Mechanical arm cleaning robot system based on zero-code digital twinning

Abstract

The invention provides a mechanical arm cleaning robot system based on zero-code digital twinning and a design method, and relates to the technical field of cleaning robot design. The method comprises the following steps: establishing a locomotive object scene three-dimensional digital twinborn model based on a locomotive body cleaning scene; a mechanical arm digital twinning model of the cleaning mechanical arm is collected, digital twinning driving is carried out, and mechanical arm basic driving data are formed; locomotive body recognition and classification data are obtained, and mechanical arm optimized cleaning scheme data are formed; acquiring locomotive cleaning image data, and establishing a cleaning effect evaluation model; locomotive cleaning operation is carried out based on the mechanical arm optimized cleaning scheme data, and real-time monitoring maintenance analysis result data is formed; and performing intelligent operation analysis according to the real-time monitoring maintenance analysis result data to form intelligent operation analysis result data. According to the method, the intelligence, self-adaption and collaboration capabilities of the cleaning robot are improved, the quality and efficiency of the cleaning service are improved, and technical support is provided for innovative development of the cleaning industry.

Description

Technical Field

[0001] The present invention relates to the technical field of cleaning robot design, and in particular to a mechanical arm cleaning robot system based on zero-code digital twin. Background Art

[0002] With the development of industrialization and urbanization, the cleaning work of large facilities such as high-rise buildings, aircraft, and ships is increasing, but the traditional manual cleaning methods have problems such as low efficiency, high cost, and poor safety, which limit the quality and scale of cleaning services. According to statistics, the scale of the global cleaning service market reached 1.74 trillion US dollars in 2020, and is expected to grow to 2.41 trillion US dollars in 2027, with a compound annual growth rate of 4.8%. However, the cleaning industry is still facing challenges in technical bottlenecks and development trends, such as the complexity of the surface shape and structure of the cleaning object, the variability of the environmental conditions of the cleaning object, the differences in the cleaning requirements and standards of the cleaning object, and the complexity of the collaborative tasks of the cleaning object and the cleaning tools. Therefore, the development of an intelligent, efficient, and safe cleaning robot system has important engineering significance and social value.

[0003] Locomotives are a widely used means of transportation, and the cleaning of their bodies and parts is an important part of ensuring the performance and safety of locomotives. However, locomotive cleaning currently relies mainly on manual or track-based, vehicle-mounted, drone-based cleaning robots, which have some disadvantages, such as insufficient adaptability to the complex shapes and structures of locomotive bodies and parts, weak positioning and identification capabilities of locomotive bodies and parts, difficulty in achieving accurate and flexible cleaning actions, low ability to adjust cleaning parameters, poor ability to evaluate cleaning effects, low efficiency in the use of cleaning resources, low adaptability to the cleaning environment, and poor cleaning safety and reliability.

[0004] Therefore, it is an urgent problem to design a robotic arm cleaning robot system based on zero-code digital twin technology, realize real-time digital twin drive of locomotive body and components, improve the intelligence, adaptability and collaboration capabilities of the cleaning robot, improve the quality and efficiency of cleaning services, and provide technical support for the innovative development of the cleaning industry. Summary of the invention

[0005] The purpose of the present invention is to provide a robotic arm cleaning robot system based on zero-code digital twins, which can improve the intelligence, adaptability and collaboration of the cleaning robot by realizing real-time digital twin driving of the locomotive body and components, improve the quality and efficiency of cleaning services, and provide technical support for the innovative development of the cleaning industry.

[0006] In the first aspect, the present invention provides a design method for a robotic arm cleaning robot based on zero-code digital twins, comprising: based on a locomotive body cleaning scene, establishing a three-dimensional digital twin model of a locomotive object scene for a cleaning object through a zero-code platform; acquiring a robotic arm digital twin model of the cleaning robotic arm, and combining it with the three-dimensional data twin model of the locomotive object scene to perform digital twin driving to form basic robotic arm driving data; acquiring locomotive body identification and classification data, and performing generation optimization analysis of cleaning strategies and plans in combination with the robotic arm basic driving data to form robotic arm optimized cleaning plan data; acquiring locomotive cleaning image data, and performing evaluation and analysis of the cleaning effect to establish a cleaning effect evaluation model; performing locomotive cleaning operations based on the robotic arm optimized cleaning plan data, and acquiring real-time cleaning data for operation monitoring and maintenance analysis to form real-time monitoring and maintenance analysis result data; performing intelligent operation analysis based on the real-time monitoring and maintenance analysis result data to form intelligent operation analysis result data.

[0007] In the present invention, the method realizes real-time digital twin driving of the locomotive body and components, improves the intelligence, adaptability and coordination capabilities of the cleaning robot, improves the quality and efficiency of cleaning services, and provides technical support for the innovative development of the cleaning industry.

[0008] As a possible implementation method, based on the locomotive body cleaning scenario, a three-dimensional digital twin model of the locomotive cleaning scenario for the cleaning object is established through the zero-code platform, including: based on the locomotive body cleaning scenario, a digital twin model of the main equipment in the scene is established through the zero-code platform to form a three-dimensional digital twin model of the locomotive cleaning scene that matches the real cleaning scene; real-time data of the locomotive body and components is collected, and data twin models of different locomotive bodies and components are established using the zero-code platform to form a three-dimensional data twin model of the locomotive body.

[0009] In the present invention, the locomotive body cleaning scene is used as a specific application scene, and the self-developed zero-code platform and zero-code technology are used to quickly build the digital twin model of the locomotive body and locomotive cleaning scene, as well as the main equipment in the scene, such as the cleaning tray with components, dry ice tank, ice maker, AGV trolley, AGV parking space, AGV charging pile, etc. Among them, through the zero-code digital twin development platform, digital twin applications can be quickly built and delivered by dragging and dropping, without writing any code, which greatly reduces the threshold and difficulty of development.

[0010] As a possible implementation method, the digital twin model of the cleaning robot arm is collected, and combined with the three-dimensional data twin model of the locomotive object, digital twin driving is performed to form the basic driving data of the robot arm, including: combining plug-in technology, Internet of Things technology and big data technology for locomotive cleaning, according to the three-dimensional data twin model of the locomotive cleaning scene and the three-dimensional data twin model of the locomotive body, to determine the basic driving information of the cleaning robot and form the basic driving data of the robot arm.

[0011] In the present invention, based on the digital twin model, plug-in technology, Internet of Things technology, and big data technology are used to realize the real-time digital twin drive of the mechanical arm of the cleaning robot, as well as the coordinated control with the AGV car, to realize the real-time cleaning operation and feedback of the locomotive body. Through rich industry templates and resources, the basic parameters of the digital twin model, such as joint sensitivity, joint range of motion, upper limits of linear velocity and acceleration of each joint, motion data compensation, etc., can be set according to the actual parameters and models of the locomotive cleaning mechanical arm, so as to realize the digital twin drive of the locomotive cleaning mechanical arm.

[0012] As a possible implementation method, locomotive body identification and classification data is obtained, and the generation and optimization analysis of cleaning strategies and plans is performed in combination with the basic driving data of the robot arm to form the robot arm optimized cleaning plan data, including: obtaining image data of different locomotive bodies and parts, and performing feature extraction and classification to form locomotive body classification and identification data; collecting cleaning action control data, and combining the locomotive body classification and identification data with the basic driving data of the robot arm to generate a strategy plan based on model training, and obtaining cleaning effect data during the model training process, using the cleaning effect data as a direction to guide the model training, optimizing the strategy plan, and forming the robot arm optimized cleaning plan data.

[0013] In the present invention, based on the digital twin model, machine learning technology is used to realize intelligent identification and classification of locomotive bodies and components, and corresponding cleaning strategies and plans are generated and optimized according to different vehicle models, vehicle numbers, vehicle lengths, vehicle widths, vehicle heights, vehicle ages, vehicle conditions and other information.

[0014] As a possible implementation method, image data of different locomotive bodies and parts are obtained, and feature extraction and classification are performed to form locomotive body classification and recognition data, including: using a transfer learning method, selecting a pre-trained ResNet-50 model as the basic network, and fine-tuning and training according to a specific data set, wherein: during the model training process, the feature extraction capability of ResNet-50 is retained by freezing some layers of the ResNet-50, and fine-tuning is performed on the unfrozen layers to make the model better adapted to the locomotive recognition needs; training uses a labeled data set, and uses a loss function and an optimization algorithm to adjust the model parameters to improve recognition accuracy, and the generalization ability of the model is guaranteed through cross-validation and test set evaluation.

[0015] In the present invention, the intelligent recognition and classification scheme constructs a convolutional neural network (CNN) model through deep learning technology, which is used to extract and classify the images of locomotive bodies and parts. The transfer learning method is adopted, and the pre-trained ResNet-50 model is selected as the basic network, and fine-tuned and trained according to a specific data set. The data set contains images and labels of various types of locomotive bodies and parts, including high-speed rail, EMU, ordinary trains, as well as front, carriage, wheels, windows and other parts. During the model training process, the feature extraction capability of ResNet-50 is retained by freezing some layers of ResNet-50, and fine-tuning is performed on the thawed layers to make the model better adapted to the locomotive recognition needs. The training uses annotated data sets, and the model parameters are adjusted using loss functions and optimization algorithms to improve the recognition accuracy. The generalization ability of the model is guaranteed by cross-validation and test set evaluation. Through the digital twin development tool of the self-developed zero-code platform, the model is quickly deployed to the intelligent recognition system to realize real-time image acquisition and processing. The model extracts and classifies features on the collected images, and outputs corresponding labels to help operation and maintenance personnel identify different types of locomotive bodies and parts. At the same time, the system continuously accumulates data through a feedback mechanism, continuously optimizes model performance, and can realize fault identification and automatic warning functions. For example, it triggers warnings and maintenance tasks when identifying body damage or component wear, thereby improving the intelligence and automation level of locomotive maintenance.

[0016] As a possible implementation method, the strategy solution generation is carried out in the following way: a policy-based approach is adopted, and the Actor-Critic algorithm is used as the learning framework to learn and update according to the environment and reward function, where: the environment includes the state of the locomotive body and parts and the state of the cleaning robot; in the process of determining the cleaning strategy, the cleaning efficiency, cleaning quality and cleaning cost are comprehensively considered, and the reward function guides the model to generate high-quality cleaning strategies.

[0017] In the present invention, a deep reinforcement learning (DRL) model is constructed using reinforcement learning technology to generate and optimize the cleaning strategies and solutions for locomotive bodies and parts. A policy-based approach is adopted, and the Actor-Critic algorithm is used as a learning framework to learn and update according to the environment and reward function. The environment includes the state of the locomotive body and parts, such as the degree of dirt, surface temperature, surface material, etc., and the state of the cleaning robot, such as position, angle, speed, water pressure, water temperature, etc. In the process of determining the cleaning strategy, the cleaning efficiency, cleaning quality and cleaning cost are comprehensively considered, and the reward function guides the model to generate high-quality cleaning strategies. In terms of cleaning efficiency, the reward function gives rewards based on the coverage area or progress of cleaning to encourage the improvement of cleaning speed; in terms of cleaning quality, a target quality standard is set to score the thoroughness of dirt removal and surface cleanliness. If the target value is reached or close to it, a high reward is given; in terms of cleaning cost, factors such as dry ice, electricity, and time are taken into consideration. Operations that consume less resources but achieve the same cleanliness will receive higher rewards, thereby guiding the model to select a low-cost solution. During the model training process, through sampling and exploration, the model continuously tries different cleaning action combinations and gradually finds the best cleaning strategy. Using the zero-code digital twin development tool provided by the self-developed zero-code platform, the intelligent cleaning strategy and solution model can be quickly built and deployed to achieve real-time cleaning operations and feedback on the locomotive body and parts.

[0018] As a possible implementation method, locomotive cleaning image data is obtained, and the cleaning effect is evaluated and analyzed, and a cleaning effect evaluation model is established, including: based on the locomotive cleaning image data, a full convolutional network method is adopted, and a U-Net model is used as a segmentation network, and training and testing are performed according to the data set, wherein: the data set includes images of the locomotive body and parts before and after cleaning, and the corresponding segmentation labels.

[0019] In the present invention, computer vision technology is used to build an image segmentation model for detecting and evaluating the cleaning effect of the locomotive body and parts. The fully convolutional network (FCN) method is adopted, and the U-Net model is used as the segmentation network, and training and testing are performed according to the data set. The data set includes images of the locomotive body and parts before and after cleaning, and corresponding segmentation labels, such as clean areas, dirty areas, water stains, etc. Use the zero-code digital twin development tool provided by the self-developed zero-code platform to quickly build and deploy intelligent cleaning effect and evaluation models, and realize the analysis and display of real-time cleaning effects and evaluations of locomotive bodies and parts.

[0020] As a possible implementation method, locomotive cleaning operations are performed based on the robot arm's optimized cleaning plan data, and real-time cleaning data is collected for operation monitoring, maintenance and analysis to form real-time monitoring, maintenance and analysis result data, including: utilizing the Internet of Things technology to collect and transmit real-time operation data of the cleaning robot arm through various sensors and communication modules installed on the cleaning robot; utilizing big data technology to establish a database of historical data and fault data for the real-time operation data of the cleaning robot arm, and utilizing artificial intelligence technology to realize operation monitoring, maintenance and analysis of the cleaning robot arm to form real-time monitoring, maintenance and analysis result data.

[0021] In the present invention, the Internet of Things technology is used to realize the real-time collection and transmission of data such as the working status, operating parameters, and environmental information of the cleaning robot through various sensors and communication modules installed on the cleaning robot. Remote control and scheduling of the cleaning robot are realized. Through the zero-code digital twin development tool provided by the self-developed zero-code platform, the digital twin model of the cleaning robot is quickly built and deployed to realize real-time visualization and interaction of the cleaning robot. Using big data technology, a database of historical data and fault data is established for the cleaning robot to realize data mining and statistical analysis of the cleaning robot. Using artificial intelligence technology, a deep neural network (DNN) model is constructed to detect and identify faults of the cleaning robot. Using the transfer learning method, the pre-trained ResNet-50 model is used as the basic network, and fine-tuning and training are performed according to the data set. The data set includes fault data of different types of cleaning robots, as well as corresponding labels, such as motor failure, water pump failure, nozzle clogging, sensor failure, etc. Using the zero-code digital twin development tool provided by the self-developed zero-code platform, the intelligent diagnosis model is quickly built and deployed to realize real-time fault detection and identification of the cleaning robot. In addition, data analysis and visualization technology are used to provide comprehensive performance evaluation and optimization suggestions for cleaning robots. Using data analysis technology, the operation data of the cleaning robot is analyzed in multiple dimensions, such as cleaning efficiency, cleaning quality, cleaning cost, cleaning energy consumption, etc., and the key performance indicators (KPI) of the cleaning robot are calculated and compared with industry standards and similar robots. Using visualization technology, the performance evaluation results of the cleaning robot are intuitively displayed, such as line graphs, bar graphs, pie charts, dashboards, etc., and optimization suggestions and improvement measures for the cleaning robot are provided.

[0022] As a possible implementation method, intelligent operation analysis is performed based on real-time monitoring and maintenance analysis result data to form intelligent operation analysis result data, including: using data analysis and machine learning technology to establish an operation cost evaluation model for the cleaning robot arm based on the real-time monitoring and maintenance analysis result data, so as to realize the accounting and evaluation of the operation cost of the cleaning robot.

[0023] In the present invention, data analysis and machine learning techniques are used to establish an operation effect evaluation model for the cleaning robot, so as to quantify and evaluate the operation effect of the cleaning robot. Data analysis and machine learning techniques are used to establish an operation cost evaluation model for the cleaning robot, so as to calculate and evaluate the operation cost of the cleaning robot. Data analysis and machine learning techniques are used to establish an operation risk evaluation model for the cleaning robot, so as to identify and evaluate the operation risk of the cleaning robot. In addition, on this basis, an operation effect optimization model is established for the cleaning robot by using optimization algorithms and reinforcement learning techniques, so as to improve and optimize the operation effect of the cleaning robot. An operation cost optimization model is established for the cleaning robot by using optimization algorithms and reinforcement learning techniques, so as to reduce and optimize the operation cost of the cleaning robot. An operation risk optimization model is established for the cleaning robot by using optimization algorithms and reinforcement learning techniques, so as to control and optimize the operation risk of the cleaning robot. Visualization and virtual reality techniques are used to provide a visualization and interactive display of operation data for the cleaning robot, so that the operator can intuitively understand the operation status and operation effect of the cleaning robot. Utilize the zero-code digital twin development tools provided by the self-developed zero-code platform to quickly build and deploy intelligent display models to achieve real-time visualization and interaction of the cleaning robot.

[0024] In the second aspect, the present invention provides a robotic arm cleaning robot system based on zero-code digital twins, which adopts the robotic arm cleaning robot design method based on zero-code digital twins mentioned in the first aspect, including: an Internet of Things system for collecting Internet of Things data including locomotive body cleaning scenes, locomotive body identification and classification data, locomotive cleaning image data, and real-time cleaning data; a big data system for collecting big data information including a robotic arm data twin model; an analysis and processing system for using the robotic arm data twin model collected by the big data system to generate basic driving data for the robotic arm by following the data twin drive, and using the locomotive body identification and classification data collected by the Internet of Things system and combining the basic driving data of the robotic arm The IoT system uses real-time cleaning data to generate and optimize cleaning strategies and solutions, and forms data for optimizing cleaning solutions for robotic arms. The IoT system uses the locomotive cleaning image data collected to evaluate and analyze the cleaning effect, and establishes a cleaning effect evaluation model. The IoT system uses real-time cleaning data collected to perform operation monitoring and maintenance analysis, and forms real-time monitoring and maintenance analysis result data. The real-time monitoring and maintenance analysis result data is used to perform intelligent operation analysis, and forms intelligent operation analysis result data. The zero-code platform enables the establishment of a three-dimensional digital twin model of the locomotive object scene, as well as the rapid construction and deployment of the robotic arm optimization cleaning solution data, cleaning effect evaluation model, real-time monitoring and maintenance analysis result data, and intelligent operation analysis result data on the platform.

[0025] In the present invention, the system forms a tight integrated robotic arm cleaning robot design system through the Internet of Things system, big data system, analysis and processing system, and zero-code platform, which fully guarantees the real-time monitoring and intelligent cleaning of the locomotive body and parts, and improves the operation and maintenance efficiency and reliability of the cleaning robot. It is an important material basis for establishing a good robotic arm cleaning robot design system.

[0026] The beneficial effects of a mechanical arm cleaning robot system based on zero-code digital twin provided by the present invention are:

[0027] This method realizes real-time digital twin driving of locomotive bodies and components, improves the intelligence, adaptability and collaboration capabilities of cleaning robots, improves the quality and efficiency of cleaning services, and provides technical support for the innovative development of the cleaning industry.

[0028] The system forms a tight integrated robotic arm cleaning robot design system through the Internet of Things system, big data system, analysis and processing system, and zero-code platform, which fully ensures the real-time monitoring and intelligent cleaning of the locomotive body and components, and improves the operation and maintenance efficiency and reliability of the cleaning robot. It is an important material basis for establishing a good robotic arm cleaning robot design system. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments of the present invention are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0030] Figure 1 A step diagram of a robotic arm cleaning robot system based on zero-code digital twin provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present invention will be described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0032] With the development of industrialization and urbanization, the cleaning work of large facilities such as high-rise buildings, aircraft, and ships is increasing, but the traditional manual cleaning methods have problems such as low efficiency, high cost, and poor safety, which limit the quality and scale of cleaning services. According to statistics, the scale of the global cleaning service market reached 1.74 trillion US dollars in 2020, and is expected to grow to 2.41 trillion US dollars in 2027, with a compound annual growth rate of 4.8%. However, the cleaning industry is still facing challenges in technical bottlenecks and development trends, such as the complexity of the surface shape and structure of the cleaning object, the variability of the environmental conditions of the cleaning object, the differences in the cleaning requirements and standards of the cleaning object, and the complexity of the collaborative tasks of the cleaning object and the cleaning tools. Therefore, the development of an intelligent, efficient, and safe cleaning robot system has important engineering significance and social value.

[0033] Locomotives are a widely used means of transportation, and the cleaning of their bodies and parts is an important part of ensuring the performance and safety of locomotives. However, locomotive cleaning currently relies mainly on manual or track-based, vehicle-mounted, drone-based cleaning robots, which have some disadvantages, such as insufficient adaptability to the complex shapes and structures of locomotive bodies and parts, weak positioning and identification capabilities of locomotive bodies and parts, difficulty in achieving accurate and flexible cleaning actions, low ability to adjust cleaning parameters, poor ability to evaluate cleaning effects, low efficiency in the use of cleaning resources, low adaptability to the cleaning environment, and poor cleaning safety and reliability.

[0034] refer to Figure 1 The embodiment of the present invention provides a design method for a robotic arm cleaning robot based on zero-code digital twin. The method realizes real-time digital twin driving of the locomotive body and components, improves the intelligence, adaptability and collaboration capabilities of the cleaning robot, improves the quality and efficiency of cleaning services, and provides technical support for the innovative development of the cleaning industry.

[0035] The robotic arm cleaning robot system based on zero-code digital twin is configured as follows:

[0036] S1: Based on the locomotive body cleaning scenario, a three-dimensional digital twin model of the locomotive object scene for the cleaning object is established through the zero-code platform.

[0037] Based on the locomotive body cleaning scenario, a three-dimensional digital twin model of the locomotive cleaning scenario for the cleaning object is established through the zero-code platform, including: based on the locomotive body cleaning scenario, a digital twin model of the main equipment in the scene is established through the zero-code platform to form a three-dimensional digital twin model of the locomotive cleaning scene that matches the real cleaning scene; real-time data of the locomotive body and components is collected, and data twin models of different locomotive bodies and components are established using the zero-code platform to form a three-dimensional data twin model of the locomotive body.

[0038] Taking the locomotive body cleaning scenario as a specific application scenario, the self-developed zero-code platform and zero-code technology are used to quickly build the digital twin model of the locomotive body and locomotive cleaning scenario, as well as the main equipment in the scenario, such as the cleaning tray with components, dry ice tank, ice maker, AGV trolley, AGV parking space, AGV charging pile, etc. Among them, the zero-code digital twin development platform allows the digital twin application to be quickly built and delivered by dragging and dropping without writing any code, which greatly reduces the threshold and difficulty of development.

[0039] S2: Collect the digital twin model of the cleaning robot arm, and combine it with the three-dimensional data twin model of the locomotive object scene to perform digital twin driving to form the basic driving data of the robot arm.

[0040] The digital twin model of the cleaning robot arm is collected, and combined with the three-dimensional data twin model of the locomotive object, digital twin drive is performed to form the basic driving data of the robot arm, including: combining plug-in technology, Internet of Things technology and big data technology for locomotive cleaning, according to the three-dimensional data twin model of the locomotive cleaning scene and the three-dimensional data twin model of the locomotive body, to determine the basic driving information of the cleaning robot and form the basic driving data of the robot arm.

[0041] Based on the digital twin model, plug-in technology, Internet of Things technology, and big data technology are used to realize real-time digital twin drive of the cleaning robot's mechanical arm, as well as coordinated control with the AGV car, to achieve real-time cleaning operation and feedback of the locomotive body. Through rich industry templates and resources, the basic parameters of the digital twin model can be set according to the actual parameters and models of the locomotive cleaning mechanical arm, such as joint sensitivity, joint range of motion, upper limits of linear velocity and acceleration of each joint, motion data compensation, etc., to realize the digital twin drive of the locomotive cleaning mechanical arm.

[0042] S3: Obtain the locomotive body identification and classification data, and combine it with the basic driving data of the robot arm to perform generation and optimization analysis of the cleaning strategy and plan to form the robot arm optimized cleaning plan data.

[0043] The locomotive body identification and classification data is obtained, and the generation and optimization analysis of cleaning strategies and plans is performed in combination with the basic driving data of the robot arm to form the robot arm optimized cleaning plan data, including: obtaining image data of different locomotive bodies and parts, and performing feature extraction and classification to form locomotive body classification and identification data; collecting cleaning action control data, and combining the locomotive body classification and identification data and the basic driving data of the robot arm to generate a strategy plan based on model training, and obtaining cleaning effect data during the model training process, using the cleaning effect data as a direction to guide the model training, optimizing the strategy plan, and forming the robot arm optimized cleaning plan data.

[0044] Based on the digital twin model and using machine learning technology, intelligent identification and classification of locomotive bodies and components can be achieved, and corresponding cleaning strategies and plans can be generated and optimized based on different vehicle models, vehicle numbers, vehicle lengths, vehicle widths, vehicle heights, vehicle ages, vehicle conditions and other information.

[0045] Among them, image data of different locomotive bodies and parts are obtained, and feature extraction and classification are performed to form locomotive body classification and recognition data, including: using a transfer learning method, selecting a pre-trained ResNet-50 model as the basic network, and fine-tuning and training according to a specific data set, wherein: during the model training process, the feature extraction capability of ResNet-50 is retained by freezing some layers of the ResNet-50, and fine-tuning is performed on the unfrozen layers to make the model better adapted to the locomotive recognition needs; training uses annotated data sets, and uses loss functions and optimization algorithms to adjust model parameters to improve recognition accuracy, and cross-validation and test set evaluation are used to ensure the generalization ability of the model.

[0046] The intelligent recognition and classification solution uses deep learning technology to build a convolutional neural network (CNN) model for feature extraction and classification of locomotive body and component images. Using the transfer learning method, the pre-trained ResNet-50 model is selected as the basic network, and fine-tuned and trained according to a specific data set. The data set contains images and labels of various types of locomotive bodies and components, including high-speed trains, EMUs, ordinary trains, as well as locomotive heads, carriages, wheels, windows and other components. During the model training process, the feature extraction capability of ResNet-50 is retained by freezing some layers, and fine-tuning is performed on the unfrozen layers to make the model better adapted to the needs of locomotive recognition. Training uses annotated data sets, and uses loss functions and optimization algorithms to adjust model parameters to improve recognition accuracy. Cross-validation and test set evaluation ensure the generalization ability of the model. Through the digital twin development tool of the self-developed zero-code platform, the model is quickly deployed to the intelligent recognition system to achieve real-time image acquisition and processing. The model extracts and classifies features on the acquired images and outputs corresponding labels to help operation and maintenance personnel identify different types of locomotive bodies and components. At the same time, the system continuously accumulates data through a feedback mechanism, continuously optimizes model performance, and can realize fault identification and automatic warning functions. For example, it triggers warnings and maintenance tasks when identifying body damage or component wear, thereby improving the intelligence and automation level of locomotive maintenance.

[0047] The generation of strategy solutions is carried out in the following way: a strategy-based approach is adopted, the Actor-Critic algorithm is used as the learning framework, and learning and updating are performed according to the environment and reward function, where: the environment includes the state of the locomotive body and parts and the state of the cleaning robot; in the process of determining the cleaning strategy, the cleaning efficiency, cleaning quality and cleaning cost are comprehensively considered, and the reward function guides the model to generate high-quality cleaning strategies.

[0048] Using reinforcement learning technology, a deep reinforcement learning (DRL) model is constructed to generate and optimize the cleaning strategies and solutions for locomotive bodies and parts. A policy-based approach is adopted, using the Actor-Critic algorithm as a learning framework, and learning and updating are performed according to the environment and reward function. The environment includes the status of the locomotive body and parts, such as the degree of dirt, surface temperature, surface material, etc., as well as the status of the cleaning robot, such as position, angle, speed, water pressure, water temperature, etc. In the process of determining the cleaning strategy, the cleaning efficiency, cleaning quality and cleaning cost are comprehensively considered, and the reward function guides the model to generate high-quality cleaning strategies. In terms of cleaning efficiency, the reward function gives rewards based on the coverage area or progress of cleaning to encourage the improvement of cleaning speed; in terms of cleaning quality, a target quality standard is set to score the thoroughness of dirt removal and surface cleanliness. If the target value is reached or close to it, a high reward is given; in terms of cleaning cost, factors such as dry ice, electricity, and time are taken into consideration. Operations that consume less resources but achieve the same cleanliness will receive higher rewards, thereby guiding the model to choose a low-cost solution. During the model training process, through sampling and exploration, the model continuously tries different cleaning action combinations and gradually finds the best cleaning strategy. Using the zero-code digital twin development tool provided by the self-developed zero-code platform, the intelligent cleaning strategy and solution model can be quickly built and deployed to achieve real-time cleaning operations and feedback on the locomotive body and parts.

[0049] S4: Obtain locomotive cleaning image data, conduct evaluation and analysis of the cleaning effect, and establish a cleaning effect evaluation model.

[0050] Obtain locomotive cleaning image data, conduct cleaning effect evaluation and analysis, and establish a cleaning effect evaluation model, including: based on the locomotive cleaning image data, adopting a full convolutional network method, using a U-Net model as a segmentation network, and training and testing according to the data set, wherein: the data set includes images of the locomotive body and parts before and after cleaning, and corresponding segmentation labels.

[0051] Using computer vision technology, an image segmentation model is constructed to detect and evaluate the cleaning effect of locomotive bodies and parts. The fully convolutional network (FCN) method is adopted, and the U-Net model is used as the segmentation network, and training and testing are performed according to the data set. The data set includes images of locomotive bodies and parts before and after cleaning, as well as corresponding segmentation labels, such as clean areas, dirty areas, water stains, etc. Using the zero-code digital twin development tool provided by the self-developed zero-code platform, intelligent cleaning effect and evaluation models are quickly built and deployed to realize the analysis and display of real-time cleaning effects and evaluations of locomotive bodies and parts.

[0052] S5: Perform locomotive cleaning operations based on the robot arm's optimized cleaning solution data, and collect real-time cleaning data for operation monitoring, maintenance and analysis to form real-time monitoring, maintenance and analysis result data.

[0053] Locomotive cleaning operations are performed based on the robot arm's optimized cleaning plan data, and real-time cleaning data is collected for operation monitoring, maintenance and analysis to form real-time monitoring, maintenance and analysis result data, including: utilizing the Internet of Things technology to collect and transmit real-time operation data of the cleaning robot arm through various sensors and communication modules installed on the cleaning robot; utilizing big data technology to establish a database of historical data and fault data for the real-time operation data of the cleaning robot arm, and utilizing artificial intelligence technology to realize operation monitoring, maintenance and analysis of the cleaning robot arm to form real-time monitoring, maintenance and analysis result data.

[0054] Using the Internet of Things technology, various sensors and communication modules installed on the cleaning robot can be used to collect and transmit the working status, operating parameters, environmental information and other data of the cleaning robot in real time. Remote control and scheduling of the cleaning robot can be realized. The zero-code digital twin development tool provided by the self-developed zero-code platform can be used to quickly build and deploy the digital twin model of the cleaning robot to realize real-time visualization and interaction of the cleaning robot. Using big data technology, a database of historical data and fault data is established for the cleaning robot to realize data mining and statistical analysis of the cleaning robot. Using artificial intelligence technology, a deep neural network (DNN) model is constructed to detect and identify faults of the cleaning robot. Using the transfer learning method, the pre-trained ResNet-50 model is used as the basic network, and fine-tuned and trained according to the data set. The data set includes fault data of different types of cleaning robots, as well as corresponding labels, such as motor failure, water pump failure, nozzle blockage, sensor failure, etc. Using the zero-code digital twin development tool provided by the self-developed zero-code platform, the intelligent diagnosis model is quickly built and deployed to realize real-time fault detection and identification of the cleaning robot. In addition, data analysis and visualization technology are used to provide comprehensive performance evaluation and optimization suggestions for the cleaning robot. Using data analysis technology, we can analyze the operation data of the cleaning robot in multiple dimensions, such as cleaning efficiency, cleaning quality, cleaning cost, cleaning energy consumption, etc., calculate the key performance indicators (KPI) of the cleaning robot, and compare them with industry standards and similar robots. Using visualization technology, we can intuitively display the performance evaluation results of the cleaning robot, such as line graphs, bar graphs, pie charts, dashboards, etc., and provide optimization suggestions and improvement measures for the cleaning robot.

[0055] S6: Perform intelligent operation analysis based on the real-time monitoring maintenance analysis result data to form intelligent operation analysis result data.

[0056] Intelligent operation analysis is performed based on the real-time monitoring and maintenance analysis result data to form intelligent operation analysis result data, including: using data analysis and machine learning technology to establish an operation cost evaluation model for the cleaning robot arm based on the real-time monitoring and maintenance analysis result data, so as to realize the accounting and evaluation of the operation cost of the cleaning robot.

[0057] Using data analysis and machine learning technology, an operation effect evaluation model is established for the cleaning robot to quantify and evaluate the operation effect of the cleaning robot. Using data analysis and machine learning technology, an operation cost evaluation model is established for the cleaning robot to calculate and evaluate the operation cost of the cleaning robot. Using data analysis and machine learning technology, an operation risk evaluation model is established for the cleaning robot to identify and evaluate the operation risk of the cleaning robot. In addition, on this basis, using optimization algorithms and reinforcement learning technology, an operation effect optimization model is established for the cleaning robot to improve and optimize the operation effect of the cleaning robot. Using optimization algorithms and reinforcement learning technology, an operation cost optimization model is established for the cleaning robot to reduce and optimize the operation cost of the cleaning robot. Using optimization algorithms and reinforcement learning technology, an operation risk optimization model is established for the cleaning robot to control and optimize the operation risk of the cleaning robot. Using visualization and virtual reality technology, a visualization and interactive display of operation data is provided for the cleaning robot, so that the operator can intuitively understand the operation status and operation effect of the cleaning robot. Utilize the zero-code digital twin development tools provided by the self-developed zero-code platform to quickly build and deploy intelligent display models to achieve real-time visualization and interaction of the cleaning robot.

[0058] The present invention also provides a mechanical arm cleaning robot system based on zero-code digital twins, which adopts the above-mentioned mechanical arm cleaning robot design method based on zero-code digital twins, including: an Internet of Things system for collecting Internet of Things data including locomotive body cleaning scenes, locomotive body identification and classification data, locomotive cleaning image data, and real-time cleaning data; a big data system for collecting big data information including a mechanical arm data twin model; an analysis and processing system for using the mechanical arm data twin model collected by the big data system to generate mechanical arm basic drive data by following the data twin drive, and using the locomotive body identification and classification data collected by the Internet of Things system and combining it with the mechanical arm basic drive data to perform Perform cleaning strategy and solution generation optimization analysis to form robot arm optimization cleaning solution data, use the locomotive cleaning image data collected by the Internet of Things system, and evaluate and analyze the cleaning effect, establish a cleaning effect evaluation model, use the real-time cleaning data collected by the Internet of Things system to perform operation monitoring and maintenance analysis to form real-time monitoring and maintenance analysis result data, and use the real-time monitoring and maintenance analysis result data to perform intelligent operation analysis to form intelligent operation analysis result data; zero-code platform to achieve the establishment of a three-dimensional digital twin model of the locomotive object scene, and complete the rapid construction and deployment of robot arm optimization cleaning solution data, cleaning effect evaluation model, real-time monitoring and maintenance analysis result data and intelligent operation analysis result data on the platform.

[0059] The system forms a tight integrated robotic arm cleaning robot design system through the Internet of Things system, big data system, analysis and processing system, and zero-code platform, which fully ensures the real-time monitoring and intelligent cleaning of the locomotive body and components, and improves the operation and maintenance efficiency and reliability of the cleaning robot. It is an important material basis for establishing a good robotic arm cleaning robot design system.

[0060] In summary, the beneficial effects of the mechanical arm cleaning robot system based on zero-code digital twin provided by the embodiment of the present invention are:

[0061] This method realizes real-time digital twin driving of locomotive bodies and components, improves the intelligence, adaptability and collaboration capabilities of cleaning robots, improves the quality and efficiency of cleaning services, and provides technical support for the innovative development of the cleaning industry.

[0062] The system forms a tight integrated robotic arm cleaning robot design system through the Internet of Things system, big data system, analysis and processing system, and zero-code platform, which fully ensures the real-time monitoring and intelligent cleaning of the locomotive body and components, and improves the operation and maintenance efficiency and reliability of the cleaning robot. It is an important material basis for establishing a good robotic arm cleaning robot design system.

[0063] In the embodiments of the present application, "indication" may include direct indication and indirect indication, and may also include explicit indication and implicit indication. The information indicated by a certain information is called information to be indicated. In the specific implementation process, there are many ways to indicate the information to be indicated, such as but not limited to, the information to be indicated can be directly indicated, such as the information to be indicated itself or the index of the information to be indicated. The information to be indicated can also be indirectly indicated by indicating other information, wherein there is an association between the other information and the information to be indicated. It is also possible to indicate only a part of the information to be indicated, while the other parts of the information to be indicated are known or agreed in advance. For example, the indication of specific information can also be achieved by means of the arrangement order of each piece of information agreed in advance (for example, specified by the protocol), thereby reducing the indication overhead to a certain extent. At the same time, the common parts of each piece of information can also be identified and indicated uniformly to reduce the indication overhead caused by indicating the same information separately.

[0064] In addition, the specific indication method may also be various existing indication methods, such as but not limited to the above-mentioned indication methods and various combinations thereof. The specific details of the various indication methods can refer to the prior art and will not be repeated herein. As can be seen from the above, for example, when it is necessary to indicate multiple information of the same type, different indication methods may be used for different information. In the specific implementation process, the desired indication method can be selected according to specific needs. The embodiment of the present application does not limit the selected indication method. In this way, the indication method involved in the embodiment of the present application should be understood to cover various methods that can enable the party to be indicated to obtain the information to be indicated.

[0065] It should be understood that the information to be indicated can be sent as a whole, or divided into multiple sub-information and sent separately, and the sending period and / or sending time of these sub-information can be the same or different. The specific sending method is not limited in the embodiment of the present application. Among them, the sending period and / or sending time of these sub-information can be pre-defined, for example, pre-defined according to a protocol, or can be configured by the sending end device by sending configuration information to the receiving end device.

[0066] "Pre-definition" or "pre-configuration" can be implemented by pre-saving corresponding codes, tables or other methods that can be used to indicate relevant information in the device, and the embodiments of the present application do not limit the specific implementation method. Among them, "saving" can mean saving in one or more memories. The one or more memories can be set separately or integrated in an encoder or decoder, a processor, or a communication device. The one or more memories can also be partially set separately and partially integrated in a decoder, a processor, or a communication device. The type of memory can be any form of storage medium, which is not limited by the embodiments of the present application.

[0067] The "protocol" involved in the embodiments of the present application may refer to a protocol family in the communication field, a standard protocol with a similar protocol family frame structure, or a related protocol used in future communication systems, and the embodiments of the present application do not make specific limitations on this.

[0068] In the embodiments of the present application, descriptions such as "when...", "in the case of...", "if" and "if" all mean that the device will make corresponding processing under certain objective circumstances. It does not limit the time, nor does it require the device to have a judgment action when implementing it, nor does it mean that there are other limitations.

[0069] In the description of the embodiments of the present application, unless otherwise specified, " / " indicates that the objects associated before and after are in an "or" relationship, for example, A / B can represent A or B; "and / or" in the embodiments of the present application is only a description of the association relationship of the associated objects, indicating that there can be three relationships, for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. In addition, in the description of the embodiments of the present application, unless otherwise specified, "multiple" refers to two or more than two. "At least one of the following" or its similar expression refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple. In addition, in order to facilitate the clear description of the technical solution of the embodiments of the present application, in the embodiments of the present application, the words "first" and "second" are used to distinguish the same or similar items with basically the same functions and effects. Those skilled in the art will appreciate that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit the difference. At the same time, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a concrete way for easy understanding.

[0070] It should be understood that the processor in the embodiments of the present application may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0071] It should also be understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0072] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0073] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0074] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0075] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0076] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0077] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0078] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0079] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0080] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0081] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0082] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A design method for a robotic arm cleaning robot based on zero-code digital twin, characterized in that: include: Based on the locomotive body cleaning scenario, a three-dimensional digital twin model of the locomotive object scenario for the cleaning object is established through the zero-code platform; Collecting a digital twin model of the cleaning robot arm, and combining it with the three-dimensional data twin model of the locomotive object scene, performing digital twin driving to form basic driving data of the robot arm; Acquire the locomotive body identification classification data, and combine the basic driving data of the mechanical arm to perform generation optimization analysis of the cleaning strategy and plan, and form the mechanical arm optimization cleaning plan data; Obtain locomotive cleaning image data, conduct cleaning effect evaluation and analysis, and establish a cleaning effect evaluation model; Based on the robot arm optimized cleaning scheme data, the locomotive cleaning operation is performed, and real-time cleaning data is collected for operation monitoring, maintenance and analysis to form real-time monitoring, maintenance and analysis result data; Intelligent operation analysis is performed based on the real-time monitoring maintenance analysis result data to form intelligent operation analysis result data.

2. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 1 is characterized in that: Based on the locomotive body cleaning scene, a three-dimensional digital twin model of the locomotive cleaning scene for the cleaning object is established through the zero-code platform, including: Based on the locomotive body cleaning scene, a digital twin model of the main equipment in the scene is established through a zero-code platform to form a three-dimensional digital twin model of the locomotive cleaning scene that matches the real cleaning scene; Collect real-time data of locomotive bodies and components, and use the zero-code platform to establish data twin models of different locomotive bodies and components to form a three-dimensional data twin model of the locomotive body.

3. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 2 is characterized in that: The digital twin model of the robot arm of the cleaning robot arm is collected, and combined with the three-dimensional data twin model of the locomotive object, digital twin driving is performed to form basic driving data of the robot arm, including: Combining plug-in technology, Internet of Things technology and big data technology for locomotive cleaning, the basic driving information of the cleaning robot is determined according to the three-dimensional data twin model of the locomotive cleaning scene and the three-dimensional data twin model of the locomotive body to form the basic driving data of the robotic arm.

4. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 3 is characterized in that: The acquisition of the locomotive body identification classification data and the generation optimization analysis of the cleaning strategy and scheme in combination with the basic driving data of the mechanical arm to form the mechanical arm optimization cleaning scheme data include: Obtain image data of different locomotive bodies and parts, and perform feature extraction and classification to form locomotive body classification and recognition data; The cleaning action control data is collected, and combined with the locomotive body classification and identification data and the basic driving data of the robotic arm to generate a strategy plan based on model training, and the cleaning effect data is obtained during the model training process. The cleaning effect data is used as a guide for the direction of model training, and the strategy plan is optimized to form the robotic arm optimized cleaning plan data.

5. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 4 is characterized in that: The image data of different locomotive bodies and parts are obtained, and feature extraction and classification are performed to form locomotive body classification and recognition data, including: Using the transfer learning method, the pre-trained ResNet-50 model is selected as the base network, and fine-tuned and trained according to a specific dataset, where: During the model training process, some layers of ResNet-50 are frozen to retain its feature extraction capability, and fine-tuning is performed on the unfrozen layers to make the model better suited to the needs of locomotive recognition; The training uses labeled data sets, and the loss function and optimization algorithm are used to adjust the model parameters to improve the recognition accuracy. The generalization ability of the model is guaranteed through cross-validation and test set evaluation.

6. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 4 is characterized in that: The generation of the policy solution is done in the following way: A policy-based approach is adopted, using the Actor-Critic algorithm as the learning framework to learn and update according to the environment and reward function, where: The environment includes the status of the locomotive body and components and the status of the cleaning robot; In the process of determining the cleaning strategy, the cleaning efficiency, cleaning quality and cleaning cost are comprehensively considered, and the reward function guides the model to generate high-quality cleaning strategies.

7. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 4 is characterized in that: The obtaining of locomotive cleaning image data, and performing cleaning effect evaluation and analysis, and establishing a cleaning effect evaluation model, includes: According to the locomotive cleaning image data, a full convolutional network method is adopted, and the U-Net model is used as the segmentation network, and training and testing are performed according to the data set, wherein: The dataset includes images of locomotive bodies and parts before and after cleaning, as well as the corresponding segmentation labels.

8. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 7 is characterized in that: Based on the robot arm optimized cleaning scheme data, the locomotive cleaning operation is performed, and real-time cleaning data is collected for operation monitoring, maintenance and analysis to form real-time monitoring, maintenance and analysis result data, including: Using the Internet of Things technology, various sensors and communication modules installed on the cleaning robot can be used to collect and transmit real-time operation data of the cleaning robot arm. By using big data technology, a database of historical data and fault data is established for the real-time operation data of the cleaning robot arm, and artificial intelligence technology is used to realize the operation monitoring, maintenance and analysis of the cleaning robot arm to form the real-time monitoring, maintenance and analysis result data.

9. The design method of a mechanical arm cleaning robot based on zero-code digital twin according to claim 8 is characterized in that: The performing intelligent operation analysis according to the real-time monitoring maintenance analysis result data to form intelligent operation analysis result data includes: By using data analysis and machine learning technology, an operating cost evaluation model is established for the cleaning robot arm according to the real-time monitoring, maintenance and analysis result data, so as to realize the accounting and evaluation of the operating cost of the cleaning robot.

10. A robotic arm cleaning robot system based on zero-code digital twin, adopting the robotic arm cleaning robot design method based on zero-code digital twin according to any one of claims 1 to 9, characterized in that: include: The IoT system is used to collect IoT data including locomotive body cleaning scenes, locomotive body identification and classification data, locomotive cleaning image data, and real-time cleaning data; Big data system, used to collect big data information including the robotic arm data twin model; An analysis and processing system is used to use the robot data twin model collected by the big data system to generate basic driving data of the robot arm by following the data twin drive, use the locomotive body identification and classification data collected by the Internet of Things system and combine it with the basic driving data of the robot arm to perform cleaning strategy and solution generation optimization analysis to form robot arm optimization cleaning solution data, use the locomotive cleaning image data collected by the Internet of Things system, and perform cleaning effect evaluation analysis to establish a cleaning effect evaluation model, use the real-time cleaning data collected by the Internet of Things system to perform operation monitoring and maintenance analysis to form real-time monitoring and maintenance analysis result data, and use the real-time monitoring and maintenance analysis result data to perform intelligent operation analysis to form intelligent operation analysis result data; The zero-code platform enables the establishment of a three-dimensional digital twin model of the locomotive object scene, as well as the rapid construction and deployment of robotic arm optimization cleaning solution data, cleaning effect evaluation model, real-time monitoring and maintenance analysis result data, and intelligent operation analysis result data on the platform.

Images