CN202311310307.7A
Authority: CN
Inventors: 李红斌; 易玉辉; 陈名文
Original assignee: Shenzhen Yuxuntong Photoelectricity Co ltd
Current assignee: Shenzhen Yuxuntong Photoelectricity Co ltd
Priority date: 2023-10-10
Filing date: 2023-10-10
Publication date: 2024-05-10
Anticipated expiration: 2043-10-10
Also published as: CN117214908A

Abstract

The invention relates to the technical field of positioning control, in particular to a positioning control method and system based on an intelligent cable cutting machine. The method comprises the following steps: acquiring cable cutting mechanical environment data by using a camera; performing environment scanning through a laser radar to obtain obstacle position data; carrying out three-dimensional point cloud modeling on the cable cutting mechanical environment data and the obstacle position data to construct an environment map; performing target positioning on a target object through an environment map to generate a target boundary box; performing pixel level segmentation on the target boundary box to generate a target instance segmentation map; acquiring a first position parameter of a cable cutting machine; planning a path of a first position parameter of the cable cutting machine through a target instance segmentation map so as to generate a first working path; obstacle path optimization is performed on the first working path through the environment map to construct a second working path. The invention realizes the high-efficiency and accurate positioning control of the cable cutting machine.

Description

Positioning control method and system based on intelligent cable cutting machine

Technical Field

The invention relates to the technical field of positioning control, in particular to a positioning control method and system based on an intelligent cable cutting machine.

Background

With the rapid development of automation of cable cutting machinery industry and the popularization of intelligent manufacturing, positioning control technology plays a vital role in various production fields. Positioning control methods face a number of challenges in dealing with complex, high precision, high efficiency production requirements, including problems with accurate positioning, real-time response, automated coordination, and security. With the increasing intelligence and interconnection of manufacturing equipment, positioning control systems face more and more complex situations, such as dynamic working environments, adaptive production, highly customized requirements, and the like, and conventional positioning control methods generally rely on preprogrammed tracks and fixed sensor systems, which limit their applicability in coping with diverse production environments, and the positioning control methods often have problems of low positioning control precision and low efficiency, so that an intelligent positioning control method and system based on intelligent cable cutting machinery are needed.

Disclosure of Invention

The invention provides a positioning control method and a positioning control system based on an intelligent cable cutting machine for solving at least one technical problem.

In order to achieve the above purpose, the invention provides a positioning control method based on an intelligent cable cutting machine, which comprises the following steps:

Step S1: acquiring cable cutting mechanical environment data by using a camera; performing environment scanning through a laser radar to obtain obstacle position data; carrying out three-dimensional point cloud modeling on the cable cutting mechanical environment data and the obstacle position data to construct an environment map;

Step S2: performing target positioning on a target object through an environment map to generate a target boundary box; performing pixel level segmentation on the target boundary box to generate a target instance segmentation map;

Step S3: acquiring a first position parameter of a cable cutting machine; planning a path of a first position parameter of the cable cutting machine through a target instance segmentation map so as to generate a first working path; performing obstacle path optimization on the first working path through an environment map to construct a second working path;

Step S4: carrying out dynamic characteristic analysis on the cable cutting machine to generate motion constraint data; planning the path speed of the second working path based on the motion constraint data to generate the path speed; performing action serialization analysis on the second working path through the path speed to generate action serialization data; performing track fitting on the second working path according to the motion serialization data to generate a cable cutting mechanical track;

Step S5: detecting the position of the cable cutting machine in real time to obtain a second position parameter of the cable cutting machine; carrying out target attitude error analysis on the target instance segmentation map through a second position parameter of the cable cutting machine so as to generate target attitude error data; performing self-adaptive attitude adjustment analysis through the target attitude error data to generate a self-adaptive attitude adjustment strategy;

step S6: carrying out real-time dynamic track optimization on the cable cutting mechanical track based on the self-adaptive posture adjustment strategy so as to generate a dynamic optimization track; and carrying out data mining modeling on the dynamic optimization track to construct a cable cutting machine dynamic track model so as to execute positioning control operation.

According to the invention, the camera is used for acquiring the mechanical environment data of the cable cutting machine, and the laser radar is used for carrying out environment scanning so as to acquire the position data of the obstacle. The data sources provide real-time information about the robot working environment, and the acquired data are used for creating a three-dimensional point cloud model to construct a mechanical environment map. This helps the robot understand the shape and obstacle distribution of its surroundings, locate the target object using the environment map, and generate a target bounding box. This facilitates the robot to identify and lock the target object to be operated upon, pixel-level segmentation of the target bounding box, generating a target instance segmentation map. This allows the robot to accurately identify the edge and shape of the target object, and to generate a first working path by path planning through the target instance segmentation map using the first position parameters of the cable cutting machine. This ensures that the robot can reach the target, uses the environment map to optimize the obstacle path for the first working path to ensure that the robot can avoid the obstacle when performing tasks, and performs kinetic characteristics analysis on the cable cutting machine to generate motion constraint data. This helps to ensure that the movement of the robot is physically feasible, and the path speed planning is performed on the second working path based on the movement constraint data to generate the path speed. The method ensures that the robot keeps stable and safe in the moving process, performs action serialization analysis on the second working path through the path speed to generate action serialization data, and then performs track fitting on the second working path to generate the actual movement track of the cable cutting machine. This helps the motion of actual control robot, detects in real time to cut cable machinery orbit to obtain cut cable machinery's second position parameter. This ensures that the robot can track its position in the task execution, and through the second position parameter of the cable cutting machine, the target instance segmentation map is subjected to target attitude error analysis to generate target attitude error data. This helps determine if the robot is properly aimed at the target, generates an adaptive pose adjustment strategy based on the target pose error data, allowing the robot to pose adjust while performing tasks to ensure accuracy. And carrying out real-time dynamic track optimization on the cable cutting mechanical track based on the self-adaptive posture adjustment strategy so as to generate a dynamic optimization track. The method is beneficial to the robot to adapt to continuously changing conditions when the robot executes tasks, and the dynamic optimization track is subjected to data mining modeling so as to construct a dynamic track model of the cable cutting machine. This model can be used for future mission planning and control.

Preferably, step S1 comprises the steps of:

Step S11: acquiring cable cutting mechanical environment data by using a camera;

Step S12: performing environment sensing scanning through a laser radar to obtain obstacle position data;

Step S13: performing three-dimensional point cloud data conversion on the cable cutting mechanical environment data and the obstacle position data to generate environment point cloud data and obstacle point cloud data;

Step S14: respectively performing point cloud segmentation on the environmental point cloud data and the obstacle point cloud data to generate environmental points cloud mass and obstacle point cloud clusters;

step S15: and carrying out three-dimensional point cloud modeling on the environment points cloud mass and the obstacle point cloud clusters to construct an environment map.

According to the invention, the camera is used for acquiring the environmental data around the cable cutting machine. The camera provides a visible light image, is used for detecting and identifying visible objects and visual information for positioning and navigation, and performs environment sensing scanning through the laser radar to acquire position data of obstacles in the environment. The laser radar can accurately measure the distance and shape of an object, provide high-resolution environment sensing information, process data acquired by the camera and the laser radar, and convert the data into three-dimensional point cloud data. This will help integrate the information of the different sensors for subsequent analysis and modeling, splitting the environmental and obstacle point cloud data into different points cloud mass and point cloud clusters. The method is helpful for separating different objects and barriers in the environment, so that the subsequent modeling is more accurate, and the environment map is constructed by performing three-dimensional point cloud modeling by using the environment points cloud mass and the barrier point cloud clusters. The environment map is an accurate three-dimensional model reflecting the actual conditions of the machine operating environment.

Preferably, step S2 comprises the steps of:

step S21: positioning a target through an environment map to obtain a target object position parameter;

step S22: performing boundary marking on the target positioning based on the target object position parameters to generate a target boundary frame;

Step S23: performing boundary clipping on the target boundary frame to generate a target boundary frame area;

Step S24: performing pixel level segmentation on the target boundary box area to generate a segmentation result, wherein the segmentation result comprises a target pixel label and a target boundary box background;

Step S25: and carrying out segmentation mapping on the environment map through the segmentation result to generate a target instance segmentation map.

According to the invention, through the environment map, the system can position the target according to the position parameters of the target object in the map. This step helps the machine to accurately know the location of the target object, based on which the system can generate a target bounding box. This bounding box is a rectangular box for marking the approximate location of the target object. The method is helpful for visually representing the target, and performing boundary clipping on the target boundary box to obtain a region containing the target object. This region is typically smaller than the entire image, reducing the computational effort of subsequent segmentation operations, performing pixel-level segmentation on the target bounding box region, dividing each pixel in the image into a target pixel label and a target bounding box background, using deep learning techniques, the contours and shapes of the target object can be identified very accurately, and by mapping the segmentation results back to the original environment map, the system can generate an example segmentation map of the target. This example segmentation map not only includes pixel labels for objects, but also distinguishes between different objects, so that multiple objects can be identified and assigned unique identifiers.

Preferably, step S3 comprises the steps of:

Step S31: acquiring a first position parameter of the cable cutting machine through a sensor;

step S32: calculating a target object path of the first position parameter of the cable cutting machine through the target instance segmentation map to generate a target object path parameter;

step S33: planning a path of the environment map according to the path parameters of the target object so as to generate a first working path;

step S34: performing obstacle marking on the first working path through an environment map to obtain path obstacle data;

Step S35: performing obstacle avoidance strategy analysis on the path obstacle data based on the first working path to generate an obstacle avoidance strategy;

Step S36: and performing obstacle path optimization on the first working path based on the obstacle avoidance strategy to construct a second working path.

According to the invention, through the sensor, the system acquires the first position parameter of the cable cutting machine. This may be positional information of a robotic arm, vehicle, or other type of mobile device, with the target instance segmentation map, the system calculating path parameters for the target object. This may include information of the position, direction, speed, etc. of the target for subsequent path planning, based on target object path parameters, the system performs path planning to generate a first working path of the cutting machine. This path takes into account the position of the target object so that it can work in conjunction with the target object when performing the task, and the system detects and marks obstacles on the first working path through the environment map. This may be other objects, obstacles or non-passable areas. The data are acquired to ensure the safety of the path, and the system analyzes the obstacle avoidance strategy based on the first working path and the path obstacle data. This may include strategies for avoiding obstacles, slowing down, bypassing, etc., to ensure that the machine is avoiding collisions or accidents while performing tasks, and the system optimally adjusts the first work path to construct the second work path based on the obstacle avoidance strategy. The path considers the obstacle avoidance strategy to ensure that the machine can avoid obstacles when executing the task, improve the efficiency and safety of task execution, the system can ensure that the machine cannot collide or accident when executing the task by detecting and avoiding the obstacles on the path, thereby enhancing the safety of operation, the system can generate a path which cooperates with the target object to improve the cooperative efficiency of task execution by considering the position and the path of the target object, the system can adaptively adjust the path according to the change of different situations and obstacles by calculating the path parameters of the target object in real time and analyzing the obstacle avoidance strategy, the adaptability and the robustness of the system are improved, the path optimization and the obstacle avoidance strategy analysis can ensure that the machine executes the task in a more efficient mode, unnecessary pauses and bypasses are reduced, and the execution efficiency of the task is improved.

Preferably, step S35 includes the steps of:

Step S351: performing path crossing identification on the path obstacle data based on the first working path to generate path crossing data;

Step S352: the method comprises the steps of performing motion state analysis on path obstacle data to identify path obstacle types, wherein the path obstacle types are divided into static path obstacles and dynamic path obstacles;

Step S353: when the path obstacle type is a static path obstacle, carrying out detour obstacle avoidance strategy analysis on the static path obstacle based on the first working path so as to generate a detour obstacle avoidance strategy;

step S354: when the path obstacle type is a dynamic path obstacle, performing motion characteristic analysis on the dynamic path obstacle to generate motion characteristic data, wherein the motion characteristic data comprises the speed, the direction and the acceleration of the path obstacle;

step S355: path prediction is carried out on the dynamic path obstacle through the motion characteristic data so as to generate a dynamic obstacle prediction path;

step S356: performing collision probability calculation on the dynamic obstacle predicted path and the first working path through a dynamic obstacle collision probability calculation formula so as to generate obstacle collision probability;

step S357: and carrying out avoidance strategy analysis on the first working path based on the collision probability of the obstacle so as to generate an avoidance strategy.

The invention analyzes path obstacle data through the first working path to identify the path crossing condition. Path crossing refers to the situation where a path obstruction may cross a first working path. Generating path crossing data facilitates subsequent collision probability calculation and avoidance maneuvers, and the system analyzes the path obstacle data to identify the type of path obstacle. These types are classified into two types, a static path obstacle (which does not move) and a dynamic path obstacle (which moves). This classification is to take different obstacle avoidance strategies based on the characteristics of the obstacles, and when static path obstacles are identified, the system analyzes how to bypass the obstacles to generate a bypass obstacle avoidance strategy. This may include selecting an appropriate alternative path or adjusting the path of the machine to avoid static obstacles, and when the path obstacles are identified as dynamic, the system may analyze the motion characteristics of the obstacles, including speed, direction, acceleration, and the like. These data are used for subsequent path prediction and collision probability calculation, based on the motion characteristics data, the system predicts the future path of the dynamic obstacle. This helps to mechanically predict where an obstacle may be present in order to take appropriate obstacle avoidance measures, and calculates the collision probability using a collision probability calculation formula in combination with the predicted path and the first working path of the dynamic obstacle. This helps to evaluate whether the machine will collide with a dynamic obstacle, based on the collision probability, the system generates an evasive obstacle avoidance strategy. This includes adjusting the path of the machine, slowing, stopping, or taking other action to ensure that the machine can safely pass near dynamic obstacles without collision.

Preferably, the dynamic obstacle collision probability calculation formula in step S356 is specifically:

wherein, P _collision is a calculation formula of collision probability of the dynamic obstacle, H is a adjustment factor of collision probability of the dynamic obstacle, t is a crossing time of the dynamic obstacle on the first working path, v ² (t) is an average moving speed of the dynamic obstacle in time t, a is a moving speed attenuation coefficient of the dynamic obstacle, d is a path distance between the dynamic obstacle and the cable cutting machine, H is a moving speed of the cable cutting machine, k is a volume parameter of the dynamic obstacle, t ₁ is an ending time of path movement of the dynamic obstacle, and t ₀ is an ending time of path movement of the dynamic obstacle.

The invention is realized byThe probability adjustment factor H is used for adjusting the collision probability P _collision, so that the decision system is helped to consider the factors for avoiding collision when planning a path, the safety of operation is improved, and the probability adjustment factor H is used for adjusting the collision probability P _collision. By adjusting the factor, the collision probability can be corrected under different scenes to adapt to different situations and requirements, and the crossing time of the dynamic obstacle on the first working path is calculated, namely, when the dynamic obstacle possibly crosses the cable cutting machine path is predicted. This is a key time parameter that helps predict the timing of potential collisions byReflecting the average speed of movement of the dynamic barrier over time. It helps consider the motion of the obstacle to estimate the collision probability more accurately, k representing the motion speed decay coefficient of the dynamic obstacle, for simulating the change in the obstacle's motion speed. The coefficient can consider the situation that the obstacle possibly decelerates or accelerates, so that the collision probability calculation is more accurate, and the path distance d represents the shortest distance between the dynamic obstacle and the cable cutting machine. This distance is an important parameter for calculating the collision probability for estimating whether an obstacle will be close to the machine, and the movement speed v of the cable cutting machine is used to take into account the movement situation of the machine. Taking the machine speed into consideration can estimate the collision probability more accurately, and the formula can better plan the path and action of the machine by estimating the collision probability accurately so as to ensure the safety and the high efficiency of the operation. This helps to reduce the risk of accidents and to improve the reliability of the automated operation.

Preferably, step S4 comprises the steps of:

step S41: carrying out rotational inertia analysis on the cable cutting machine to generate rotational inertia data;

Step S42: carrying out stress load analysis on the cable cutting machine based on the moment of inertia data so as to generate stress load parameters;

step S43: carrying out dynamic characteristic analysis on the stressed load parameters to generate motion constraint data;

Step S44: planning the path speed of the second working path based on the motion constraint data to generate the path speed;

step S45: performing action time sequence segmentation on the cable cutting machine through the path speed to generate action time sequence segmentation data;

step S46: performing action serialization analysis on the second working path based on the action timing sequence segmentation data to generate action serialization data;

step S47: and performing track fitting on the second working path according to the motion serialization data to generate a cable cutting mechanical track.

The invention analyzes the moment of inertia of the cable cutting machine, which is an important parameter about the property of the rotary motion of a mechanical object. The generated moment of inertia data can be used for subsequent stress load and dynamics analysis, and stress load analysis is performed on the cable cutting machine based on the moment of inertia data. This means that the stress situation of the machine under different conditions is determined, including the stress situation of the individual critical components and structures. These parameters are critical to the stability and safety of the machine and dynamic characteristics of the stressed load parameters are analyzed. This includes information about the dynamic response, inertia, acceleration, etc. of the machine. Generating such data facilitates understanding of the movement characteristics of the machine to better plan the path and actions, and path speed planning is performed on the second work path. This means that the speed distribution of the machine on the path is determined to ensure that the motion constraints are met and stability is maintained when the task is performed, the actions of the cable cutting machine are time-series segmented, the whole task is decomposed into a series of small action paragraphs, so that the actions of the machine are better controlled and coordinated, and the action serialization analysis is performed on the second working path based on the action time-series segmentation data. This includes determining when to perform a particular action so that the machine may perform the task in a predetermined sequence, path speed planning and action sequence segmentation help optimize the movement of the machine so that it is more efficient and smooth in performing the task, action sequencing and trajectory fitting ensure that the task performs in an expected manner, improving the accuracy and controllability of the task planning, reducing unnecessary wear of the machine by better understanding the stress situation, and extending its life.

Preferably, step S5 comprises the steps of:

Step S51: detecting the position of the cable cutting machine in real time to obtain a second position parameter of the cable cutting machine;

Step S52: performing target positioning error analysis on the target instance segmentation map to generate target positioning error parameters;

Step S53: carrying out attitude error calculation on the cable cutting machine through an attitude error calculation formula of the cable cutting machine based on the second position parameters of the cable cutting machine so as to generate attitude error data of the cable cutting machine;

step S54: carrying out target attitude error analysis on the target positioning error parameters according to the attitude error data of the cable cutting machine so as to generate target attitude error data;

step S55: performing self-adaptive attitude adjustment analysis through the target attitude error data to generate a self-adaptive attitude adjustment strategy;

According to the invention, the system is allowed to acquire accurate position information of the cable cutting machine through real-time position detection, the second position parameter of the cable cutting machine provides more details about the machine position, possibly including speed, acceleration and other information, basic data are provided for subsequent error analysis and adjustment, target instance segmentation graph analysis allows the system to identify and position a target object, target positioning error parameters provide accurate information of the target position, the system can identify and quantify target positioning errors, a basis is provided for subsequent adjustment, attitude error calculation can estimate attitude deviation of the cable cutting machine based on the machine position parameters, attitude error data provide attitude information of the machine, which is very important for accurate control and adjustment, based on the machine attitude error data, information about the relation between the target position accuracy and the machine attitude can be analyzed, how the machine attitude affects the target positioning, guidance is provided for further adaptive adjustment, an adaptive attitude adjustment strategy is generated according to the target attitude error data, so that the target positioning error can be minimized, the system can be adjusted accurately according to the specific attitude error of the target, and the system is not adaptive to the condition of the system, and the accuracy of the system is ensured to be changed adaptively under the conditions of being not accurate.

Preferably, the calculation formula of the attitude error of the cable cutting machine in step S53 is specifically:

Wherein E _attcalc is the attitude error parameter of the cable cutting machine, delta is the rotational inertia parameter of the cable cutting machine, P is the current position parameter coordinate value of the cable cutting machine, O is the joint torque of the cable cutting machine, E _loc is the target positioning error parameter, alpha is the torsion speed of the cable cutting machine, E _att is the target attitude error parameter, G is the maximum torsion angle of the cable cutting machine, and F is the working vibration frequency of the cable cutting machine.

The invention is realized byThe position deviation and the attitude deviation of the cable cutting machine are calculated, and the position deviation and the attitude deviation can be comprehensively taken into consideration by adding the squares and the squares of the position deviation and the attitude deviation and performing natural logarithmic operation. This helps to evaluate the overall degree of deviation of the cable cutting machine, rather than focusing on only a single aspect of deviation, the natural logarithm operation can convert the combined result of the positional deviation and the attitude deviation into a more easily understood and comparable value. The result can be mapped into a wider numerical range by taking the logarithm, so that the degree of the attitude error of the cable cutting machine can be better expressed, and the method is characterized by/>Calculating the negative index of the product of the vibration frequency and the maximum torsion angle allows to quantify the degree of influence of the vibration frequency on the attitude error, a higher vibration frequency possibly resulting in a larger attitude error, mapping the input value (negative index of the vibration frequency) to an output value between 0 and 1. The modeling of the nonlinear response enables a formula to better reflect the nonlinear influence of the vibration frequency on the attitude error, and the formula comprehensively considers the position deviation, the attitude deviation and the vibration frequency of the cable cutting machine, and obtains a quantized attitude error parameter through standardization and calculation. This helps to evaluate the performance and stability of the cable cutting machine in terms of attitude control and provides a reference for further optimization and improvement.

Preferably, step S6 comprises the steps of:

Step S61: carrying out real-time dynamic track optimization on the cable cutting mechanical track based on the self-adaptive posture adjustment strategy so as to generate a dynamic optimization track;

step S62: performing expansion convolution on the dynamic optimization track to generate a dynamic track curve;

step S63: carrying out pooled sampling on the dynamic track curve to generate a dynamic track network;

step S64: and carrying out data mining modeling on the dynamic track network to construct a cable cutting machine dynamic track model so as to execute positioning control operation.

The invention can move according to the preset path more accurately by optimizing the track in real time, reduce unnecessary movement and adjustment, improve efficiency and accuracy, enhance and highlight the characteristics of the track by expansion convolution, make the dynamic track curve more representative and distinguishable, facilitate subsequent processing and analysis, reduce the complexity of data by pooling sampling, and retain key information, thereby generating a compact dynamic track network, which facilitates further data processing and analysis, saves calculation resources, allows deep patterns and relations to be extracted from the dynamic track network by data mining modeling, provides more accurate positioning control strategy for the cable cutting machine, and ensures that the cable cutting machine performs tasks more stably and accurately.

In this specification, a positioning control system based on an intelligent cable cutting machine is provided, including:

the environment map module is used for acquiring environment data of the cable cutting machine by using the camera; performing environment scanning through a laser radar to obtain obstacle position data; carrying out three-dimensional point cloud modeling on the cable cutting mechanical environment data and the obstacle position data to construct an environment map;

The target boundary module is used for carrying out target positioning on a target object through the environment map so as to generate a target boundary frame; performing pixel level segmentation on the target boundary box to generate a target instance segmentation map;

The path optimization module is used for acquiring first position parameters of the cable cutting machine; planning a path of a first position parameter of the cable cutting machine through a target instance segmentation map so as to generate a first working path; performing obstacle path optimization on the first working path through an environment map to construct a second working path;

The track fitting module is used for carrying out dynamic characteristic analysis on the cable cutting machine so as to generate motion constraint data; planning the path speed of the second working path based on the motion constraint data to generate the path speed; performing action serialization analysis on the second working path through the path speed to generate action serialization data; performing track fitting on the second working path according to the motion serialization data to generate a cable cutting mechanical track;

The gesture adjusting module is used for detecting the position of the cable cutting machine in real time to acquire the second position parameter of the cable cutting machine; carrying out target attitude error analysis on the target instance segmentation map through a second position parameter of the cable cutting machine so as to generate target attitude error data; performing self-adaptive attitude adjustment analysis through the target attitude error data to generate a self-adaptive attitude adjustment strategy;

The track model module is used for carrying out real-time dynamic track optimization on the cable cutting mechanical track based on the self-adaptive posture adjustment strategy so as to generate a dynamic optimization track; and carrying out data mining modeling on the dynamic optimization track to construct a cable cutting machine dynamic track model so as to execute positioning control operation.

According to the invention, the environment data is acquired through the camera and the laser radar, and then three-dimensional point cloud modeling is carried out, so that an accurate environment map is created. This helps the cable cutting machine to understand the surrounding environment, identify the location and shape of obstacles, thereby improving safety and navigation capabilities, and through the environment map, can locate target objects and generate target bounding boxes. The pixel-level segmentation then separates the target object from the surrounding environment, generating a target instance segmentation map. The method is beneficial to accurately identifying and positioning the target of the cable cutting machine, provides important target information for subsequent operation, obtains the position parameters of the cable cutting machine by the path optimization module, and plans the first working path by using the target instance segmentation map. The second working path is generated by obstacle path optimization. The method enables the machine to bypass obstacles more effectively, reduces the moving time and energy consumption, improves the efficiency, analyzes the dynamics characteristics of the cable cutting machine and generates motion constraint data. Path speed planning is then performed based on the data, generating path speeds so that the machine can move at the appropriate speed on the path. Finally, generating the track of the cable cutting machine through action serialization and track fitting. This facilitates a smoother and more controllable movement of the machine on the path, real-time detection of the position of the cable cutting machine, analysis of the target attitude error data, and then generation of an adaptive attitude adjustment strategy. The method can ensure that the machine maintains a correct posture when executing tasks, improves the working quality and precision, optimizes the track of the cable cutting machine in real time based on the self-adaptive posture adjustment strategy, and generates a dynamic optimization track. And constructing a dynamic track model through data mining modeling to execute positioning control operation. The machine can better adapt to different working conditions and requirements, and the reliability and the robustness of automatic operation are improved.

Drawings

FIG. 1 is a schematic flow chart of steps of a positioning control method and system based on an intelligent cable cutting machine;

FIG. 2 is a detailed implementation step flow diagram of step S1;

FIG. 3 is a detailed implementation step flow diagram of step S2;

Fig. 4 is a detailed implementation step flow diagram of step S3.

Detailed Description

It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

The embodiment of the application provides a positioning control method and system based on an intelligent cable cutting machine. The execution main body of the positioning control method and the system based on the intelligent cable cutting machine comprises, but is not limited to, the system: mechanical devices, data processing platforms, cloud server nodes, network uploading devices, etc. may be considered general purpose computing nodes of the present application, including but not limited to: at least one of an audio image management system, an information management system and a cloud data management system.

Referring to fig. 1 to 4, the present invention provides a method comprising the steps of:

In the embodiment of the present invention, as described with reference to fig. 1, a step flow diagram of a positioning control method and a system based on an intelligent cable cutting machine according to the present invention is provided, where in this example, the steps of the positioning control method based on the intelligent cable cutting machine include:

In this embodiment, a camera is mounted at a proper position on the cable cutting machine. The position and the angle of the camera can provide clear environment images, the camera is connected to a control system or a computer, video streams transmitted by the camera can be ensured to be acquired, the control system or the computer starts to acquire image data of the camera, and a laser radar is arranged at a proper position on the cable cutting machine. Lidar is often required to be placed at the high level of the machine in order to be able to cover a large range of environments, start the lidar and start the environment scan. The laser radar emits a laser beam and measures the reflection thereof to acquire position and shape data of the obstacle, transmits laser radar scanning data to a computer, and performs data processing to extract the position data of the obstacle. The data are usually expressed in a point cloud form, each point represents the position of an obstacle, the image data acquired by a camera are registered with the point cloud data acquired by a laser radar, different sensor data are fused into a key of a coordinate system, and three-dimensional point cloud modeling is started by using the registered data. The method comprises the steps of matching objects on a camera image with obstacles in a laser radar point cloud to obtain a more accurate environment map, and starting to construct the environment map according to matched data. The map may be three-dimensional to represent the location, shape, and size of obstacles and objects in the environment. The map may also include other information such as texture and color: along with the movement of the cable cutting machine, the data of the camera and the laser radar are continuously collected, and the environment map is updated in real time, so that the machine can respond to the environment change in time.

In this embodiment, the environment map is combined with the camera image data so that the object detection algorithm is applied to the environment map. Typically, the environment map is used to provide positional information of the target, while the camera image is used to obtain visual information of the target, and the input data is analyzed using a selected target detection algorithm. The algorithm will return positional information for each detected object, typically in the form of a bounding box, converting the positional information detected by the object into an object bounding box. These bounding boxes will be used for pixel level segmentation at the next step, the target bounding box being used as the region of interest (Region of Interest, ROI) for the target instance segmentation. These bounding boxes focus the help algorithm on the target for pixel-level segmentation, using the selected target instance segmentation algorithm, the pixels within each target bounding box are segmented. The algorithm will generate a separate binary mask for each object, with object pixels marked as foreground and background pixels marked as background, if there are multiple objects, combining their instance segmentation results into one image to generate the final object instance segmentation map. Each target may be assigned a unique identifier or color to distinguish them.

In this embodiment, a sensor or a position measurement device on the cable cutting machine is used to obtain a position parameter of the current machine, and the obtained position parameter data is transmitted to a control system or a computer for processing, so as to obtain a first position parameter of the machine. And determining a target object or region which needs to be reached by the cable cutting machine according to the target instance segmentation map. This may be a cable or other object that needs to be trimmed, using the first position parameter of the machine as the coordinates of the starting point, using the position of the target object as the coordinates of the ending point, using a selected path planning algorithm to generate a first working path in the environment map to ensure that the machine can reach the target object or area, taking into account the obstacle information in the environment map in path planning, ensuring that the generated second working path avoids the obstacle, using a path planning algorithm to generate the second working path while optimizing the path to ensure the efficiency and safety of the path, combining the first and second working paths to form the final working path. This final path will take into account target object position, obstacle avoidance, efficiency, etc.

In this embodiment, based on the position, speed and acceleration information of the machine, the kinematic characteristics of each joint of the machine, such as joint position, speed and acceleration, are calculated, and a kinetic model of the machine is used to perform a kinetic analysis, taking into account the effects of external forces and torques to determine the mechanical dynamics. This includes joint torque, inertial force, etc., generating motion constraint data of the machine from the results of the kinetic analysis, the data describing motion limits of the machine when performing the task, such as maximum speed, maximum acceleration, maximum joint torque, etc., generating path speeds based on the second working path and the motion constraint data using a selected algorithm to ensure that the machine adheres to the motion constraints when performing the task, segmenting the second working path into a plurality of time segments, each time segment comprising a continuous path, converting the path speed data into a time-dependent speed sequence to describe speed changes of the machine in each time segment, analyzing the speed sequence in each time segment, identifying critical motion segments, such as acceleration, deceleration, constant speed, etc., converting the identified motion segment information into motion serialization data, including start time, end time, speed change, etc., of each motion segment, generating a series of trace points from the motion serialization data, the points describing the position and orientation of the machine when performing the task, using a selected fitting algorithm, fitting the generated trace points as a smoothed trace, to describe motion trace, such as a fitted trace, and generating motion trace, including the motion trace, acceleration, position, speed, etc. representing the motion trace, and the like.

In this embodiment, the position data of the machine is acquired in real time by the above-mentioned sensor during the movement of the machine. The data can comprise information such as coordinates, postures and speeds of the machine, the acquired position data are subjected to filtering and denoising processing, so that accuracy and stability of the data are improved, second position parameters of the cable cutting machine are extracted from the processed data, and the current posture of the target is estimated by using the target segmentation image and the second position parameters. This may be accomplished by a pose estimation algorithm, such as visual SLAM (Simultaneous Localization AND MAPPING) or a pose estimation neural network, that compares the estimated pose of the target with the ideal pose of the target, calculating the target pose error. The attitude error can comprise information such as rotation angle, translation deviation and the like, calculated target attitude error data are stored as a data set, the data set is used for subsequent self-adaptive attitude adjustment analysis, and the influence of different attitude errors on a cable cutting task is analyzed by using the target attitude error data set. This may include analysis in terms of error magnitude, direction, frequency, etc., and based on the results of the target attitude error analysis, an adaptive attitude adjustment strategy is generated. The strategy can comprise the steps of adjusting parameters such as the posture, the speed and the acceleration of the machine to minimize a target posture error, implementing the generated self-adaptive posture adjustment strategy on the cable cutting machine, ensuring that the machine is adjusted according to the real-time target posture error when executing a task, continuously monitoring the position and the target posture error of the machine during the execution of the cable cutting task, and adjusting the movement of the machine according to the requirement so as to ensure the accuracy and the stability of the task.

In this embodiment, parameters such as the gesture and the speed of the cable cutting machine are adjusted according to the data acquired in real time through the adaptive gesture adjustment strategy to minimize the target gesture error, and dynamic track optimization algorithms suitable for the cable cutting machine are selected, wherein the algorithms can adjust the motion track of the machine according to the real-time data and the adaptive gesture adjustment strategy to adapt to different working scenes and targets, and the motion track of the cable cutting machine is optimized in real time according to the selected algorithms to ensure that the machine remains stable and efficient when executing tasks: storing the optimized track as a dynamic optimized track, selecting a proper data mining model, such as a machine learning model (such as regression, classification, clustering and the like) or a deep learning model (such as a neural network) to model dynamic optimized track data, adjusting the model according to an evaluation result, and possibly adjusting feature selection, model parameters and the like, and performing positioning control operation on the cable cutting machine by using the constructed dynamic track model and the data acquired in real time during the execution of the cable cutting task. The method comprises the steps of adjusting the movement of the machine according to the current position and the target position to keep on a dynamic optimization track, continuously monitoring the position and the gesture of the machine, and adjusting the movement of the machine in real time according to the guidance provided by a dynamic track model so as to ensure the smooth progress of a task.

In this embodiment, as described with reference to fig. 2, a detailed implementation step flow diagram of the step S1 is described, and in this embodiment, the detailed implementation step of the step S1 includes:

In this embodiment, the camera is started to continuously capture the environmental image data. The captured images may be pre-processed, including denoising, color correction, distortion correction, etc., using a single or multiple cameras, depending on the needs of the task, to improve data quality, ensure that the lidar has been properly mounted on the cable cutter, and calibrated to ensure accurate laser scanning, start the lidar, perform environmental scanning, and acquire obstacle location data. The laser radar can provide high-precision distance and position information, post-process laser scanning data, including removing outliers, coordinate transformation and the like, so as to improve data accuracy, fuse data from the camera and the laser radar, align the data under the same coordinate system, and convert the fused data into three-dimensional point cloud data. For camera data, depth information can be generated by using parallax methods, structured light and other methods to construct point clouds, the environmental point cloud data is segmented into different points cloud mass by using a point cloud segmentation algorithm, each point cloud mass represents an environmental feature or object, the obstacle point cloud data is segmented into different point cloud clusters by using a point cloud segmentation algorithm, each point cloud cluster represents an obstacle or obstacle group, and three-dimensional point cloud modeling is performed on each environmental point cloud mass and obstacle point cloud cluster. This may include identifying features in the point cloud using machine learning algorithms or geometric modeling methods and converting them into map data, combining the modeling results of all the environmental points cloud mass and obstacle point cloud clusters to construct a complete environmental map. The map can be used for navigation, obstacle avoidance, task planning and other applications.

In this embodiment, as described with reference to fig. 3, a detailed implementation step flow diagram of the step S2 is shown, and in this embodiment, the detailed implementation step of the step S2 includes:

In this embodiment, a target detection algorithm is used to identify possible target objects in the environment map. This may be the use of a deep learning model (such as a convolutional neural network) for target detection or the localization of a target by matching specific features, once the target object is detected, its positional parameters, such as three-dimensional coordinates or positional parameters relative to the machine, are obtained. This will help determine the location of the object in the environment, and based on the location parameters of the object, a bounding box is generated on the environment map to mark the location of the object. This bounding box is typically a rectangle and the environment map is cropped using the target bounding box, leaving only the data for the target area. This will reduce the amount of data processed, increase the efficiency of segmentation, and use a pixel-level segmentation algorithm (such as semantic segmentation or instance segmentation) to divide pixels in the target bounding box region into target pixels and background pixels. This may be achieved by a deep learning model that assigns a label to each pixel indicating whether it belongs to the target object, generates a segmented image based on the results of the pixel level segmentation, wherein the target pixel is labeled as the target object and the background pixel is labeled as the background, and maps the pixel level segmentation results back to the original environment map to generate an example segmentation map of the target. In this figure, each object has a unique identifier to track and identify the different objects, and finally an object instance segmentation map is output in which each object is segmented and marked. This map may be used for robotic navigation, target tracking, obstacle avoidance, and other tasks to aid the mechanical system in interacting with the surrounding environment.

In this embodiment, as described with reference to fig. 4, a detailed implementation step flow diagram of the step S3 is shown, and in this embodiment, the detailed implementation step of the step S3 includes:

In this embodiment, a suitable sensor (e.g., a laser sensor or a camera) is used to obtain the position parameters of the cable cutting machine. The sensors can provide information such as the position, the direction, the gesture and the like of the machine, and calculate and generate target object path parameters to be followed by the cable cutting machine based on the position parameters of the cable cutting machine and the position information of the target object. These path parameters may include position, attitude, velocity, acceleration, etc. of the target object, and the path planning algorithm is used to generate a first working path of the cutting machine based on the environment map and the target object path parameters. Path planning takes into account the physical constraints of the machine and the safety of the path, ensuring that the machine can move correctly along the path, marking obstacles on the first working path on the environment map. These obstacles may be other objects, walls, terrain variations, etc. that are marked to facilitate subsequent obstacle avoidance strategy analysis and path optimization, and obstacle data on the first working path, including location, shape, and other relevant information, is extracted from the environment map, and the obstacle avoidance strategy is determined by analyzing the path obstacle data. This may be to avoid obstacles, adjust paths, change speeds, or take other action to ensure that the machine is safe and reliable in executing the work paths, the first work path being optimized according to an obstacle avoidance strategy to generate the second work path. The goal of optimization may be to minimize path length, maximize safe distance, or meet specific operational requirements.

In this embodiment, step S35 includes the steps of:

In this embodiment, the first working path and the path obstacle data are used to identify and determine whether a path crossing condition exists. The path crossing refers to the condition that the paths and the path obstacles are crossed, overlapped or collided, path crossing data including crossing positions, crossing types and other related information are generated for the recognized path crossing condition, and motion state analysis is carried out for the path obstacle data to judge the types of the path obstacles. The path obstacle is classified into a static path obstacle, such as a wall, a fixed obstacle, and a dynamic path obstacle, such as other work machines, for which an analysis is performed based on the first work path, determining a detour strategy. This may include selecting an appropriate detour path, adjusting a path planning algorithm, or utilizing special functions of the machine to avoid static path obstructions, performing motion characteristic analysis for dynamic path obstructions, determining motion characteristic data such as velocity, direction, and acceleration thereof. This may be derived by analyzing historical trajectories, sensor data or other information of the path obstructions, and using the motion characteristics data of the dynamic path obstructions, performing path prediction to predict their future likely paths. This may be achieved by a method such as a motion model, a machine learning algorithm, or the like, and the collision probability on the path is calculated by a collision probability calculation formula using the dynamic obstacle prediction path and the first working path. This may estimate the probability of collision occurring based on the contact area, relative velocity, and other relevant factors of the path, and perform an avoidance maneuver analysis on the first working path based on the obstacle collision probability. According to the collision probability, the method can select to adjust the path, reduce the speed and take avoidance action or other obstacle avoidance measures to ensure the safety.

In this embodiment, the dynamic obstacle collision probability calculation formula in step S356 is specifically:

In this embodiment, step S4 includes the following steps:

In this embodiment, the moment of inertia of each component of the cable cutting machine needs to be measured and calculated. Moment of inertia is a physical quantity describing the inertial properties of an object as it rotates about an axis, and is related to the shape, mass distribution, and position of the axis of the object. By measuring and calculating the rotational inertia of each component, the rotational inertia data of the cable cutting machine can be obtained. And analyzing the stress load of the cable cutting machine in the working state. By considering the mass and the moment of inertia of each component of the cable cutting machine and the external force and moment applied to the cable cutting machine during the working process, the parameters of the stressed load, such as the magnitude and the direction of the force and the moment, can be calculated. Analysis of kinetic properties was performed. By considering the mass, the moment of inertia, the stressed load parameters and the motion constraint conditions of the cable cutting machine, the dynamic characteristics of acceleration, speed, displacement and the like of the cable cutting machine in the motion process can be calculated, so that motion constraint data are generated. And planning the path speed of the second working path. Path speed planning refers to determining the speed of a mechanical system at different locations on a given path to meet motion constraints and operational requirements. By path speed planning, speed data at various locations on the second working path may be generated. And performing time sequence segmentation on the action of the cable cutting machine according to the obtained path speed data. Action sequence segmentation refers to the division of the entire course of action into several small time periods, in each of which the speed of the mechanical system remains substantially constant. By the action time sequence segmentation, information such as the position and the speed of the mechanical system in each time period can be determined, so that action time sequence segmentation data is generated. And performing action serialization analysis on the second working path. Motion serialization refers to converting information such as the position and speed of a mechanical system in different time periods into a specific motion sequence so as to realize the motion of the cable cutting machine on a second working path. Through the action serialization analysis, data describing the sequence of mechanical actions of the cable cutting machine may be generated. And performing track fitting on the second working path. Track fitting refers to determining the actual track of the cable cutting machine on the second working path according to a given action sequence. Through track fitting, the change rule of the position and the speed of the cable cutting machine on the second working path can be obtained, so that track data of the cable cutting machine are generated.

In this embodiment, step S5 includes the following steps:

In this embodiment, the position of the cable cutting machine is detected in real time using a suitable sensor (e.g., laser rangefinder, camera, encoder, etc.). These sensors will collect data about the machine position, including coordinate information and time stamps, from which the current position parameters of the cable cutting machine on the second working path, such as position coordinates (x, y, z) and direction (attitude) parameters (as euler angle or quaternion representation) are calculated. The object instance is segmented using computer vision techniques, separated from the background to obtain a contour or mask of the object, the actual position of the object is compared with the position in the segmented image, and the positioning error of the object, including position error and orientation error, is calculated. These error parameters reflect the difference between the position of the target in the image and the actual position, and the actual pose of the cable cutting machine is compared with the desired pose using a pose error calculation formula to calculate the pose error. The attitude error is typically expressed in terms of euler angles or quaternions, and attitude error data is generated, including direction errors and angle errors, to describe the difference between the current attitude and the desired attitude of the cable cutting machine, and the relationship between the attitude error data and the target positioning error parameters of the cable cutting machine is analyzed in combination. This may involve the use of mathematical models or relational functions to calculate the target attitude error, i.e. the effect of the attitude error of the cable cutting machine on the target positioning. This may be a directional error, an angular error, or other relevant parameter, generating target attitude error data to describe the impact of the attitude error on target positioning, analyzing the target attitude error data, knowing the extent of impact of different attitude errors on task execution. The method can comprise the steps of developing a self-adaptive attitude adjustment strategy according to the influences of deviation, stability and the like of target positions under different attitude errors, automatically adjusting the attitude of the cable cutting machine according to the size and the direction of the target attitude errors, minimizing the errors and ensuring successful completion of tasks, implementing the self-adaptive attitude adjustment strategy, and adjusting the attitude of the cable cutting machine by controlling an actuator of the cable cutting machine so as to adapt to different target attitude error conditions.

In this embodiment, the calculation formula of the attitude error of the cable cutting machine in step S53 is specifically:

In this embodiment, step S6 includes the following steps:

In this embodiment, the track of the cable cutting machine is optimized in real time according to the current position parameter and the dynamic gesture adjustment strategy. This may be achieved by mathematical models, control algorithms or optimization methods. The optimized target is to make the mechanical track more suitable for the attitude error and the positioning error of the target so as to improve the accuracy of positioning control, and a dynamic optimized track is obtained according to the optimized result, which is a suggested motion path of the cable cutting machine under the current attitude adjustment strategy, the dynamic optimized track is converted into a curve representation form, a smooth track curve can be obtained through an interpolation or fitting method, and the expansion convolution operation is carried out on the track curve. Dilation convolution is an operation of expanding a curve or trajectory to a certain width, typically implemented using a filter or convolution kernel. The method can enable the track curve to be more stable, and the dynamic track curve is generated by taking the motion constraint and the working environment condition of the machine into consideration, and the dynamic track curve is subjected to expansion convolution treatment, wherein the dynamic track curve is subjected to pooling operation by taking the posture adjustment of the machine and the smooth track representation of the environment condition into consideration, and the track curve is discretized into a series of sampling points. The pooling mode can be equal interval sampling or adaptive sampling according to the needs, the number and the interval of sampling points are determined according to specific requirements and system requirements, and in general, the factors such as the movement range of the machine, the track precision requirement, the data processing efficiency and the like are considered to generate a dynamic track network, and the dynamic track network is formed by pooled sampling points, wherein each sampling point comprises relevant information of the machine at a corresponding position, such as position coordinates, attitude parameters and the like. The network is used for subsequent data mining modeling, and data mining and modeling analysis are performed by using data in the dynamic track network. This may include techniques such as statistical analysis, machine learning, deep learning, etc. to extract useful patterns, rules, and associations from the data to construct a dynamic trajectory model of the machine that will describe the trajectory characteristics and performance behavior of the machine under different attitude adjustment strategies. The model can be a mathematical model, a probability model, a neural network model and the like, and can be applied to positioning control operation after the dynamic track model is completed by selecting an appropriate model type according to the needs of specific application. The model predicts the optimal track of the cable cutting machine based on a real-time attitude adjustment strategy and environmental conditions so as to realize more accurate positioning control.

In this embodiment, a positioning control system based on an intelligent cable cutting machine is provided, including:

The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

It will be understood that, although the terms "first," "second," etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.

The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. The positioning control method based on the intelligent cable cutting machine is characterized by comprising the following steps of:

2. The positioning control method based on the intelligent cable cutting machine according to claim 1, wherein the specific steps of the step S1 are as follows:

3. The positioning control method based on the intelligent cable cutting machine according to claim 1, wherein the specific steps of the step S2 are as follows:

4. The positioning control method based on the intelligent cable cutting machine according to claim 1, wherein the specific steps of the step S3 are as follows:

5. The positioning control method based on intelligent cable cutting machinery according to claim 4, wherein the obstacle avoidance policy analysis includes a detour policy analysis and an avoidance policy analysis, the obstacle avoidance policy includes a detour obstacle avoidance policy and an avoidance obstacle avoidance policy, and the specific steps of step S35 are:

6. The positioning control method based on intelligent cable cutting machine according to claim 5, wherein the dynamic obstacle collision probability calculation formula in step S356 specifically includes:

；

Wherein, For dynamic obstacle collision probability calculation formula,/>For dynamic obstacle collision probability adjustment factor,/>For the crossing time of dynamic obstacles on the first working path,/>For time/>The average speed of movement of the inner dynamic barrier,Is the motion speed attenuation coefficient of dynamic obstacle,/>For the path distance between the dynamic obstacle and the cable cutting machine,/>For the movement speed of the cable cutting machine,/>As a volume size parameter of dynamic obstacle,/>For the end time of the dynamic obstacle path motion,/>Is the starting time of the dynamic obstacle path movement.

7. The positioning control method based on the intelligent cable cutting machine according to claim 1, wherein the specific steps of the step S4 are as follows:

8. The positioning control method based on the intelligent cable cutting machine according to claim 1, wherein the specific steps of the step S5 are as follows:

the calculation formula of the attitude error of the cable cutting machine in step S53 specifically includes:

；

Wherein, For the attitude error parameter of cable cutting machinery,/>For the rotational inertia parameter of the cable cutting machine,/>For the current position parameter coordinate value of the cable cutting machine,/>For cutting cable mechanical joint torque,/>For the target positioning error parameter,/>For cutting cable mechanical torsion speed,/>For the target attitude error parameter,/>For the maximum torsion angle of the cable cutting machine,/>The vibration frequency of the cable cutting machine is the working vibration frequency of the cable cutting machine.

9. The positioning control method based on the intelligent cable cutting machine according to claim 1, wherein the specific steps of the step S6 are as follows:

10. A positioning control system based on an intelligent cable cutting machine, for executing the positioning control method based on the intelligent cable cutting machine as set forth in claim 1, comprising: