Interactive data processing method for intelligent robot system
Through electric field detection, three-dimensional modeling and real-time environment perception, path planning and robotic arm control are optimized, and the problem of slow semantic understanding and response speed of intelligent robot systems in complex environments is solved, achieving more efficient and safe task execution.
Patent Information
- Application Number
- CN202411051233.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-01
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing intelligent robot systems have problems with inaccurate semantic understanding, slow response speed and lack of personalized services when processing complex interactive data, especially in natural language understanding, multimodal perception, and dynamic environment adaptability.
By obtaining robot sensing data for electric field detection and digitization, establishing a three-dimensional structural model, real-time environmental perception, object detection and obstacle pre-cleaning, optimizing path planning and robotic arm control strategies, integrating data for robot control parameters, ensuring that in Execute tasks efficiently and stably in complex dynamic environments.
It improves the robot's responsiveness, navigation accuracy, security and operation efficiency in dynamic environments, enhances the flexibility and intelligent perception of complex scenarios, and ensures efficient and stable execution of mobile and operation tasks.
Smart Images

Figure CN118968469A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of human-computer interaction technology, and in particular to an interactive data processing method for an intelligent robot system. Background Art
[0002] With the rapid development of intelligent robot technology, its application in various fields has gradually expanded, including but not limited to home services, industrial production, healthcare, and customer service. Intelligent robot systems are able to obtain environmental information through sensors and perform tasks through data processing and decision-making algorithms. However, existing robot systems still face challenges in processing complex interactive data, such as limitations in natural language understanding, multimodal perception, and adaptability to dynamic environments. Traditional robot interaction methods are usually based on predefined instruction sets or fixed operating procedures, which limits the robot's ability to respond to complex and dynamic environments. In addition, when processing user natural language input, existing technologies often have problems such as inaccurate semantic understanding, slow response speed, and lack of personalized services. Summary of the invention
[0003] Based on this, it is necessary for the present invention to provide an interactive data processing method for an intelligent robot system to solve at least one of the above technical problems.
[0004] To achieve the above object, an interactive data processing method for an intelligent robot system includes the following steps:
[0005] Step S1: acquiring robot sensor data, and performing electric field detection on the robot sensor data, thereby obtaining electric signal change data; and digitally processing the electric signal change data, thereby obtaining a calibration electric signal instruction;
[0006] Step S2: Acquire the robot application area structural data, perform three-dimensional building structure modeling according to the robot application area structural data, so as to obtain the robot application area structural model; perform radar sensing spatial position recognition on the robot application area structural model according to the robot sensing data, so as to obtain the robot real-time position data; perform path planning on the robot real-time position data according to the calibration electrical signal instruction, so as to obtain the initial path data, and upload it to the robot management cloud platform to execute the robot movement control task;
[0007] Step S3: Performing real-time environmental perception according to the robot sensing data, thereby obtaining real-time robot environmental data; acquiring robot camera data, and performing target detection on the robot camera data according to the calibration electrical signal instruction, thereby obtaining a camera target object image; classifying the real-time robot environmental data according to the camera target object image, thereby obtaining obstacle data and pre-cleared obstacle data;
[0008] Step S4: performing obstacle avoidance optimization on the initial path data according to the obstacle data, thereby obtaining obstacle avoidance optimized path data; acquiring the robot arm data, and performing posture control strategy analysis on the robot arm data according to the robot sensor data, thereby obtaining the robot arm control data;
[0009] Step S5: Integrate the robot control parameters of the obstacle avoidance optimization path data and the robot arm control data to obtain the robot optimization control data, and upload it to the robot management cloud platform to perform the robot movement correction task.
[0010] The present invention obtains calibration electric signal instructions through electric field detection and digital processing, ensures that the robot can accurately obtain the electric signal change data of the surrounding environment, thereby identifying potential electric field interference or changes, and providing an accurate starting point for subsequent position identification and path planning. A three-dimensional structural model of the robot application area is established, and the real-time position is identified by radar sensing, and then the path is planned according to the calibration electric signal instructions, so as to ensure that the robot can accurately understand its own position in a complex structural environment, and dynamically adjust the path according to the real-time situation, perform mobile tasks in an optimized manner, and enhance the robot's responsiveness and navigation accuracy in a dynamic environment. Real-time robot environment data is obtained, and target detection and classification are performed through camera data, allowing the robot to perceive obstacles and their classification in the environment in real time, predict and clear obstacles in advance, thereby avoiding collisions or errors, and improving the safety and operation accuracy of the robot in a complex dynamic environment. The path is optimized according to the obstacle data, and the robot arm control strategy is analyzed at the same time to ensure that the robot can avoid the identified obstacles, avoid unnecessary detours or pauses by optimizing the path, and accurately analyze the posture control of the robot arm, thereby improving the flexibility and operation efficiency of the robot in complex scenes. Integrate obstacle avoidance optimization path and robotic arm control data, upload them to the robot management cloud platform to perform mobile correction tasks, ensure that the robot can make dynamic adjustments based on real-time environmental data and optimized paths during actual execution, and ensure that the robot's movement and operation tasks can be performed efficiently and stably in a complex and dynamic environment.
[0011] Optionally, step S1 specifically includes:
[0012] Step S11: acquiring robot sensor data, and performing capacitive touch sensor data extraction on the robot sensor data, thereby obtaining capacitive touch sensor data;
[0013] Step S12: Drawing an electric field distribution diagram of the capacitive touch sensing data, thereby obtaining capacitive touch electric field distribution diagram data;
[0014] Step S13: performing low-frequency electric field screening according to the capacitive touch electric field distribution map data, thereby obtaining electric signal change data;
[0015] Step S14: digitally process the electrical signal change data to obtain a calibration electrical signal instruction.
[0016] The present invention reads data from the sensor, identifies and extracts relevant information of the touch signal, and through the capacitive touch sensing data, can monitor in real time whether the robot is touched or operated, such as detecting the interaction between the user and the robot or starting a specific function. The electric field distribution on the surface of the robot is plotted based on the sensor data. The electric field distribution map can help determine the electric field strength and distribution in different areas, and provide basic data for subsequent electrical signal analysis. According to the data in the electric field distribution map, the changes in the low-frequency electric field are extracted and analyzed. Low-frequency electric field screening helps to identify and capture the weak electric field signals generated when the human body approaches or contacts the robot, such as for automatically triggering specific reactions or safety measures. The screened and analyzed data is converted into an instruction format that can be understood and executed by the robot system, and the calibration electric signal instruction is used to control the specific reaction or behavior of the robot, such as responding to the user's touch action or adjusting the operating state of the robot.
[0017] Optionally, step S13 is specifically:
[0018] Step S131: classifying the electric field distribution map data by electric field strength, thereby obtaining strong electric field distribution map data and weak electric field distribution map data;
[0019] Step S132: performing electric field frequency statistics on the strong electric field distribution map data, thereby obtaining low-frequency strong electric field distribution map data;
[0020] Step S133: performing electric field frequency statistics on the weak electric field distribution map data, thereby obtaining low-frequency weak electric field distribution map data;
[0021] Step S134: screening the electric field distribution map data for low-frequency strong electric field distribution map data and low-frequency weak electric field distribution map data, thereby obtaining electric signal change data.
[0022] The present invention divides the data into two categories: strong electric field and weak electric field by the electric field strength of each area in the electric field distribution map. The classified data helps to further analyze and understand the electric field environment around the robot, determine which areas are subject to stronger interference or have a higher probability of triggering, and thus improve the accuracy of recognition and response. The distribution of low-frequency electric field signals in the strong electric field area is analyzed. The frequency-intensity electric field distribution map data reflects the frequent changes in electric field signals, such as common user touch actions or other interactive signals, which helps to establish a precise interaction and response mechanism between the robot and the user. The distribution of low-frequency electric field signals in the weak electric field area is analyzed. The weak electric field usually indicates less common or weak electric field changes, which are used to detect the static state of the surrounding environment or specific non-triggering signals, such as environmental monitoring or state feedback. For the low-frequency electric field distribution map data, the electric field signal changes with specific frequency and intensity are screened out, and the obtained electric signal change data can be directly used to generate calibration electric signal instructions for controlling the behavior of the robot or responding to the user's operation. These data reflect important trigger signals in the robot's surrounding environment, which helps to improve the robot's intelligent perception and interaction capabilities.
[0023] Optionally, step S14 is specifically:
[0024] Step S141: performing signal enhancement on the electrical signal change data to obtain enhanced electrical signal change data;
[0025] Step S142: performing bandpass filtering on the enhanced electrical signal change data to obtain corrected enhanced electrical signal change data;
[0026] Step S143: digitally quantizing the correction enhanced electrical signal change data to obtain a calibration electrical signal instruction.
[0027] The present invention can improve the signal quality by amplifying the amplitude of the signal or increasing the signal-to-noise ratio through signal enhancement. This step can make the electrical signal change data clearer and easier to analyze, reduce errors and interference in the subsequent processing stage, and improve the reliability and accuracy of the data. Bandpass filtering can selectively retain signals within a specific frequency range while suppressing unnecessary frequency components. This can further optimize the spectral characteristics of the signal so that the corrected enhanced signal is more in line with the expected standards or requirements. Digital quantization converts the corrected enhanced signal into a digital form, making signal processing more accurate and controllable. Digital signals can be easily stored, transmitted and analyzed, and can reduce the distortion and noise effects of analog signals during transmission and processing, ensuring the integrity and stability of the data.
[0028] Optionally, step S2 specifically includes:
[0029] Step S21: obtaining robot application area structure data, and performing regional geometric structure feature extraction and identification area spatial feature extraction on the robot application area structure data, thereby obtaining application area geometric structure data and identification area spatial data;
[0030] Step S22: performing three-dimensional building structure modeling according to the application area geometric structure data and the identification area spatial data, thereby obtaining a robot application area structure model;
[0031] Step S23: extracting laser radar sensor data from the robot sensor data to obtain laser radar sensor data, and performing radar sensor spatial position recognition on the laser radar sensor data according to the robot application area structure model to obtain geographic coordinate data and speed direction data;
[0032] Step S24: performing position calculation based on the geographic coordinate data and the speed direction data, thereby obtaining the real-time position data of the robot;
[0033] Step S25: dividing the robot real-time position data into application ranges, thereby obtaining the robot real-time position application range data;
[0034] Step S26: Perform path planning on the robot's real-time position application range data according to the calibration electrical signal instruction, thereby obtaining initial path data, and uploading it to the robot management cloud platform to execute the robot movement control task.
[0035] The present invention can obtain the geometric structure data and spatial features of the robot application area through regional geometric structure feature extraction and identification area spatial feature extraction. These data provide a basis for subsequent modeling and path planning, ensuring the accurate perception and operation ability of the robot in the environment. Based on the regional structure data, three-dimensional modeling can accurately simulate the physical environment of the robot application area, which can provide accurate spatial reference for the robot's path planning, obstacle avoidance and task execution, and improve work efficiency and safety. By extracting and processing laser radar sensor data and combining the robot application area structure model, the precise position of the robot in space can be identified. These data not only include geographic coordinates and speed direction information, but also help the robot perceive changes in the surrounding environment in real time to ensure safe navigation and task execution. According to the laser radar sensor data and structure model, position calculation can accurately determine the real-time position of the robot, which is the basis for executing path planning and mobile control tasks, ensuring that the robot can move safely and efficiently according to the predetermined path. Dividing the application range can enable the robot to automatically select an appropriate operation area in the working area according to the current task requirements, which can optimize the task execution order and efficiency, while avoiding unnecessary energy and time consumption. Path planning based on calibrated electrical signal instructions and real-time position data can ensure that the robot moves safely and efficiently along the predetermined path. These path data are uploaded to the robot management cloud platform, allowing operators to monitor and adjust task progress in real time, thereby achieving more intelligent and automated robot control tasks.
[0036] Optionally, step S26 is specifically:
[0037] Step S261: performing path planning on the reasonable range data of the real-time position of the robot according to the calibration electrical signal instruction, thereby obtaining planned path data;
[0038] Step S262: extracting path wind resistance characteristics and path terrain characteristics from the planned path data, thereby obtaining path wind resistance data and path terrain data;
[0039] Step S263: calculating the drag coefficient of the path wind resistance data, thereby obtaining path low wind resistance data;
[0040] Step S264: Calculate the ruggedness of the path terrain data to obtain path flatness data;
[0041] Step S265: performing an intersection operation based on the low wind resistance data of the path and the flat path data, thereby obtaining low risk path data;
[0042] Step S266: Perform low-risk path screening on the path data according to the low-risk path data to obtain initial path data, and upload it to the robot management cloud platform to execute the robot movement control task.
[0043] The present invention determines the reasonable range within which the robot can move according to the current real-time position of the robot and the electrical signal instruction, and the path planning ensures that the robot can move within the predetermined area without exceeding its operating range. This precise position planning helps to ensure that the robot will not deviate from the predetermined path during the task execution process, thereby improving the reliability and efficiency of task execution. By extracting the wind resistance and terrain characteristics of the planned path data, the path wind resistance data reflects factors such as wind speed and wind direction that affect the movement of the robot on the path, while the path terrain data reflects terrain information such as slope and obstacles on the path. The extraction of these data helps the system better understand the physical environment of the path and provides an accurate reference for subsequent steps. The wind resistance coefficient reflects the degree of wind resistance on the path. A low wind resistance path means that the robot is subject to less wind resistance when moving on this path. Selecting a low wind resistance path can reduce the energy consumption of the robot, extend its working time, and improve the operating efficiency of the system. By calculating the ruggedness of the path terrain data, the flatness data of the path is obtained. The path flatness data reflects the flatness or slope changes on the path. Choosing a flat path can reduce the physical challenges faced by the robot during the movement, such as slopes and uneven terrain, and help improve the stability and movement efficiency of the robot. The low wind resistance data of the path and the flatness data of the path are combined to perform intersection operations. The low-risk path data represents the path segments with both low wind resistance and flat characteristics. These path segments are considered to be the best choice for robot movement. Choosing a low-risk path can minimize the energy consumption and risk during the robot's movement and improve the robot's movement efficiency and safety. Based on the low-risk path data, the initial path data is formed. These path data are strictly screened and calculated to ensure that the robot follows the best path selection during the movement to ensure the smooth execution of the task. These path data are uploaded to the robot management cloud platform to achieve effective management and execution of the robot's mobile control tasks.
[0044] Optionally, step S3 specifically includes:
[0045] Step S31: Perform real-time environmental perception based on the laser radar sensing data to obtain real-time robot environmental data;
[0046] Step S32: performing target detection on the laser radar sensing data according to the calibration electrical signal instruction, thereby obtaining the laser radar sensing target data;
[0047] Step S33: acquiring a robot camera image, and performing target detection on the robot camera image according to the calibration electrical signal instruction, thereby obtaining a camera target object image;
[0048] Step S34: integrating the laser radar sensing target data and the camera target object image to obtain the initial confirmed target object data;
[0049] Step S35: obtaining robot knowledge database data, and performing parameter matching on the initially confirmed target object data according to the robot knowledge database data, thereby obtaining pre-cleared obstacle data;
[0050] Step S36: Pre-cleaning obstacle data is screened out from the real-time robot environment data according to the pre-cleaning obstacle data, thereby obtaining obstacle data.
[0051] The present invention obtains real-time environmental data through a laser radar sensor, including information such as the location, size and shape of obstacles. These real-time data are crucial for the robot to navigate safely and avoid obstacles, and can help the robot perceive the surrounding environment in real time and make corresponding action plans. The laser radar sensing data is processed by calibrating electrical signal instructions to identify target objects in the environment, such as obstacles, other robots or specific marked objects. These target data provide basic information for the robot's navigation and path planning, helping the robot avoid obstacles or perform specific tasks. By acquiring camera images, using calibrated electrical signal instructions to perform target detection on the image, and identifying target objects within the camera's field of view, the camera can provide more detailed visual information to help the robot more accurately understand the objects and their positions in the environment, which is particularly important in tasks that require high-precision visual information. By integrating and fusing the target data from the laser radar and the camera, the accuracy of target object recognition can be improved. The target object data initially confirmed includes the target object information jointly confirmed by the laser radar and the camera, providing a basis for subsequent data processing and decision-making. According to the data in the robot knowledge database, detailed parameter matching and identification are performed on the initially confirmed target object data, so that the characteristics and behaviors of the target object can be further confirmed, and its impact on the robot's movement can be predicted, such as the formulation of avoidance strategies, etc. According to the previously obtained pre-cleared obstacle data, the real-time robot environment data is further screened and identified for obstacles. The robot can accurately identify obstacles in the environment, and perform corresponding path planning and action adjustments to ensure safe avoidance of obstacles or interaction with them.
[0052] Optionally, step S33 is specifically:
[0053] Step S331: extracting target object features from the calibration electrical signal instruction, thereby obtaining target object data;
[0054] Step S332: acquiring a robot camera image, and performing grayscale conversion on the robot camera image, thereby obtaining a robot camera grayscale image;
[0055] Step S333: performing background subtraction on the robot camera grayscale image to obtain the robot camera object grayscale image;
[0056] Step S334: performing parameter matching on the robot camera object grayscale image according to the target object data, thereby obtaining the camera target object image.
[0057] The present invention can accurately extract the features of the target object by calibrating the electrical signal instructions. These features include information such as the size, shape, and color of the object. This information is crucial to subsequent image processing and analysis. Accurate feature extraction can help the robot identify and locate the target object. Converting a color image to a grayscale image can simplify the complexity of subsequent image processing while retaining enough image information for subsequent analysis. The grayscale image is insensitive to changes in illumination and can better highlight the shape and texture of the object, which is conducive to the identification and analysis of the target object. By removing irrelevant background parts in the image, the detection and recognition accuracy of the target object can be improved. Especially in complex backgrounds, background subtraction can effectively reduce noise interference, making subsequent processing more accurate and reliable. According to the target object data extracted in the previous step, the image after background subtraction is parameter matched, and the target object can be further accurately located and extracted. These parameters are related to the shape, size, boundary and other features of the target object. The matching process can ensure that the identified target object meets the expected characteristics.
[0058] Optionally, step S4 is specifically:
[0059] Step S41: extracting obstacle path features from the initial path data according to the obstacle data, thereby obtaining initial obstacle path data;
[0060] Step S42: performing environmental perception on the initial obstacle path data according to the laser radar sensing data, thereby obtaining initial obstacle path position data;
[0061] Step S43: performing obstacle avoidance simulation on the initial obstacle path position data according to the real-time position data of the robot, thereby obtaining obstacle avoidance simulation data of the robot;
[0062] Step S44: Calculating the minimum detour angle of the robot's detour obstacle avoidance simulation data, thereby obtaining the robot's minimum detour obstacle avoidance angle data;
[0063] Step S45: Optimizing the initial obstacle path data by the minimum detour angle according to the robot's minimum detour angle data, thereby obtaining obstacle avoidance optimization path data;
[0064] Step S46: Acquire the robot arm data, and perform posture control strategy analysis on the robot arm data according to the laser radar sensor data, so as to obtain the robot arm control data.
[0065] The present invention extracts path features according to obstacle data. These features include information such as the shape, size, and position of the obstacle. This is very critical for subsequent path planning, because understanding the characteristics of the obstacle can help the robot plan the obstacle avoidance path more accurately and avoid collisions with obstacles, thereby ensuring safety and effectiveness. Using devices such as laser radar sensors, the system performs environmental perception on the initial obstacle path data in this step. Environmental perception involves obtaining detailed information about the surrounding environment, including the real-time position and motion state of the obstacle. These data help to update the robot's cognition of the environment in real time, so that it can adjust the path planning strategy in time to cope with dynamic environmental changes. The detour obstacle avoidance process is simulated according to the robot's real-time position data. This simulation is based on the current environmental perception data. The purpose is to evaluate different detour strategies and find the optimal obstacle avoidance path. Through simulation, the system can predict the situation encountered by the robot in actual execution and plan the best detour route in advance. The minimum detour angle is calculated according to the detour obstacle avoidance simulation data. The minimum detour angle refers to the minimum turning angle required for the robot to detour the obstacle. By accurately calculating this angle, the system can effectively reduce the length and time of the robot's detour path, improve the efficiency of path planning and the smoothness of the robot's movement. Based on the calculated minimum detour angle, the system optimizes the initial obstacle path data, which includes adjusting the direction and shape of the path so that the robot can move safely along the optimal obstacle avoidance path. Path optimization helps avoid unnecessary detours and increase the speed and efficiency of the robot's actions. The data acquisition and control strategy analysis of the robot arm is carried out under the consideration of the relationship between the robot arm and the surrounding environment. The posture of the robot arm is precisely controlled through the laser radar sensor data. This analysis ensures the accuracy and safety of the robot arm when performing tasks, while coordinating the overall movement of the robot.
[0066] Optionally, step S46 is specifically:
[0067] Step S461: Acquire the robot arm data, and perform control direction positioning on the robot arm data according to the laser radar sensor data, thereby obtaining the robot arm control direction data;
[0068] Step S462: estimating the extension distance of the robot arm data according to the laser radar sensing data, thereby obtaining the robot arm extension distance data;
[0069] Step S463: performing a robot arm control simulation according to the robot arm control direction data and the robot arm extension distance data, thereby obtaining the robot arm control simulation data;
[0070] Step S464: performing time error calculation on the robot arm control simulation data, thereby obtaining the robot arm control simulation time error data;
[0071] Step S465: extracting minimum time error control simulation parameters from the robot arm control simulation data according to the robot arm control simulation time error data, thereby obtaining the robot arm control data.
[0072] The present invention obtains environmental information, such as the position of obstacles and the current position of the manipulator, through a laser radar sensor. The control direction positioning is based on these data to ensure that the manipulator can accurately face the target direction, so that the manipulator will not collide with obstacles during operation, ensuring the safety and accuracy of the operation. The current extension distance of the manipulator is estimated using laser radar data, which is crucial for accurately positioning and operating the target object in space. By accurately estimating the extension distance, it can be ensured that the manipulator reaches the required position and height during operation, thereby completing complex tasks such as object grabbing or placement. The manipulator is controlled and simulated, which means that by simulating the movement and operation of the manipulator, the behavior in actual control can be predicted and optimized. This simulation can help engineers or system optimization algorithms test different control strategies, thereby improving the operating efficiency and accuracy of the manipulator. By comparing the simulation data with the actual execution time, the deviation between the simulation and the actual operation can be determined, which helps to adjust the control algorithm or improve the simulation accuracy, so that the manipulator is more reliable and stable in actual operation. By extracting the minimum time error control simulation parameters, the control algorithm can be adjusted or the model can be improved to ensure that the robot arm can be close to the expected motion and position control in actual operation. This optimization can significantly improve the operating accuracy and efficiency of the robot arm. BRIEF DESCRIPTION OF THE DRAWINGS
[0073] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:
[0074] Figure 1 A schematic diagram of the steps of the interactive data processing method for an intelligent robot system of the present invention;
[0075] Figure 2 Detailed step flow diagram of step S1 in the present invention;
[0076] Figure 3 Detailed flowchart of step S13 in the present invention.
[0077] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0078] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.
[0079] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.
[0080] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.
[0081] To achieve this, please refer to Figures 1 to 3 The present invention provides an interactive data processing method for an intelligent robot system, the method comprising the following steps:
[0082] Step S1: acquiring robot sensor data, and performing electric field detection on the robot sensor data, thereby obtaining electric signal change data; and digitally processing the electric signal change data, thereby obtaining a calibration electric signal instruction;
[0083] In this embodiment, the raw data of various sensors of the robot, such as temperature sensors and pressure sensors, are obtained. These sensors usually output analog signals, which are detected by electric field and digitally processed for subsequent control and analysis. For example, for a temperature sensor, the noise in the signal can be removed by an amplification and filtering circuit, and then the analog temperature value is converted into a digital signal using an analog-to-digital converter (ADC). For example, an operational amplifier and a filter circuit are used to pre-process the analog signal to ensure signal stability and accuracy. The operational amplifier is used to amplify the tiny signal output by the sensor, and the filter can select a low-pass or band-pass filter, and select a suitable filter type according to the signal frequency characteristics. Next, the processed analog signal is converted into digital data by a high-precision ADC, and an ADC with a resolution of 12 bits or higher is usually used to ensure data accuracy and dynamic range. The digital data obtained after such processing is the calibration electrical signal instruction, which can be used for subsequent control algorithms or uploaded to the cloud for real-time monitoring and remote control. The whole process not only ensures the accuracy of the sensor data, but also improves the stability and reliability of the robot under various environmental conditions.
[0084] Step S2: Acquire the robot application area structural data, perform three-dimensional building structure modeling according to the robot application area structural data, so as to obtain the robot application area structural model; perform radar sensing spatial position recognition on the robot application area structural model according to the robot sensing data, so as to obtain the robot real-time position data; perform path planning on the robot real-time position data according to the calibration electrical signal instruction, so as to obtain the initial path data, and upload it to the robot management cloud platform to execute the robot movement control task;
[0085] In this embodiment, the structural data of the robot application area is obtained by laser scanning, camera image processing or manual input of CAD design files. These data will be used to create a three-dimensional model of the robot application area, usually using computer-aided design (CAD) software. For example, the point cloud data obtained by laser scanning can be converted into a three-dimensional model, reflecting the geometric shape and structural characteristics of the area. The robot's sensor data, such as a radar sensor, is used to perform real-time spatial position recognition on this three-dimensional model. The radar sensor can detect the distance and position information of the surrounding environment, match this information with the pre-modeled structural data, and thus determine the real-time position of the robot. This step ensures the precise positioning of the robot in the application area. According to the calibration electrical signal instruction, the real-time position data of the robot is path planned. The path planning algorithm can calculate the optimal movement path based on factors such as the robot's current position, target position, and obstacle avoidance requirements. Common path planning algorithms include A* algorithm and Dijkstra algorithm, which can effectively take into account the factors of structural complexity and dynamic changes of the robot. The calculated initial path data is uploaded to the robot management cloud platform for executing the robot's mobile control task. The management cloud platform can not only receive path data, but also monitor the robot's position and motion status in real time to ensure the smooth execution and safety of the task.
[0086] Step S3: Performing real-time environmental perception according to the robot sensing data, thereby obtaining real-time robot environmental data; acquiring robot camera data, and performing target detection on the robot camera data according to the calibration electrical signal instruction, thereby obtaining a camera target object image; classifying the real-time robot environmental data according to the camera target object image, thereby obtaining obstacle data and pre-cleared obstacle data;
[0087] In this embodiment, the robot obtains real-time environmental data through a laser radar sensor, which includes information such as the location, distance and shape of surrounding objects. The sensor data is preprocessed and digitized for subsequent environmental perception and analysis. The robot camera captures environmental image data, which is subjected to target detection by an image processing algorithm. For example, a convolutional neural network (CNN) can be used to identify and locate target objects in an image, such as identifying people, vehicles or other obstacles. Based on the target object image, the robot classifies and analyzes the real-time environmental data, identifies various obstacles and performs pre-cleaning processing. This involves classifying obstacles into different categories, such as static obstacles and mobile obstacles, and making corresponding response strategies, such as avoidance or avoidance path planning. For example, high-precision sensors are used to obtain environmental data, and deep learning technology is combined to perform target detection and classification of camera data. For example, a deep learning framework such as TensorFlow or PyTorch is used to train the model, and the accuracy of target detection is improved by a large amount of pre-collected labeled data. Subsequently, using the real-time classification results, the robot can adjust its own actions or paths in a timely manner to ensure safe movement and operation in complex environments.
[0088] Step S4: performing obstacle avoidance optimization on the initial path data according to the obstacle data, thereby obtaining obstacle avoidance optimized path data; acquiring the robot arm data, and performing posture control strategy analysis on the robot arm data according to the robot sensor data, thereby obtaining the robot arm control data;
[0089] In this embodiment, the robot uses sensors to obtain obstacle data in the environment, including the location, size and shape information of the obstacles. Use a path planning algorithm (such as A algorithm or RRT algorithm) to optimize the initial path in combination with obstacle data to generate optimal path data to avoid obstacles. These algorithms take into account the dynamics of obstacles and the real-time position of the robot to ensure the safety and efficiency of the path. The robot's robotic arm data obtains its posture information through sensors. According to the environmental sensing data, posture control strategy analysis is performed to adjust the posture and movement of the robotic arm. For example, combining visual and force sensor data, accurate object grasping and placement operations can be achieved to ensure the accuracy and safety of the operation. For example, high-precision sensors (such as lidar, visual sensors and force sensors) are used to obtain environmental and robotic arm data, combined with advanced path planning and control algorithms for real-time optimization and adjustment. For example, using visual feedback and kinematic algorithms, adaptive posture control of the robotic arm in a complex environment is achieved to meet different operational task requirements.
[0090] Step S5: Integrate the robot control parameters of the obstacle avoidance optimization path data and the robot arm control data to obtain the robot optimization control data, and upload it to the robot management cloud platform to perform the robot movement correction task.
[0091] In this embodiment, the robot will obtain obstacle avoidance optimization path data and robotic arm control data through sensors. Use data fusion technology to integrate this information, such as using state estimation and filtering algorithms (such as Kalman filters) to reduce errors and improve data accuracy. Next, upload the integrated optimized control data to the robot management cloud platform. This requires the use of secure communication protocols and data transmission mechanisms, such as TLS encryption and HTTPs protocols, to ensure the confidentiality and integrity of the data. On the cloud platform, the robot's mobile correction tasks are performed through the remote control interface. Using remote control APIs and real-time data feedback, operators can monitor the status of the robot and adjust the movement path or posture. For example, the cloud server can send control instructions to the robot and verify the real-time effectiveness of the operation through a feedback mechanism.
[0092] Optionally, step S1 specifically includes:
[0093] Step S11: acquiring robot sensor data, and performing capacitive touch sensor data extraction on the robot sensor data, thereby obtaining capacitive touch sensor data;
[0094] In this embodiment, the acquisition of robot sensor data involves the use of multiple sensors to capture information about the environment and its own state. For example, the robot is equipped with a visual sensor, an inertial measurement unit (IMU), and a position encoder. In order to extract capacitive touch sensing data, a dedicated sensor component can be designed or the versatility of existing sensors can be utilized. For example, a visual sensor and a specific algorithm can be used to detect and track the tiny capacitance changes when the robot arm touches the surface, or a dedicated capacitive sensor array can be used.
[0095] Step S12: Drawing an electric field distribution diagram of the capacitive touch sensing data, thereby obtaining capacitive touch electric field distribution diagram data;
[0096] In this embodiment, the collected capacitive touch sensing data is preprocessed and analyzed. This includes filtering and denoising the data to ensure that only valid signals related to touch are retained. Next, the electric field distribution map is drawn and analyzed using data processing software (such as MATLAB or Python). By establishing a mathematical model and algorithm, the capacitance change obtained by the touch sensor is converted into electric field distribution map data to display the electric field strength distribution of the touch area.
[0097] Step S13: performing low-frequency electric field screening according to the capacitive touch electric field distribution map data, thereby obtaining electric signal change data;
[0098] In this embodiment, low-frequency electric field screening is performed to extract electrical signal changes related to touch events from the capacitive touch electric field distribution map data. For example, filtering methods in digital signal processing technology, such as low-pass filters or spectrum analysis, can be used to filter out high-frequency noise and retain low-frequency touch signals. Technical features include optimization of signal processing algorithms to ensure accurate detection of touch position and intensity. For example, using tools such as MATLAB or Python, the electric field distribution map data is input into a digital filter, and a suitable cutoff frequency is set to filter out high-frequency noise. Subsequently, time domain analysis or frequency domain analysis techniques are applied to capture and analyze low-frequency electric field changes, which reflect the electrical signal changes of the capacitive sensor when touched. Finally, effective electrical signal change data is extracted by setting an appropriate signal threshold or change rate for subsequent digital processing and control instruction generation.
[0099] Step S14: digitally process the electrical signal change data to obtain a calibration electrical signal instruction.
[0100] In this embodiment, the purpose of digitizing the electrical signal change data is to convert the analog signal obtained from the capacitive touch sensor into a calibration electrical signal instruction that the machine can understand and execute. The analog signal is converted into digital form using an analog-to-digital converter (ADC) to ensure that the data can be further analyzed and processed by a digital processing system. Technical features include high-precision ADC sampling rate and resolution to capture and retain subtle electrical signal changes. For example, the accuracy of the data can be ensured by selecting an ADC with a resolution of 16 bits or higher and setting an appropriate sampling rate (e.g., the number of samples per second). Subsequently, the digitized electrical signal data is cleaned and optimized using digital signal processing (DSP) techniques such as filtering, correction, and interpolation. This involves applying a digital filter (e.g., a low-pass filter) to further remove noise or using a differential amplifier to amplify tiny signal changes. Professional signal processing software (e.g., LabVIEW or a custom Python program) is used for data analysis and processing to identify the precise timing and intensity of the touch event. Finally, a calibration electrical signal instruction is generated according to a preset touch trigger condition, such as being sent to a robot control system to perform a corresponding action or task. This ensures high accuracy and stability from the capacitive touch sensor to the robot performing the action.
[0101] Optionally, step S13 is specifically:
[0102] Step S131: classifying the electric field distribution map data by electric field strength, thereby obtaining strong electric field distribution map data and weak electric field distribution map data;
[0103] In the present embodiment, a professional electric field sensor is used to obtain electric field distribution map data. The sensor should have high precision and a wide measurement range to ensure that the changes of various electric field strengths are captured. The electric field distribution map data is divided into strong electric field and weak electric field using data mining and machine learning techniques, such as cluster analysis or classification algorithms. For example, a threshold-based method or a feature-based classification method can be used for classification. For each classified data set, feature extraction and analysis are performed to determine the characteristic parameters of each electric field strength level, such as the mean value, standard deviation, etc. of the electric field strength. The classified strong electric field distribution map data and weak electric field distribution map data are stored in a suitable data structure, such as a database or a file system, for further processing and analysis. For example, for practical applications, an automated data processing script based on LabVIEW or Python programming can be used, combined with appropriate data visualization tools, such as Matplotlib or Plotly, to achieve electric field strength classification and data display.
[0104] Step S132: performing electric field frequency statistics on the strong electric field distribution map data, thereby obtaining low-frequency strong electric field distribution map data;
[0105] In the present embodiment, utilize digital signal processing technology, such as Fourier transform or power spectrum density estimation, the strong electric field distribution diagram data are carried out frequency analysis.This can help determine the periodicity and frequency component of electric field variation.According to the preset frequency range (such as 0-100Hz), the spectrum is carried out low-pass filtering process, to extract the low-frequency strong electric field distribution diagram data.Verify the accuracy of frequency statistics, and carry out necessary calibration to ensure the accuracy of frequency analysis and the reliability of data.For example, the SciPy library in MATLAB or Python can be used to realize spectrum analysis and digital filtering, thereby effectively extract the low-frequency strong electric field distribution diagram data.
[0106] Step S133: performing electric field frequency statistics on the weak electric field distribution map data, thereby obtaining low-frequency weak electric field distribution map data;
[0107] In the present embodiment, the frequency analysis technology is used to perform frequency analysis on the weak electric field distribution diagram data. This can help separate the electric field changes of the low-frequency part. The digital filtering technology is applied to remove the part with a frequency higher than the required low-frequency range to obtain the low-frequency weak electric field distribution diagram data. The frequency statistics are verified and calibrated to ensure the accuracy and reliability of the obtained low-frequency weak electric field distribution diagram data. For example, similar tools and methods can be used in the same data processing environment, such as spectrum analysis and filtering techniques in MATLAB or Python.
[0108] Step S134: screening the electric field distribution map data for low-frequency strong electric field distribution map data and low-frequency weak electric field distribution map data, thereby obtaining electric signal change data.
[0109] In this embodiment, the low-frequency strong electric field distribution map data and the low-frequency weak electric field distribution map data are integrated together. Using data processing and analysis techniques, such as differential operation or threshold detection, the change part of the electric field signal is detected and extracted. The detected electric signal changes are further processed and analyzed to generate final electric signal change data. These data can be time series, event trigger information or other forms for subsequent system control or decision support. For example, real-time processing and output of electric signal change data can be achieved by writing an automated script or using a real-time data processing system, such as Apache Kafka and Spark Streaming.
[0110] Optionally, step S14 is specifically:
[0111] Step S141: performing signal enhancement on the electrical signal change data to obtain enhanced electrical signal change data;
[0112] In the present embodiment, the Fourier transform algorithm is selected to enhance the specific frequency components of the electrical signal or improve the signal-to-noise ratio. According to the spectral characteristics and requirements of the electrical signal, a suitable digital filter is designed. Common filters include low-pass, high-pass, band-pass, etc., which are used to extract or suppress signal components within a specific frequency range. Select real-time processing or batch processing to perform signal enhancement. Real-time processing is suitable for real-time monitoring systems that require rapid response, while batch processing is suitable for offline data analysis and batch processing scenarios. Before performing signal enhancement, necessary data preprocessing is performed, such as removing baseline drift, data normalization, etc., to ensure the effectiveness and accuracy of the signal enhancement algorithm. For example, the wavelet transform toolbox in MATLAB or the SciPy library in Python can be used to implement the signal enhancement algorithm. For real-time processing, hardware acceleration technology, such as FPGA or GPU, can be considered to improve processing efficiency and response speed.
[0113] Step S142: performing bandpass filtering on the enhanced electrical signal change data to obtain corrected enhanced electrical signal change data;
[0114] In the present embodiment, a bandpass filter is designed to select signal components within a specific frequency range. The bandpass filter allows signals within the target frequency range to pass while suppressing signal components of other frequencies. According to the frequency characteristics of the electrical signal changes, appropriate bandpass filter parameters, such as center frequency and bandwidth, are selected. These parameters should be optimized according to the specific application scenario and signal spectrum. On the basis of signal enhancement, a bandpass filter is applied in real time, or the entire data set is filtered in batch processing. Necessary post-processing, such as smoothing or data interpolation, can be performed later. The filtering effect is evaluated to ensure that the signal components within the target frequency range are effectively enhanced and corrected while minimizing the interference of non-target frequencies. For example, Butterworth filters or FIR (finite impulse response) filters can be used to implement bandpass filtering, and the specific implementation can rely on digital signal processing tools and programming environments, such as SciPy libraries in MATLAB and Python.
[0115] Step S143: digitally quantizing the correction enhanced electrical signal change data to obtain a calibration electrical signal instruction.
[0116] In this embodiment, an appropriate digital quantization algorithm is selected to convert the continuous enhanced electrical signal change data into a discrete digital signal. Common algorithms include Pulse Code Modulation (PCM), Delta Modulation, etc. Quantization parameters, such as sampling rate and bit depth, are set to balance the accuracy of the signal and the efficiency of data transmission. The sampling rate determines the number of data points sampled per second, and the bit depth determines the quantization accuracy of each sample. Before digital quantization, the signal needs to be shaped or pre-coded to meet specific transmission or storage requirements. Analyze the errors introduced in the quantization process, such as quantization error and quantization noise, and consider whether calibration or compression algorithm is needed to optimize data transmission efficiency.
[0117] Optionally, step S2 specifically includes:
[0118] Step S21: obtaining robot application area structure data, and performing regional geometric structure feature extraction and identification area spatial feature extraction on the robot application area structure data, thereby obtaining application area geometric structure data and identification area spatial data;
[0119] In this embodiment, a combination of multiple sensors such as laser radar, RGB-D camera and inertial measurement unit (IMU) is used to obtain rich environmental data, and the data obtained by different sensors are fused to obtain more complete and accurate application area structure data. Through the point cloud data generated by laser radar, geometric features such as plane and curvature are extracted by deep learning or traditional feature extraction algorithms (such as RANSAC). The specific geometric structure information of the wall, the ground, etc. is fitted using a fitting algorithm (such as the least squares method). Semantic segmentation is performed based on RGB-D data to divide the environment into different categories, such as walls, furniture, etc. Analyze the spatial relationship between different areas, such as the placement of furniture, the relative position of walls, etc. The specific implementation method includes developing and deploying algorithms using a point cloud library (such as PCL) and a deep learning framework (such as TensorFlow, PyTorch) in ROS (Robot Operating System). For example, laser radar is used to obtain environmental point cloud data, and point cloud segmentation and feature extraction are performed in combination with a deep learning model, and finally the geometric structure data and spatial feature data of the robot application area are obtained, providing the necessary environmental cognition capabilities for robot navigation and task execution.
[0120] Step S22: performing three-dimensional building structure modeling according to the application area geometric structure data and the identification area spatial data, thereby obtaining a robot application area structure model;
[0121] In this embodiment, the geometric structure data obtained from the sensor is processed to ensure data accuracy and completeness. The geometric structure data is used to extract geometric features of walls, ceilings, floors, etc., such as planes, curvatures, and edges. A three-dimensional mesh model is generated based on the geometric feature data, including walls, furniture, and other structural elements. The model is adjusted using fitting algorithms and optimization techniques to ensure consistency and accuracy with the actual environment. Use modeling software based on point cloud data or develop custom modeling algorithms, such as using PointCloud Library (PCL) for point cloud processing and three-dimensional modeling; combine deep learning models for automatic segmentation and recognition to more efficiently generate building structure models; use modern CAD software or specialized robot navigation modeling tools, such as the modeling package in Robot Operating System (ROS), to achieve three-dimensional modeling and model optimization.
[0122] Step S23: extracting laser radar sensor data from the robot sensor data to obtain laser radar sensor data, and performing radar sensor spatial position recognition on the laser radar sensor data according to the robot application area structure model to obtain geographic coordinate data and speed direction data;
[0123] In this embodiment, the robot is equipped with a 360-degree rotating laser radar sensor that can perform hundreds of thousands of scans per second to generate high-density point cloud data. Using a previously established indoor building structure model, such as using BIM (Building Information Modeling) or SLAM (Simultaneous Localization and Mapping) technology, the point cloud data acquired in real time is compared and matched with the model. By matching the point cloud data with the structural model, the specific position coordinates and posture (direction and tilt angle) of the robot in the indoor space are identified. Based on the identified position coordinates and posture, the indoor coordinates are converted into geographic coordinates, and the robot's movement speed and direction are calculated by analyzing the changes in the point cloud data.
[0124] Step S24: performing position calculation based on the geographic coordinate data and the speed direction data, thereby obtaining the real-time position data of the robot;
[0125] In this embodiment, the real-time position of the robot is calculated by using a GPS receiver to obtain geographic coordinate data, and an encoder or inertial measurement unit (IMU) to obtain speed and direction data. GPS provides global positioning accuracy, while the encoder or IMU can provide instant speed and direction change information. The strength and availability of the GPS signal are monitored in real time, and high-precision encoders or IMU sensors are used to ensure the accuracy of position and motion data. For example, a GPS module, encoder or IMU is installed on the robot, and data fusion and calibration are performed in combination with algorithms to achieve high-precision real-time position calculation and navigation capabilities.
[0126] Step S25: dividing the robot real-time position data into application ranges, thereby obtaining the robot real-time position application range data;
[0127] In this embodiment, the collection method and transmission protocol of the real-time position data of the robot are determined. For example, the position data is collected using GPS, an inertial navigation system, or a visual sensor, and transmitted to a control center via a wireless network (such as Wi-Fi, Bluetooth) or a wired interface. A processing mechanism is designed to parse and analyze the real-time position data to meet specific application requirements. For example, the current position of the robot is compared with the target position, or the robot's motion trajectory is monitored in real time. A path planning algorithm based on machine learning is developed, which combines historical position data and environmental information to adjust the robot's path in real time, avoid obstacles, and optimize energy consumption.
[0128] Step S26: Perform path planning on the robot's real-time position application range data according to the calibration electrical signal instruction, thereby obtaining initial path data, and uploading it to the robot management cloud platform to execute the robot movement control task.
[0129] In this embodiment, the robot's real-time position path planning is performed using calibration electrical signal instructions. The robot's initial position and direction are calibrated by electrical signal instructions to ensure accuracy and reliability. High-precision indoor positioning technology (such as laser SLAM or ultrasonic positioning), real-time path planning algorithms and data transmission protocols are used. For example, a positioning sensor network is installed in a building, and the robot's accurate position and orientation are obtained through calibration electrical signal instructions that communicate with the robot, and then uploaded to the cloud platform for path planning. Such a system can ensure that the robot can safely and efficiently perform mobile control tasks in complex indoor environments.
[0130] Optionally, step S26 is specifically:
[0131] Step S261: performing path planning on the reasonable range data of the real-time position of the robot according to the calibration electrical signal instruction, thereby obtaining planned path data;
[0132] In this embodiment, during the robot navigation process, the robot's current real-time position data is accurately planned according to the calibration electrical signal instructions. The robot's current position is determined by advanced indoor positioning technology (such as laser SLAM or visual recognition), and then the optimal path is calculated by a path planning algorithm (such as A* algorithm or Dijkstra algorithm). For example, in the application scenario of a storage center, the robot determines its own position through the received signal calibration, and then dynamically adjusts the path according to the inventory layout to achieve efficient cargo transportation and automated operation.
[0133] Step S262: extracting path wind resistance characteristics and path terrain characteristics from the planned path data, thereby obtaining path wind resistance data and path terrain data;
[0134] In this embodiment, the path wind resistance feature extraction and path terrain feature extraction are performed based on the planned path data. The wind resistance information and terrain information on the path, such as wind speed, slope, etc., are obtained through sensors. High-precision sensor technology and data acquisition systems are used to ensure accurate acquisition of path feature data. For example, in an outdoor robot inspection task, the wind speed and terrain slope information on the path are obtained through GPS and meteorological sensors.
[0135] Step S263: calculating the drag coefficient of the path wind resistance data, thereby obtaining path low wind resistance data;
[0136] In this embodiment, in the robot path planning, sensors are used to obtain wind speed and wind direction data on the path, and the path drag coefficient is calculated in combination with a mathematical model. High-precision meteorological sensors (such as anemometers and wind direction sensors), real-time data acquisition systems, and drag coefficient calculation algorithms (such as aerodynamic models) are used. For example, wind speed and wind direction sensors are installed in an outdoor environment, and the drag coefficient of each path is calculated by regularly collecting data and inputting it into the calculation model, so as to select the optimal path and ensure that the robot is minimized by the wind during movement.
[0137] Step S264: Calculate the ruggedness of the path terrain data to obtain path flatness data;
[0138] In this embodiment, in the robot path planning, the terrain height data on the path is obtained by using sensors, and the ruggedness of the path is calculated by an algorithm to obtain the flatness data of the path. Technical features include accurate terrain sensors (such as laser rangefinders or high-precision terrain radars), real-time data acquisition, and terrain analysis algorithms (such as digital elevation model processing). For example, a terrain sensor network is deployed in the robot movement area, and the sensor regularly measures the terrain height, uploads the data to the central processing unit for ruggedness calculation, and generates a path flatness map to help the robot select a flat and stable path to ensure the safety and efficiency of the mobile task.
[0139] Step S265: performing an intersection operation based on the low wind resistance data of the path and the flat path data, thereby obtaining low risk path data;
[0140] In this embodiment, in the robot path planning, the low wind resistance data of the path and the flat path data are combined to perform intersection operations to determine a low-risk path. High-precision sensors (such as meteorological sensors and terrain sensors), data analysis algorithms and path optimization models are used. For example, the robot is equipped with meteorological and terrain sensors to collect real-time wind speed, wind direction and terrain height data, and then the algorithm is used to process and perform intersection operations on these data to generate a low-risk path map. For example, by comparing the wind resistance coefficients and terrain ruggedness of different paths, the best path is selected so that the robot can travel safely and efficiently even in bad weather or complex terrain conditions. This ensures that the robot avoids unnecessary risks and delays when performing tasks.
[0141] Step S266: Perform low-risk path screening on the path data according to the low-risk path data to obtain initial path data, and upload it to the robot management cloud platform to execute the robot movement control task.
[0142] In this embodiment, in the robot path planning, the path screening is performed through the low-risk path data to obtain the initial path data, and uploaded to the robot management cloud platform to perform the mobile control task. A path optimization algorithm, cloud data storage and real-time update mechanism are adopted. Using a path planning algorithm such as A* or Dijkstra, potential paths are evaluated and screened based on the collected low-risk path data. The optimal path data is then uploaded to the cloud platform for the robot to access and execute tasks in real time. For example, in factory automation, after receiving the low-risk path data through the cloud management system, the robot can make autonomous decisions and adjust the movement trajectory, thereby ensuring safe and efficient operation in a dynamic environment.
[0143] Optionally, step S3 specifically includes:
[0144] Step S31: Perform real-time environmental perception based on the laser radar sensing data to obtain real-time robot environmental data;
[0145] In this embodiment, a laser radar sensor is used to scan the surrounding environment in real time to obtain a high-precision environment map and obstacle location data. 360-degree scanning capability, millimeter-level precision distance measurement and real-time data processing are adopted. The laser radar sensor is installed on the robot, scans multiple times per second, and generates an environment map through a data processing algorithm so that the robot can plan and adjust the path.
[0146] Step S32: performing target detection on the laser radar sensing data according to the calibration electrical signal instruction, thereby obtaining the laser radar sensing target data;
[0147] In this embodiment, the laser radar data is used for target detection using calibrated electrical signal instructions to identify the position and shape of the target object. Signal processing algorithms, target segmentation and classification models are used. The laser radar data is input into the target detection algorithm to identify obstacles or targets and extract their position information for subsequent obstacle avoidance and navigation decisions.
[0148] Step S33: acquiring a robot camera image, and performing target detection on the robot camera image according to the calibration electrical signal instruction, thereby obtaining a camera target object image;
[0149] In this embodiment, a robot camera is used to capture real-time images, and target detection is performed according to electrical signal instructions to extract target object data in the image. Computer vision algorithms, real-time image processing and target recognition models are used. The camera captures environmental images in real time, identifies objects in the image through deep learning models or traditional image processing algorithms, and obtains their features and location information.
[0150] Step S34: integrating the laser radar sensing target data and the camera target object image to obtain the initial confirmed target object data;
[0151] In this embodiment, the laser radar target data and the camera image target object data are integrated to perform data fusion and preliminary confirmation of the target object. Sensor data fusion algorithm and multi-sensor information integration technology are used. The laser radar data and camera data are matched and integrated to enhance the confirmation and positioning accuracy of the target object.
[0152] Step S35: obtaining robot knowledge database data, and performing parameter matching on the initially confirmed target object data according to the robot knowledge database data, thereby obtaining pre-cleared obstacle data;
[0153] In this embodiment, the preset information in the robot knowledge database is obtained, and the parameters of the target object data confirmed for the first time are matched and verified. The characteristics and behavior patterns of the relevant objects are retrieved from the knowledge database using database query and analysis technology, pattern recognition and database management system, and compared and verified with the data actually perceived by the robot to confirm the identity and attributes of the target object.
[0154] Step S36: pre-cleaning obstacle data of the real-time robot environment data according to the pre-cleaning obstacle data to obtain obstacle data;
[0155] In this embodiment, the real-time robot environment data is screened according to the pre-cleaned obstacle data, and the obstacle data is identified and cleaned. A data cleaning and filtering algorithm, a real-time update and feedback mechanism are used. Based on a preset obstacle database or model, the real-time perceived obstacle data is compared with it, and the environment map is dynamically updated to ensure that the robot can avoid obstacles during navigation and complete the task safely and efficiently.
[0156] Optionally, step S33 is specifically:
[0157] Step S331: extracting target object features from the calibration electrical signal instruction, thereby obtaining target object data;
[0158] In this embodiment, the target object feature extraction is performed on the data received by the robot according to the pre-set calibration electrical signal instruction. This involves the application of signal processing and feature extraction algorithms. For example, spectrum analysis technology can be used to extract the frequency characteristics of the signal, or a pattern recognition algorithm can be used to identify a target object of a specific shape or structure. The robot will use a specific algorithm to parse and extract the key features of the target object, such as size, shape or other identifiable feature information, through the received electrical signal data.
[0159] Step S332: acquiring a robot camera image, and performing grayscale conversion on the robot camera image, thereby obtaining a robot camera grayscale image;
[0160] In this embodiment, the robot camera image is obtained and grayscale conversion is performed for subsequent image processing operations. Grayscale image conversion is a common preprocessing step in image processing. Converting a color image to a grayscale image can simplify the subsequent image analysis and feature extraction process. Use an image processing library or algorithm to convert the RGB channel values of a color image to grayscale values to obtain grayscale image data.
[0161] Step S333: performing background subtraction on the robot camera grayscale image to obtain the robot camera object grayscale image;
[0162] In this embodiment, background subtraction is performed on the grayscale image of the robot camera to extract the target object in the image. Background subtraction is a commonly used image segmentation technology that separates the background from the foreground in the image to highlight the boundaries and features of the target object. Background modeling algorithms and foreground extraction techniques are used, such as those based on differential image method or Gaussian mixture model. The robot will perform corresponding background subtraction operations on the grayscale image based on the extracted target object features to obtain a clear grayscale image of the target object.
[0163] Step S334: performing parameter matching on the robot camera object grayscale image according to the target object data, thereby obtaining the camera target object image.
[0164] In this embodiment, according to the previously extracted target object data, the robot camera object grayscale image after background subtraction is parameter matched to obtain the final camera target object image. The obtained target object features are matched and confirmed with the image to ensure that the extracted image data conforms to the expected target object features and shape. The target object data is compared with the image after background subtraction using pattern recognition technology or feature matching algorithm to obtain accurate target object image data.
[0165] Optionally, step S4 is specifically:
[0166] Step S41: extracting obstacle path features from the initial path data according to the obstacle data, thereby obtaining initial obstacle path data;
[0167] In this embodiment, obstacle path features are extracted from the initial path data based on the obstacle data, and obstacles on the path are identified and analyzed, thereby optimizing the path planning and ensuring that the robot can avoid obstacles and safely reach the target location when performing tasks. Image processing technology or sensor data analysis algorithms, such as image segmentation algorithms or lidar data processing, are used to extract obstacle features on the path. For example, for a visual system, obstacles on the path can be identified through edge detection or object recognition algorithms, and then this information can be converted into obstacle path data for subsequent environmental perception and path planning.
[0168] Step S42: performing environmental perception on the initial obstacle path data according to the laser radar sensing data, thereby obtaining initial obstacle path position data;
[0169] In this embodiment, environmental perception is performed on the initial obstacle path data based on the laser radar sensing data, aiming to obtain the precise location and spatial information of the obstacle through accurate environmental perception, so as to perform more accurate path planning and obstacle avoidance operations. During implementation, the robot will use the laser radar sensor to obtain data such as the distance, direction, and shape of obstacles in the environment. These data can help the robot accurately build an environmental map and determine the location of obstacles in the path. For example, through laser radar scanning and distance measurement, the robot can update the location and dynamic changes of obstacles in real time, thereby adjusting the path planning to ensure safe passage.
[0170] Step S43: performing obstacle avoidance simulation on the initial obstacle path position data according to the real-time position data of the robot, thereby obtaining obstacle avoidance simulation data of the robot;
[0171] In this embodiment, the initial obstacle path position data is simulated for detour and obstacle avoidance according to the real-time position data of the robot, with the purpose of simulating the detour strategy of the robot when encountering obstacles in the actual environment, and generating detour and obstacle avoidance simulation data. According to the real-time environment conditions and the current position of the robot, the path and action strategy for avoiding obstacles are simulated. For example, based on the path planning algorithm, the robot can simulate different obstacle avoidance path options, evaluate the safety and efficiency of each path, and determine the best obstacle avoidance action.
[0172] Step S44: Calculating the minimum detour angle of the robot's detour obstacle avoidance simulation data, thereby obtaining the robot's minimum detour obstacle avoidance angle data;
[0173] In this embodiment, the minimum detour angle calculation is performed on the robot's obstacle avoidance simulation data to determine the minimum angle that the robot needs to adjust when encountering an obstacle, so as to bypass the obstacle. This involves path planning and dynamic adjustment algorithms, such as minimum turning radius calculation or optimal path selection algorithm, to ensure that the robot can avoid obstacles at the smallest angle and in the shortest time.
[0174] Step S45: Optimizing the initial obstacle path data by the minimum detour angle according to the robot's minimum detour angle data, thereby obtaining obstacle avoidance optimization path data;
[0175] In this embodiment, the initial obstacle path data is optimized for the minimum detour angle according to the robot's minimum detour angle data to generate obstacle avoidance optimized path data. The optimized path can improve the robot's obstacle avoidance efficiency and safety by adjusting the angle and direction in the path. For example, through a dynamic path planning algorithm and real-time environmental perception, the robot can optimize the path while avoiding obstacles, reduce the path length or reduce the steering angle in the path to improve the efficiency of task execution.
[0176] Step S46: Acquire the robot arm data, and perform posture control strategy analysis on the robot arm data according to the laser radar sensor data, so as to obtain the robot arm control data.
[0177] In this embodiment, the robot data is acquired, and the posture control strategy analysis of the robot data is performed based on the laser radar sensor data to generate the robot control data. This involves analyzing and generating appropriate robot motion control strategies based on the position of obstacles in the environment and the robot's motion trajectory to ensure that the robot avoids obstacles and completes the predetermined task during the work process. For example, based on the real-time data scanned by the laser radar, the robot can adjust the posture and motion path of the robot arm to ensure that the work operation is completed safely and efficiently.
[0178] Optionally, step S46 is specifically:
[0179] Step S461: Acquire the robot arm data, and perform control direction positioning on the robot arm data according to the laser radar sensor data, thereby obtaining the robot arm control direction data;
[0180] In this embodiment, the sensor system of the robot arm obtains key position and posture data, such as joint angles and the position of the end effector. Then, the laser radar sensor is used to obtain the position of the target object in the environment and the position relationship of the robot arm relative to the target. By combining these data, the direction in which the robot arm needs to face can be accurately calculated to ensure that it can accurately perform subsequent operations.
[0181] Step S462: estimating the extension distance of the robot arm data according to the laser radar sensing data, thereby obtaining the robot arm extension distance data;
[0182] In this embodiment, the environmental data acquired by the laser radar sensor, especially the position and geometric features of the target object related to the operation of the robot arm, are used to estimate the extension distance that the robot arm needs to perform in order to correctly operate or move to the target position.
[0183] Step S463: performing a robot arm control simulation according to the robot arm control direction data and the robot arm extension distance data, thereby obtaining the robot arm control simulation data;
[0184] In this embodiment, the system will simulate the motion path and actions of the robot arm. For example, using the motion planning algorithm, the system calculates the motion trajectory of each joint to ensure that the robot arm meets the kinematic requirements such as speed and acceleration while controlling the direction and extension distance to achieve the expected operation tasks, such as assembling parts on an assembly line or grasping products in automated production.
[0185] Step S464: performing time error calculation on the robot arm control simulation data, thereby obtaining the robot arm control simulation time error data;
[0186] In this embodiment, the system analyzes the time difference between the actual robot arm movement and the simulated movement. By comparing the timestamps of the simulated path and the actual movement, the time error is calculated, reflecting the synchronization and accuracy between the simulation system and the actual physical robot arm. For example, in precision machining, the time error data will be used to adjust the control algorithm to improve the accuracy and efficiency of the robot arm's movement.
[0187] Step S465: extracting minimum time error control simulation parameters from the robot arm control simulation data according to the robot arm control simulation time error data, thereby obtaining the robot arm control data.
[0188] In this embodiment, based on the time error analysis, the system optimizes the parameter settings of the control algorithm to make the simulated motion of the robot arm more consistent with the actual motion. For example, by increasing the frequency of the control loop or adjusting the accuracy of the path planning algorithm, the system can reduce the time error, improve the control accuracy and response speed of the robot arm in complex environments, and ensure the accurate completion of tasks, such as the installation of precise components or high-speed packaging tasks on an automated assembly line.
[0189] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.
[0190] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.
Claims
1. An interactive data processing method for an intelligent robot system, characterized in that: The following steps are involved: Step S1: acquiring robot sensor data, and performing electric field detection on the robot sensor data, thereby obtaining electrical signal change data; Digitally process the electrical signal change data to obtain calibration electrical signal instructions; Step S2: Acquire robot application area structural data, perform three-dimensional building structure modeling based on the robot application area structural data, thereby obtaining a robot application area structural model; perform radar sensing spatial position recognition on the robot application area structural model based on the robot sensing data, thereby obtaining the robot's real-time position data; Perform path planning on the robot’s real-time position data according to the calibration electrical signal instructions to obtain initial path data, which is then uploaded to the robot management cloud platform to execute the robot’s mobile control tasks; Step S3: Perform real-time environmental perception according to the robot sensing data, thereby obtaining real-time robot environmental data; Acquire robot camera data, and perform target detection on the robot camera data according to the calibration electrical signal instruction, so as to obtain the camera target object image; Classify real-time robot environment data based on camera target object images to obtain obstacle data and pre-cleared obstacle data; Step S4: performing obstacle avoidance optimization on the initial path data according to the obstacle data, thereby obtaining obstacle avoidance optimized path data; acquiring the robot arm data, and performing posture control strategy analysis on the robot arm data according to the robot sensor data, thereby obtaining the robot arm control data; Step S5: Integrate the robot control parameters of the obstacle avoidance optimization path data and the robot arm control data to obtain the robot optimization control data, and upload it to the robot management cloud platform to perform the robot movement correction task.
2. The interactive data processing method for an intelligent robot system according to claim 1, characterized in that: Step S1 is specifically as follows: Step S11: acquiring robot sensor data, and performing capacitive touch sensor data extraction on the robot sensor data, thereby obtaining capacitive touch sensor data; Step S12: Drawing an electric field distribution diagram of the capacitive touch sensing data, thereby obtaining capacitive touch electric field distribution diagram data; Step S13: performing low-frequency electric field screening according to the capacitive touch electric field distribution map data, thereby obtaining electric signal change data; Step S14: digitally process the electrical signal change data to obtain a calibration electrical signal instruction.
3. The interactive data processing method for an intelligent robot system according to claim 2, characterized in that: Step S13 is specifically as follows: Step S131: classifying the electric field distribution map data by electric field strength, thereby obtaining strong electric field distribution map data and weak electric field distribution map data; Step S132: performing electric field frequency statistics on the strong electric field distribution map data, thereby obtaining low-frequency strong electric field distribution map data; Step S133: performing electric field frequency statistics on the weak electric field distribution map data, thereby obtaining low-frequency weak electric field distribution map data; Step S134: screening the electric field distribution map data for low-frequency strong electric field distribution map data and low-frequency weak electric field distribution map data, thereby obtaining electric signal change data.
4. The interactive data processing method for an intelligent robot system according to claim 2, characterized in that: Step S14 is specifically as follows: Step S141: performing signal enhancement on the electrical signal change data to obtain enhanced electrical signal change data; Step S142: performing bandpass filtering on the enhanced electrical signal change data to obtain corrected enhanced electrical signal change data; Step S143: digitally quantizing the correction enhanced electrical signal change data to obtain a calibration electrical signal instruction.
5. The interactive data processing method for an intelligent robot system according to claim 1, characterized in that: Step S2 is specifically as follows: Step S21: obtaining robot application area structure data, and performing regional geometric structure feature extraction and identification area spatial feature extraction on the robot application area structure data, thereby obtaining application area geometric structure data and identification area spatial data; Step S22: performing three-dimensional building structure modeling according to the application area geometric structure data and the identification area spatial data, thereby obtaining a robot application area structure model; Step S23: extracting laser radar sensor data from the robot sensor data to obtain laser radar sensor data, and performing radar sensor spatial position recognition on the laser radar sensor data according to the robot application area structure model to obtain geographic coordinate data and speed direction data; Step S24: performing position calculation based on the geographic coordinate data and the speed direction data, thereby obtaining the real-time position data of the robot; Step S25: dividing the robot real-time position data into application ranges, thereby obtaining the robot real-time position application range data; Step S26: Perform path planning on the robot's real-time position application range data according to the calibration electrical signal instruction, thereby obtaining initial path data, and uploading it to the robot management cloud platform to execute the robot movement control task.
6. The interactive data processing method for an intelligent robot system according to claim 5, characterized in that: Step S26 is specifically as follows: Step S261: performing path planning on the robot's real-time position reasonable range data according to the calibration electrical signal instruction, thereby obtaining planned path data; Step S262: extracting path wind resistance characteristics and path terrain characteristics from the planned path data, thereby obtaining path wind resistance data and path terrain data; Step S263: calculating the drag coefficient of the path wind resistance data, thereby obtaining path low wind resistance data; Step S264: Calculate the ruggedness of the path terrain data to obtain path flatness data; Step S265: performing an intersection operation based on the low wind resistance data of the path and the flat path data, thereby obtaining low risk path data; Step S266: Perform low-risk path screening on the path data according to the low-risk path data to obtain initial path data, and upload it to the robot management cloud platform to execute the robot movement control task.
7. The interactive data processing method for an intelligent robot system according to claim 1, characterized in that: Step S3 is specifically as follows: Step S31: Perform real-time environmental perception based on laser radar sensing data to obtain real-time robot environmental data; Step S32: performing target detection on the laser radar sensing data according to the calibration electrical signal instruction, thereby obtaining the laser radar sensing target data; Step S33: acquiring a robot camera image, and performing target detection on the robot camera image according to the calibration electrical signal instruction, thereby obtaining a camera target object image; Step S34: integrating the laser radar sensing target data and the camera target object image to obtain the initial confirmed target object data; Step S35: obtaining robot knowledge database data, and performing parameter matching on the initially confirmed target object data according to the robot knowledge database data, thereby obtaining pre-cleared obstacle data; Step S36: Pre-cleaning obstacle data is screened out from the real-time robot environment data according to the pre-cleaning obstacle data, thereby obtaining obstacle data.
8. The interactive data processing method for an intelligent robot system according to claim 7, characterized in that: Step S33 is specifically as follows: Step S331: extracting target object features from the calibration electrical signal instruction, thereby obtaining target object data; Step S332: acquiring a robot camera image, and performing grayscale conversion on the robot camera image, thereby obtaining a robot camera grayscale image; Step S333: performing background subtraction on the robot camera grayscale image to obtain the robot camera object grayscale image; Step S334: performing parameter matching on the robot camera object grayscale image according to the target object data, thereby obtaining the camera target object image.
9. The interactive data processing method for an intelligent robot system according to claim 1, characterized in that: Step S4 is specifically as follows: Step S41: extracting obstacle path features from the initial path data according to the obstacle data, thereby obtaining initial obstacle path data; Step S42: performing environmental perception on the initial obstacle path data according to the laser radar sensing data, thereby obtaining initial obstacle path position data; Step S43: performing obstacle avoidance simulation on the initial obstacle path position data according to the real-time position data of the robot, thereby obtaining obstacle avoidance simulation data of the robot; Step S44: Calculating the minimum detour angle of the robot's detour obstacle avoidance simulation data, thereby obtaining the robot's minimum detour obstacle avoidance angle data; Step S45: Optimizing the initial obstacle path data by the minimum detour angle according to the robot's minimum detour angle data, thereby obtaining obstacle avoidance optimization path data; Step S46: Acquire the robot arm data, and perform posture control strategy analysis on the robot arm data according to the laser radar sensor data, so as to obtain the robot arm control data.
10. The interactive data processing method for an intelligent robot system according to claim 9, characterized in that: Step S46 is specifically as follows: Step S461: Acquire the robot arm data, and perform control direction positioning on the robot arm data according to the laser radar sensor data, thereby obtaining the robot arm control direction data; Step S462: estimating the extension distance of the robot arm data according to the laser radar sensing data, thereby obtaining the robot arm extension distance data; Step S463: performing a robot arm control simulation according to the robot arm control direction data and the robot arm extension distance data, thereby obtaining the robot arm control simulation data; Step S464: performing time error calculation on the robot arm control simulation data, thereby obtaining the robot arm control simulation time error data; Step S465: extracting minimum time error control simulation parameters from the robot arm control simulation data according to the robot arm control simulation time error data, thereby obtaining the robot arm control data.
Citation Information
Cited By
Inspection method and system of inspection robot, storage medium and electronic equipment
CN122363230A
An inspection robot inspection method, system, storage medium, and electronic equipment.
CN122363230B
