Intelligent carriage cleaning method

Abstract

The invention provides an intelligent carriage cleaning method which comprises the following steps: matching and selecting a proper cleaning tool from a predefined tool library according to characteristics and material attributes of a cleaning area, and realizing automatic assembly of the tool through a mechanical arm to improve the cleaning adaptability; aiming at each cleaning area, learning and optimizing a motion track and operation parameters of a cleaning tool through a reinforcement learning algorithm in combination with a carriage internal environment constraint condition and a cleaning efficiency target, and formulating a cleaning path; in the cleaning process, the distribution change condition of residual materials is obtained in real time through a visual sensor, a material distribution diagram is dynamically updated through target detection and a semantic segmentation algorithm, and a cleaning path is adjusted in real time according to the updated distribution diagram; for the materials with high adhesion degree, the contact force between the cleaning tool and the materials is monitored in real time through a force feedback sensor, and the pressure and angle of the cleaning tool are dynamically adjusted in combination with a pre-established material characteristic database until the materials are cleaned.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to an intelligent carriage cleaning method. Background Art

[0002] In the complex interior environment of a carriage, the distribution and adhesion of residual materials vary greatly, posing a severe challenge to the adaptive ability of cleaning equipment. The interior structure of the carriage is complex and changeable, with a large number of grooves, dead corners and narrow spaces, making it difficult for cleaning tools to reach every residual material. In addition, the degree of adhesion of materials varies greatly, and some materials are firmly adsorbed on the inner wall of the carriage, making them difficult to clean. Cleaning equipment needs to be able to flexibly adjust the angle, pressure and movement trajectory of the cleaning tool, and adopt differentiated cleaning strategies based on the material characteristics of different areas to achieve the goal of thorough cleaning in the shortest possible time. This requires the cleaning equipment to have sophisticated environmental perception and decision-making planning capabilities, and to be able to analyze the material distribution map in real time and develop the optimal cleaning path. At the same time, it is also necessary to develop a set of diverse cleaning tools and realize rapid tool replacement to adapt to the complex and changeable interior environment of the carriage. Summary of the invention

[0003] The present invention provides an intelligent carriage cleaning method, which mainly includes: The three-dimensional point cloud data of the interior of the carriage is obtained through laser radar scanning, and the semantic segmentation of the point cloud data is performed in combination with a deep learning algorithm to identify the complex structure of the interior of the carriage and the distribution of residual materials, and obtain the interior environment information model of the carriage; according to the interior environment information model of the carriage, a rule-based reasoning algorithm is used to analyze the location distribution characteristics of the materials, and the adhesion degree of the materials is detected by a spectral analysis sensor, the interior of the carriage is divided into multiple cleaning areas, and the cleaning priority of each area is determined; according to the characteristics of the cleaning area and the properties of the materials, suitable cleaning tools are matched and selected from the predefined tool library, and the automatic assembly of the tools is realized through the robotic arm to improve the cleaning adaptability; for each cleaning area, a reinforcement learning algorithm is used to learn and optimize the cleaning process in combination with the internal environment constraints of the carriage and the cleaning efficiency goals. The motion trajectory and operating parameters of the sweeping tool are used to formulate a cleaning path; during the cleaning process, the visual sensor is used to obtain the distribution changes of the residual materials in real time, and the material distribution map is dynamically updated through target detection and semantic segmentation algorithms, and the cleaning path is adjusted in real time according to the updated distribution map; for materials with high adhesion, the contact force between the cleaning tool and the material is monitored in real time through a force feedback sensor, and the pressure and angle of the cleaning tool are dynamically adjusted in combination with the pre-established material property database until the material is cleaned; during the cleaning operation, the cleaning data under different carriage environments and material properties are recorded and accumulated, and the cleaning path planning and tool selection strategies are continuously optimized through convolutional neural networks and decision tree algorithms, so as to realize the self-learning and evolution of the cleaning equipment and improve its adaptive ability and cleaning effect.

[0004] The technical solution provided by the embodiment of the present invention may have the following beneficial effects: The present invention discloses an intelligent carriage cleaning method. The method obtains three-dimensional point cloud data inside the carriage through laser radar scanning, combines deep learning algorithms for semantic segmentation, identifies the internal structure of the carriage and the distribution of residual materials, and constructs an environmental information model. Based on the model, rule reasoning and spectral analysis are used to divide the cleaning area and determine the priority. According to the characteristics of the area, the appropriate cleaning tools are automatically selected and assembled. The cleaning path is optimized by using a reinforcement learning algorithm combined with environmental constraints and efficiency goals. During the cleaning process, the material distribution map is updated in real time and the path is adjusted dynamically. For adhered materials, the parameters of the cleaning tool are adjusted by a force feedback sensor. The present invention also improves the adaptive ability of the cleaning equipment through continuous learning and optimization, thereby realizing intelligent and efficient carriage cleaning. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] Figure 1 The present invention is a flow chart of an intelligent carriage cleaning method.

[0006] Figure 2 It is a schematic diagram of an intelligent carriage cleaning method of the present invention.

[0007] Figure 3 It is another schematic diagram of an intelligent carriage cleaning method of the present invention. DETAILED DESCRIPTION

[0008] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0009] like Figure 1-3 In this embodiment, an intelligent carriage cleaning method may specifically include: S101. Obtain three-dimensional point cloud data of the interior of the carriage through laser radar scanning, and perform semantic segmentation on the point cloud data in combination with a deep learning algorithm to identify the complex structure of the interior of the carriage and the distribution of residual materials, and obtain an internal environment information model of the carriage.

[0010] The original point cloud data set of the interior of the carriage is obtained, and the statistical filter in the point cloud library PCL is used to remove the noise points in the point cloud. The point cloud is downsampled using the voxelized grid method to obtain the preprocessed point cloud data set. The preprocessed point cloud data is input into the PointNet++ network. Through the semantic segmentation of the point cloud, each point is divided into two categories: the internal structure of the carriage and the residual material, and the internal structure point cloud subset and the residual material point cloud subset of the carriage are obtained. For the internal structure point cloud subset of the carriage, the regional growing algorithm is used for geometric segmentation. According to the growth criteria of the normal vector angle threshold and the curvature threshold, the point cloud is segmented into different structural components, and each component point cloud is Euclidean clustering is performed to obtain independent structural components. For the residual material point cloud subset, the number of neighboring points of each point is calculated to obtain the local density distribution of the point cloud; the point cloud is projected to the Z axis direction, the number of points in different height intervals is counted, and the height distribution histogram is generated; according to the density distribution peak area and the height histogram peak position, the accumulation position and quantity of residual materials are determined. The geometric segmentation results of the internal structure of the carriage are spatially registered with the distribution of residual materials, and a three-dimensional model of the interior environment of the carriage containing structural components and residual material information is constructed. Based on the constructed interior environment model of the carriage, combined with the distribution of residual materials and structural layout, the motion path of the cleaning robot is optimized to generate a cleaning trajectory that covers the residual material area and avoids structural obstacles; the cleaning tool is selected and the cleaning parameters are adjusted. The point cloud of the interior of the carriage after cleaning is obtained, and by comparing it with the original point cloud, the change in the volume of residual materials before and after cleaning is quantified, and the cleaning effect is evaluated. If the expected cleanliness is not achieved, the cleaning strategy is adjusted until the requirements are met.

[0011] Specifically, the interior of the carriage is scanned by a laser radar to obtain the three-dimensional point cloud data of the interior of the carriage and obtain the original point cloud data set. The statistical filter StatisticalOutlierRemoval in the point cloud library PCL is used to remove noise points, and the voxel grid method VoxelGrid is used for downsampling to obtain the preprocessed point cloud data set. The PointNet++ network is applied to the preprocessed point cloud data for point cloud semantic segmentation, and each point is divided into two categories: the internal structure of the carriage and the residual material. Based on the semantic segmentation results, the internal structure point cloud subset and the residual material point cloud subset are extracted. The internal structure point cloud subset of the carriage is geometrically segmented, and the region growing algorithm is used. The normal vector angle threshold and the curvature threshold are set as the growth criteria to segment the point cloud into different structural components, and each component point cloud is Euclidean clustering is performed to obtain independent structural components. For the residual material point cloud subset, the number of neighboring points of each point is calculated to obtain the local density distribution of the point cloud. The point cloud is projected to the Z axis direction, the number of points in different height intervals is counted, and a height distribution histogram is generated. According to the peak area of ​​the density distribution and the peak position of the height histogram, the accumulation position and quantity of the residual materials are determined. The geometric segmentation results of the internal structure of the carriage are spatially registered with the distribution of the residual materials, and a three-dimensional model of the internal environment of the carriage containing the structural components and residual materials is constructed. The Open3D library is used to realize point cloud registration and three-dimensional visualization. According to the model of the internal environment of the carriage, the automatic cleaning operation is planned. Considering the distribution of residual materials and the structural layout, the movement path of the cleaning robot is optimized to generate a trajectory that covers the residual material area and avoids structural obstacles. Select appropriate cleaning tools such as high-pressure water guns or brushes, and adjust parameters such as cleaning pressure and flow. Use the three-dimensional model to monitor the cleaning process in real time, obtain the point cloud of the interior of the carriage after cleaning, and quantify the volume change of the residual materials before and after cleaning by comparing with the original point cloud, and evaluate the cleaning effect. If the expected cleanliness is not achieved, the cleaning strategy is adjusted, such as increasing the number of cleaning times, changing the cleaning tools, etc., until the requirements are met. Through the above steps, a complete closed-loop solution from three-dimensional scanning of the interior of the carriage to automatic cleaning operations is realized. The analysis and modeling of point cloud data provide data support for cleaning decisions, optimize the cleaning path and parameter selection, and improve the cleaning efficiency and quality of the carriage. The 3D model can also be used for evaluation and feedback of cleaning effects, forming an information closed loop.

[0012] S102. According to the interior environment information model of the carriage, a rule-based reasoning algorithm is used to analyze the location distribution characteristics of the materials, and the adhesion degree of the materials is detected by a spectral analysis sensor, the interior of the carriage is divided into multiple cleaning areas, and the cleaning priority of each area is determined.

[0013] The environmental information model of the interior of the carriage is obtained, and the decision tree algorithm is used to analyze the environmental information model to obtain a material distribution thermodynamic map; the material adhesion information is obtained, and the material adhesion degree is determined according to a preset adhesion degree threshold to obtain a material adhesion distribution map; the material distribution thermodynamic map and the adhesion distribution map are data fused to extract the material distribution density characteristics and adhesion degree characteristics in the area; the K-means clustering algorithm is used to cluster the fused feature data, and the optimal number of clusters is determined according to the silhouette coefficient to obtain multiple cleaning areas inside the carriage; for each of the cleaning areas, the mean value of the material distribution density and the mean value of the adhesion degree in the area are calculated respectively, and the two means are normalized to obtain The invention provides a distribution density score and an adhesion degree score for each area; sets a distribution density score weight and an adhesion degree score weight, calculates a comprehensive score for each area by weight, and determines a cleaning priority level for each cleaning area according to the comprehensive score; determines a cleaning order for each cleaning area according to the cleaning priority level from high to low; inputs the cleaning order and the position coordinate information of each cleaning area into the cleaning robot; the cleaning robot cleans each cleaning area in order from high to low priority according to the input cleaning order and position coordinates; during the cleaning process, updates the vehicle environment information model in real time, and completes the cleaning task when all areas are cleaned.

[0014] Specifically, according to the environmental information model inside the carriage, the decision tree algorithm is used to analyze the distribution characteristics of materials inside the carriage to obtain the material distribution thermodynamic map. The material adhesion information is collected by the spectral analysis sensor, and the material adhesion degree is judged according to the preset adhesion degree threshold to obtain the material adhesion distribution map. The material distribution thermodynamic map and the adhesion distribution map are fused to extract the material distribution density and adhesion degree characteristics in the region. The K-means clustering algorithm is used to cluster the fused feature data, and the optimal number of clusters is determined according to the silhouette coefficient to obtain multiple cleaning areas inside the carriage. For each cleaning area, the mean value of the material distribution density and the mean value of the adhesion degree in the area are calculated respectively as two indicators to measure the cleaning priority of the area. The two indicators are normalized by min-max to obtain the distribution density score and adhesion degree score of each area. The distribution density score weight is set to 6, and the adhesion degree score weight is set to 4. The comprehensive score of each area is weighted to obtain the cleaning priority. According to the order of cleaning priority from high to low, the cleaning order of each cleaning area is determined. The cleaning order and the position coordinate information of each cleaning area are input into the cleaning robot. The cleaning robot cleans each cleaning area in order from high to low priority according to the input cleaning order and position coordinates. During the cleaning process, the vehicle environment information model is updated in real time. When all areas are cleaned, the cleaning task is completed.

[0015] S103. According to the characteristics of the cleaning area and the material properties, suitable cleaning tools are matched and selected from a predefined tool library, and the tools are automatically assembled through a robotic arm to improve cleaning adaptability.

[0016] Acquire image information of the cleaning area and image information of the material to be cleaned; use a fast random forest algorithm to extract and classify the image information of the cleaning area to obtain regional material and roughness regional feature parameters; use a convolutional neural network model to identify the material type of the image information of the material to be cleaned, and obtain the adhesion and corrosive material attribute parameters corresponding to the material type by querying a predefined material attribute knowledge base; perform similarity matching in a predefined tool feature database based on the regional feature parameters and the material attribute parameters to obtain several applicable cleaning tools; perform weighted scoring on the material, wear resistance and cleaning effect performance parameters of the applicable cleaning tools , select the cleaning tool with the highest score as the optimal cleaning tool, and send the model of the optimal cleaning tool to the robot arm control unit; the robot arm control unit queries the corresponding tool grabbing position and assembly steps in the predefined robot arm assembly knowledge base according to the received tool model, controls the robot arm to grab the target tool from the tool storage area and install it at the execution end, and completes the automatic replacement of the cleaning tool; the robot arm carries the optimal cleaning tool to carry out cleaning operations on the designated area, adjusts the movement speed and contact pressure of the robot arm according to the material of the area and the adhesion of the material, and controls the cleaning force; if there are stubborn stains, adopts repeated cleaning mode to clean, so as to improve the cleaning effect.

[0017] Specifically, according to the acquired image information of the cleaning area, a fast random forest algorithm is used to extract and classify characteristic parameters such as regional material and roughness; according to the acquired image of the material to be cleaned, a convolutional neural network model is used to identify the material type, and the material's attribute parameters such as adhesion and corrosiveness are obtained by querying a predefined material attribute knowledge base; according to the regional characteristic parameters and material attribute parameters, similarity matching is performed in a predefined tool feature database to screen out several suitable cleaning tools; the performance parameters such as material, wear resistance, and cleaning effect of the applicable tools are weightedly scored, and the tool with the highest score is selected as the optimal cleaning tool. Its model is sent to the robot control unit; the robot control unit queries the corresponding tool grasping position and assembly steps in the predefined robot assembly knowledge base according to the received tool model; the robot accurately grasps the target tool from the tool storage area according to the grasping position information, and installs it at the execution end according to the queried assembly steps to complete the automatic replacement of the tool; after the assembly is completed, the robot carries the cleaning tool to carry out cleaning operations on the designated area, and controls the cleaning force by adjusting the movement speed and contact pressure of the robot according to the material of the area and the adhesion of the material. For stubborn stains, a repeated cleaning mode is adopted to improve the cleaning effect while taking into account the cleaning efficiency.

[0018] S104. For each cleaning area, a reinforcement learning algorithm is used to learn and optimize the motion trajectory and operating parameters of the cleaning tool in combination with the internal environmental constraints of the vehicle compartment and the cleaning efficiency target, and a cleaning path is formulated.

[0019] The three-dimensional environmental model data of the interior of the vehicle and the cleaning area division information are obtained, and for each of the cleaning areas, the relevant environmental constraint condition parameters are extracted, and the environmental constraint condition parameters include the area of ​​the cleaning area and the obstacle distribution; a cleaning efficiency evaluation function is constructed, and the cleaning efficiency evaluation function combines multiple optimization objectives such as cleaning coverage, cleaning time, and energy consumption, and adopts a weighted summation method to assign weights to each of the optimization objectives, and linearly combines them to form the cleaning efficiency evaluation function; a deep reinforcement learning algorithm is adopted, and the motion trajectory and operation parameters of the cleaning tool are used as decision variables, and the cleaning efficiency evaluation function is used as a reward function. Through interactive exploration and learning between the intelligent agent and the environment, an optimized cleaning decision strategy is obtained; during the training process, according to the characteristics of each cleaning area, Given environmental constraints, the exploration behavior of the agent is guided and constrained, a penalty term is introduced in the reward function, and a negative reward is given for the behavior that produces an invalid or dangerous cleaning path; after multiple rounds of iterative training and optimization, the performance of the decision strategy model corresponding to each of the cleaning areas is evaluated on the validation set, and the model with the highest cleaning efficiency evaluation function value is selected as the optimal strategy model; during the actual cleaning operation, the current cleaning area information is obtained, it is determined which cleaning area the current cleaning area belongs to, the corresponding decision strategy model file is loaded, and the best cleaning tool motion trajectory and operation control parameter sequence are generated by real-time planning; the cleaning tool autonomously performs the cleaning operation through the motion control system of the robotic arm or mobile robot according to the motion trajectory and operation parameters generated by the plan.

[0020] Specifically, the three-dimensional environmental model data of the interior of the carriage and the information on the division of the cleaning area are obtained, and for each cleaning area, the relevant environmental constraint parameters are extracted, such as the area of ​​the cleaning area, the distribution of obstacles, etc. A cleaning efficiency evaluation function is constructed, and multiple optimization objectives such as cleaning coverage, cleaning time, and energy consumption are comprehensively considered. Weights can be assigned to each optimization objective in a weighted summation manner, and then they are linearly combined to form an evaluation function. The weights can be adjusted according to actual needs. The deep reinforcement learning algorithm DDPG (Deep Deterministic Policy Gradient) is adopted, the motion trajectory and operation parameters of the cleaning tool are used as decision variables, and the cleaning efficiency evaluation function is used as the reward function. Through interactive exploration and learning between the agent and the environment, the cleaning decision strategy is gradually optimized. During the training process, the exploration behavior of the agent is guided and constrained according to the specific environmental constraints of each cleaning area. A penalty term can be introduced in the reward function, and negative rewards can be given to behaviors that produce invalid or dangerous cleaning paths, guiding the agent to explore in a safe and efficient direction. After multiple rounds of iterative training and optimization, the performance of the decision strategy model corresponding to each cleaning area is evaluated on the validation set, and the model with the highest cleaning efficiency evaluation function value is selected as the optimal strategy model. In the actual cleaning operation, according to the current cleaning area information, the If-else conditional statement is used to determine which cleaning area the current cleaning area belongs to, and then the corresponding decision strategy model file is loaded to plan and generate the best cleaning tool motion trajectory and operation control parameter sequence in real time. According to the motion trajectory and operation parameters generated by the plan, the cleaning tool performs the cleaning operation autonomously through the motion control system of the robotic arm or mobile robot, realizing intelligent and efficient carriage cleaning.

[0021] S105. During the cleaning process, the visual sensor is used to obtain the distribution changes of the residual materials in real time, and the material distribution map is dynamically updated through target detection and semantic segmentation algorithms, and the cleaning path is adjusted in real time according to the updated distribution map.

[0022] Obtain an initial residual material distribution map of the cleaning area, wherein the initial residual material distribution map is obtained by performing a panoramic scan of the cleaning area through a preset visual sensor; divide the initial residual material distribution map into a number of cleaning units, each of which corresponds to an area of ​​a fixed size; determine the distribution density of residual materials in each cleaning unit according to the divided cleaning units; adopt a greedy algorithm to plan an initial cleaning path according to the distribution density, wherein the initial cleaning path gives priority to cleaning the cleaning units with a large residual material density; control the cleaning equipment to clean the cleaning units along the initial cleaning path; during the cleaning process, collect images of the cleaning area in real time through visual sensors; adopt the YOLOv3 target detection algorithm to detect residual material targets in the cleaning area image, and obtain the position coordinates of the residual material targets in the cleaning area image; use the position coordinates as prior knowledge and adopt FCN A semantic segmentation algorithm is used to segment the cleaning area image to obtain the pixel-level distribution of residual materials; the segmentation result is compared with the initial residual material distribution map, and a pixel-by-pixel difference calculation is performed to determine whether there is a newly added residual material area; if the number of newly added residual material pixels in a certain cleaning unit exceeds a preset threshold, the certain cleaning unit is determined as a newly added residual material area, and the newly added residual material area is added to the initial residual material distribution map to obtain an updated residual material distribution map; according to the updated residual material distribution map, the residual material distribution density of each cleaning unit is recalculated; the Dijkstra shortest path algorithm is used, with the cleaning unit as a node and the distance between adjacent cleaning units as an edge weight, to plan an updated cleaning path; the cleaning device is controlled to continue cleaning the cleaning unit along the updated cleaning path until all cleaning tasks are completed.

[0023] Specifically, to obtain the initial residual material distribution map of the cleaning area, the preset visual sensor can be used to perform a panoramic scan of the cleaning area to obtain an image of the cleaning area. The obtained cleaning area image is divided into several cleaning units, each of which corresponds to an area of ​​a fixed size. According to the divided cleaning units, the initial residual material distribution map is processed to obtain the distribution density of the residual material in each cleaning unit. Based on the distribution density, the greedy algorithm is used to plan the initial cleaning path, and the cleaning units with large residual material density are cleaned first. The cleaning equipment is controlled to clean the cleaning units along the planned initial cleaning path. During the cleaning process, the cleaning area image is collected in real time by the visual sensor, and the residual material target in the image is detected by the YOLOv3 target detection algorithm. The position coordinates of the detected residual material target in the image are used as prior knowledge, and the FCN semantic segmentation algorithm is used to segment the image to obtain the pixel-level distribution of the residual material. The segmentation result is compared with the initial residual material distribution map, and the pixel-by-pixel difference calculation is used to determine whether there is a new residual material area. If a large number of new residual material pixels appear in a cleaning unit, it is considered that there is a new residual material area in the cleaning unit, and it is added to the residual material distribution map. According to the updated residual material distribution map, the residual material distribution density of each cleaning unit is recalculated. The Dijkstra shortest path algorithm is used to plan the updated cleaning path with the cleaning unit as the node and the distance between adjacent cleaning units as the edge weight. The cleaning equipment is controlled to continue cleaning the cleaning unit along the updated cleaning path until all cleaning tasks are completed.

[0024] S106. For materials with a high degree of adhesion, the contact force between the cleaning tool and the material is monitored in real time by a force feedback sensor, and the pressure and angle of the cleaning tool are dynamically adjusted in combination with a pre-established material property database until the material is cleaned.

[0025] Obtain the adhesion threshold of the material in a pre-established material property database; monitor the contact force between the cleaning tool and the material in real time through a six-dimensional force sensor; compare the contact force monitored in real time with the material adhesion threshold; if the contact force is less than the material adhesion threshold, dynamically adjust the pressure and angle of the cleaning tool using a PID control algorithm so that the contact force gradually increases or reaches the adhesion threshold; optimize the pressure and angle parameters of the cleaning tool using a random forest algorithm based on the material hardness and roughness properties in the material property database; continue to perform the adjustment and optimization process until the contact force stabilizes near the adhesion threshold and the material surface is cleaned; detect the cleaned material surface through a machine vision system to determine whether it meets the preset cleanliness requirements; if the cleanliness requirements are met, the cleaning task is completed; otherwise, return to continue adjusting the cleaning tool until the cleaning is qualified.

[0026] Specifically, the adhesion threshold of the material in the pre-established material property database is obtained, and the contact force between the cleaning tool and the material is monitored in real time through a six-dimensional force sensor. The monitored contact force is compared with the material adhesion threshold. If the contact force is less than the material adhesion threshold, the PID control algorithm is used to dynamically adjust the pressure and angle of the cleaning tool so that the contact force gradually approaches or reaches the adhesion threshold. According to the material hardness, roughness and other properties in the material property database, the random forest algorithm is used to optimize the pressure and angle parameters of the cleaning tool. The above adjustment and optimization process is continuously performed until the contact force is stabilized near the adhesion threshold and the material surface is cleaned. The surface of the cleaned material is inspected by a machine vision system to determine whether the cleanliness requirements are met. If so, the cleaning task is completed, otherwise it returns to continue adjusting the cleaning tool until the cleaning is qualified.

[0027] S107. During the cleaning operation, record and accumulate cleaning data under different carriage environments and material characteristics, and continuously optimize the cleaning path planning and tool selection strategy through convolutional neural networks and decision tree algorithms to achieve self-learning and evolution of the cleaning equipment, thereby improving its adaptive ability and cleaning effect.

[0028] Acquire a cleaning operation dataset under different carriage environments and material characteristics, the dataset including environmental images, material types and cleaning effect evaluation; preprocess the dataset, including image normalization, data cleaning and annotation; randomly divide the preprocessed dataset into a training set and a test set; use a convolutional neural network to extract features of the environmental images in the training set, and extract multi-scale and multi-level feature representations through convolutional layers and pooling layers; fuse the feature representation of the environmental image with other features of the material type to obtain a comprehensive cleaning operation feature vector; use a decision tree algorithm to train the cleaning operation feature vector, and generate a decision tree model that adapts to different situations based on environmental complexity and material adhesion factors; use the test set data to train the trained convolutional neural network and decision tree models. The performance of the tree model is evaluated, and parameters are tuned and the model is improved according to the evaluation results. When the cleaning equipment is performing operations, the camera is used to collect real-time images of the car environment and identify the material type. The collected data is input into the trained convolutional neural network and decision tree model to adaptively generate the optimal cleaning path and tool selection strategy. The cleaning equipment performs cleaning operations according to the generated strategy and continuously records the cleaning process data through sensors. The recorded cleaning process data is automatically annotated and added to the training data set. At regular intervals, the convolutional neural network and decision tree model are incrementally trained and optimized using the newly added cleaning data to enable the model to continue to learn and evolve. The optimized model parameters are updated to the control system of the cleaning equipment so that the equipment can perform cleaning operations based on the latest decision rules.

[0029] Specifically, cleaning operation data under different carriage environments and material characteristics are obtained, including environmental images, material types, cleaning effect evaluation, etc., to construct a cleaning operation data set. The data set is preprocessed, including image normalization, data cleaning and annotation. The preprocessed data set is randomly divided into a training set and a test set. A convolutional neural network is used to extract features from the environmental images in the training set, and multi-scale and multi-level feature representations are extracted through convolutional layers and pooling layers. At the same time, other features such as material type are fused with image features to obtain a comprehensive cleaning operation feature vector. A decision tree algorithm (such as CART, C5, etc.) is used to train the cleaning operation feature vector to construct decision rules for cleaning path planning and tool selection. According to factors such as environmental complexity and material adhesion, decision trees that adapt to different situations are generated. The performance of the trained convolutional neural network and decision tree models is evaluated using the test set data, and parameter tuning and model improvement are performed based on the evaluation results to improve the generalization ability of the model and decision accuracy. When the cleaning equipment is performing operations, the camera collects real-time images of the carriage environment and identifies the material type. The collected data is input into the trained convolutional neural network and decision tree model to adaptively generate the optimal cleaning path and tool selection strategy. The cleaning equipment performs cleaning operations according to the generated strategy and continuously records the cleaning process data through sensors, including cleaning tracks, tool usage, cleaning effects, etc. The recorded data is automatically annotated and added to the training data set. At regular intervals (such as every month), the convolutional neural network and decision tree model are incrementally trained and optimized using the newly added cleaning data, so that the model can continue to learn and evolve and adapt to the changing carriage environment and material properties. The optimized model parameters are updated to the control system of the cleaning equipment so that the equipment can perform cleaning operations based on the latest decision rules. At the same time, the working status and cleaning effects of the cleaning equipment are displayed in real time through the human-computer interaction interface, which is convenient for staff to supervise and manage.

[0030] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An intelligent carriage cleaning method, characterized in that: The method comprises: The three-dimensional point cloud data of the interior of the carriage is obtained through laser radar scanning, and the semantic segmentation of the point cloud data is performed in combination with a deep learning algorithm to identify the complex structure of the interior of the carriage and the distribution of residual materials, and obtain the interior environment information model of the carriage; according to the interior environment information model of the carriage, a rule-based reasoning algorithm is used to analyze the location distribution characteristics of the materials, and the adhesion degree of the materials is detected by a spectral analysis sensor, the interior of the carriage is divided into multiple cleaning areas, and the cleaning priority of each area is determined; according to the characteristics of the cleaning area and the properties of the materials, suitable cleaning tools are matched and selected from the predefined tool library, and the automatic assembly of the tools is realized through the robotic arm to improve the cleaning adaptability; for each cleaning area, a reinforcement learning algorithm is used to learn and optimize the cleaning process in combination with the internal environment constraints of the carriage and the cleaning efficiency goals. The motion trajectory and operating parameters of the sweeping tool are used to formulate a cleaning path; during the cleaning process, the visual sensor is used to obtain the distribution changes of the residual materials in real time, and the material distribution map is dynamically updated through target detection and semantic segmentation algorithms, and the cleaning path is adjusted in real time according to the updated distribution map; for materials with high adhesion, the contact force between the cleaning tool and the material is monitored in real time through a force feedback sensor, and the pressure and angle of the cleaning tool are dynamically adjusted in combination with the pre-established material property database until the material is cleaned; during the cleaning operation, the cleaning data under different carriage environments and material properties are recorded and accumulated, and the cleaning path planning and tool selection strategies are continuously optimized through convolutional neural networks and decision tree algorithms, so as to realize the self-learning and evolution of the cleaning equipment and improve its adaptive ability and cleaning effect.

2. The method according to claim 1, characterized in that The three-dimensional point cloud data of the interior of the carriage is obtained by laser radar scanning, and the point cloud data is semantically segmented in combination with a deep learning algorithm to identify the complex structure of the interior of the carriage and the distribution of residual materials, and obtain the interior environment information model of the carriage, including: Obtain the original point cloud dataset inside the car, use the statistical filter in the point cloud library PCL to remove the noise points in the point cloud, and use the voxelized grid method to downsample the point cloud to obtain the preprocessed point cloud dataset; The preprocessed point cloud data is input into the PointNet++ network. Through point cloud semantic segmentation, each point is divided into two categories: the internal structure of the carriage and the residual material. The point cloud subset of the internal structure of the carriage and the point cloud subset of the residual material are obtained. For the point cloud subset of the internal structure of the carriage, the region growing algorithm is used for geometric segmentation. According to the growth criteria of the normal vector angle threshold and the curvature threshold, the point cloud is segmented into different structural components, and the point cloud of each component is clustered by Euclidean clustering to obtain independent structural components. For the residual material point cloud subset, the number of neighboring points of each point is calculated to obtain the local density distribution of the point cloud; Project the point cloud to the Z-axis direction, count the number of points in different height intervals, and generate a height distribution histogram; Determine the accumulation position and quantity of residual materials according to the density distribution peak area and the height histogram peak position; The geometric segmentation results of the internal structure of the carriage are spatially registered with the distribution of residual materials to construct a three-dimensional model of the internal environment of the carriage containing information on structural components and residual materials. Based on the constructed interior environment model of the carriage, combined with the residual material distribution and structural layout, the motion path of the cleaning robot is optimized to generate a cleaning trajectory that covers the residual material area and avoids structural obstacles. Select cleaning tools and adjust cleaning parameters; Obtain the point cloud of the interior of the carriage after cleaning, and quantify the change in the volume of residual materials before and after cleaning by comparing it with the original point cloud, and evaluate the cleaning effect. If the expected cleanliness is not achieved, adjust the cleaning strategy until the requirements are met.

3. The method according to claim 1, characterized in that The method uses a rule-based reasoning algorithm to analyze the location distribution characteristics of materials based on the internal environment information model of the carriage, detects the adhesion degree of materials through a spectral analysis sensor, divides the interior of the carriage into multiple cleaning areas, and determines the cleaning priority of each area, including: Acquire an environmental information model inside the carriage, analyze the environmental information model using a decision tree algorithm, and obtain a material distribution thermodynamic map; Obtaining material adhesion information, determining the material adhesion degree according to a preset adhesion degree threshold, and obtaining a material adhesion distribution map; Performing data fusion on the material distribution thermodynamic map and the adhesion distribution map to extract material distribution density characteristics and adhesion degree characteristics in the region; The K-means clustering algorithm is used to cluster the fused feature data, and the optimal number of clusters is determined according to the silhouette coefficient to obtain multiple cleaning areas inside the car. For each cleaning area, the mean value of material distribution density and the mean value of adhesion degree in the area are calculated respectively, and the two means are normalized to obtain the distribution density score and adhesion degree score of each area; Setting a distribution density score weight and an adhesion degree score weight, weightedly calculating a comprehensive score for each area, and determining a cleaning priority level for each cleaning area according to the comprehensive score; Determining the cleaning order of each cleaning area according to the cleaning priority level from high to low; Inputting the cleaning sequence and the position coordinate information of each cleaning area into the cleaning robot; The cleaning robot performs cleaning operations on each of the cleaning areas in order from high to low priority according to the input cleaning order and position coordinates; During the cleaning process, the vehicle compartment environment information model is updated in real time, and when all areas are cleaned, the cleaning task is completed.

4. The method according to claim 1, characterized in that: According to the characteristics of the cleaning area and the material properties, suitable cleaning tools are matched and selected from the predefined tool library, and the automatic assembly of the tools is realized by the robotic arm to improve the cleaning adaptability, including: Obtain image information of the cleaning area and image information of the material to be cleaned; A fast random forest algorithm is used to extract and classify the image information of the cleaned area to obtain regional material and roughness regional feature parameters; A convolutional neural network model is used to identify the material type of the image information of the material to be cleaned, and the adhesive and corrosive material property parameters corresponding to the material type are obtained by querying a predefined material property knowledge base; According to the regional characteristic parameters and the material attribute parameters, similarity matching is performed in a predefined tool characteristic database to obtain a number of applicable cleaning tools; Performing a weighted score on the material, wear resistance and cleaning effect performance parameters of the applicable cleaning tools, selecting the cleaning tool with the highest score as the optimal cleaning tool, and sending the model of the optimal cleaning tool to the robot arm control unit; The robot arm control unit searches for the corresponding tool grabbing position and assembly steps in the predefined robot arm assembly knowledge base according to the received tool model, controls the robot arm to grab the target tool from the tool storage area and install it at the execution end, and completes the automatic replacement of the cleaning tool; The robot arm carries the optimal cleaning tool to carry out cleaning operations on the designated area, and adjusts the movement speed and contact pressure of the robot arm to control the cleaning force according to the material of the area and the adhesion of the material; If there are stubborn stains, use the repeated cleaning mode to improve the cleaning effect.

5. The method according to claim 1, characterized in that For each cleaning area, the motion trajectory and operation parameters of the cleaning tool are learned and optimized by the reinforcement learning algorithm in combination with the internal environmental constraints of the carriage and the cleaning efficiency target, and the cleaning path is formulated, including: Acquire the three-dimensional environment model data of the interior of the vehicle and the cleaning area division information, and extract the relevant environmental constraint condition parameters for each cleaning area, wherein the environmental constraint condition parameters include the area of ​​the cleaning area and the obstacle distribution; Constructing a cleaning efficiency evaluation function, wherein the cleaning efficiency evaluation function combines multiple optimization objectives such as cleaning coverage, cleaning time and energy consumption, adopts a weighted summation method, assigns weights to each of the optimization objectives, and linearly combines them to form the cleaning efficiency evaluation function; A deep reinforcement learning algorithm is adopted, the motion trajectory and operation parameters of the cleaning tool are used as decision variables, and the cleaning efficiency evaluation function is used as a reward function. An optimized cleaning decision strategy is obtained through interactive exploration and learning between the intelligent agent and the environment. During the training process, the exploration behavior of the agent is guided and constrained according to the specific environmental constraints of each cleaning area, and a penalty term is introduced into the reward function to give negative rewards for behaviors that produce invalid or dangerous cleaning paths; After multiple rounds of iterative training optimization, the performance of the decision strategy model corresponding to each cleaning area is evaluated on the validation set, and the model with the highest cleaning efficiency evaluation function value is selected as the optimal strategy model; During the actual cleaning operation, the current cleaning area information is obtained, the current cleaning area is determined, the corresponding decision strategy model file is loaded, and the optimal cleaning tool motion trajectory and operation control parameter sequence are generated in real time; The cleaning tool autonomously performs cleaning operations according to the motion trajectory and operation parameters generated by the plan through the motion control system of the mechanical arm or the mobile robot.

6. The method according to claim 1, characterized in that During the cleaning process, the distribution change of residual materials is obtained in real time by using visual sensors, the material distribution map is dynamically updated through target detection and semantic segmentation algorithms, and the cleaning path is adjusted in real time according to the updated distribution map, including: Acquire an initial residual material distribution map of the cleaning area, wherein the initial residual material distribution map is obtained by performing a panoramic scan of the cleaning area through a preset visual sensor; Dividing the initial residual material distribution map into a plurality of cleaning units, each of the cleaning units corresponds to an area of ​​a fixed size; According to the divided cleaning units, determining the distribution density of the residual materials in each of the cleaning units; A greedy algorithm is used to plan an initial cleaning path according to the distribution density, and the initial cleaning path gives priority to cleaning units with a large residual material density; Controlling a cleaning device to perform a cleaning operation on the cleaning unit along the initial cleaning path; During the cleaning process, the visual sensor collects images of the cleaning area in real time; Using the YOLOv3 target detection algorithm to detect the residual material target in the cleaning area image, and obtain the position coordinates of the residual material target in the cleaning area image; The position coordinates are used as prior knowledge, and the FCN semantic segmentation algorithm is used to segment the cleaning area image to obtain the pixel-level distribution of the residual materials; Comparing the segmentation result with the initial residual material distribution map, and determining whether there is a newly added residual material area by calculating the difference between each pixel; If the number of newly added residual material pixels in a certain cleaning unit exceeds a preset threshold, the certain cleaning unit is determined as a newly added residual material area, and the newly added residual material area is added to the initial residual material distribution map to obtain an updated residual material distribution map; Recalculating the residual material distribution density of each cleaning unit according to the updated residual material distribution map; Using the Dijkstra shortest path algorithm, taking the cleaning units as nodes and the distances between adjacent cleaning units as edge weights, an updated cleaning path is planned; The cleaning device is controlled to continue cleaning the cleaning unit along the updated cleaning path until all cleaning tasks are completed.

7. The method according to claim 1, characterized in that For materials with high adhesion, the contact force between the cleaning tool and the material is monitored in real time by a force feedback sensor, and the pressure and angle of the cleaning tool are dynamically adjusted in combination with a pre-established material property database until the material is cleaned, including: Obtaining a material adhesion threshold value from a pre-established material property database; The contact force between the cleaning tool and the material is monitored in real time through a six-dimensional force sensor; comparing the contact force monitored in real time with the material adhesion threshold; If the contact force is less than the material adhesion threshold, the pressure and angle of the cleaning tool are dynamically adjusted using a PID control algorithm so that the contact force gradually increases or reaches the adhesion threshold; According to the material hardness and roughness attributes in the material property database, a random forest algorithm is used to optimize the pressure and angle parameters of the cleaning tool; Continuously performing the adjustment and optimization process until the contact force is stabilized near the adhesion threshold and the surface of the material is cleaned; The machine vision system is used to inspect the surface of the cleaned material to determine whether it meets the preset cleanliness requirements; If the cleanliness requirements are met, the cleaning task is completed; Otherwise, return to continue adjusting the cleaning tool until the cleaning is qualified.

8. The method according to claim 1, characterized in that During the cleaning operation, the cleaning data under different compartment environments and material characteristics are recorded and accumulated, and the cleaning path planning and tool selection strategy are continuously optimized through convolutional neural networks and decision tree algorithms to achieve self-learning and evolution of the cleaning equipment, and improve its adaptive ability and cleaning effect, including: Acquire a cleaning operation data set under different carriage environments and material characteristics, the data set including environment images, material types, and cleaning effect evaluations; Preprocessing the data set, including image normalization, data cleaning and labeling; The preprocessed data set is randomly divided into training set and test set; Using a convolutional neural network to extract features from the environment images in the training set, and extracting multi-scale and multi-level feature representations through convolutional layers and pooling layers; The feature representation of the environment image is integrated with other features of the material type to obtain a comprehensive cleaning operation feature vector; A decision tree algorithm is used to train the cleaning operation feature vector, and a decision tree model adapted to different situations is generated according to factors such as environmental complexity and material adhesion; Using the test set data to evaluate the performance of the trained convolutional neural network and decision tree model, and performing parameter tuning and model improvement based on the evaluation results; When the cleaning equipment is performing operations, the camera collects images of the carriage environment in real time and identifies the type of material; The collected data is input into the trained convolutional neural network and decision tree model to adaptively generate the optimal cleaning path and tool selection strategy; The cleaning equipment performs cleaning operations according to the generated strategy and continuously records the cleaning process data through sensors; The recorded cleaning process data is automatically annotated and added to the training data set; At regular intervals, the convolutional neural network and decision tree models are incrementally trained and optimized using newly added cleaning data, allowing the models to continue learning and evolving. The optimized model parameters are updated to the control system of the cleaning equipment, allowing the equipment to perform cleaning operations based on the latest decision rules.

Images