A method and system for planning a motion path of a bridge crane

Abstract

The present application belongs to the technical field of industrial crane control, and provides a bridge crane motion path planning method and system, which comprises: uniformly dividing the motion environment of the bridge crane into grid units, identifying dynamic obstacles and giving each grid unit a state value, trajectory predicting the dynamic obstacles, generating a time grid state matrix, constructing a time sequence of passing grid set sequence, exploring the shortest path by using an improved A* algorithm, exploring multiple feasible paths by using an RRT algorithm, marking the change nodes and screening the multiple feasible paths based on the shortest path, marking the initial path and arranging the initial path set and path node sequence, for the initial path, quantitatively calculating the path tortuosity of the initial path, and screening based on the path tortuosity, marking the optimal path, analyzing the load swing of the load at the change node, calculating the swing envelope range of the load, marking the risk change node, and screening to obtain the optimal path.

Description

Technical Field

[0001] This invention belongs to the field of industrial crane control technology, specifically a method and system for planning the motion path of a bridge crane. Background Technology

[0002] In the complex scenarios of industrial production, bridge cranes undertake the critical task of material handling, and the scientific nature of their motion path planning is crucial for improving production efficiency and ensuring operational safety. However, current path planning technologies are gradually showing many shortcomings when faced with dynamically changing operating environments.

[0003] In the environmental modeling stage, existing technologies often rely on a single design parameter or rough on-site measurements to determine spatial boundaries, resulting in deviations between the model and the actual working environment. Furthermore, the use of static mesh generation methods makes it difficult to reflect the dynamic changes of the environment over time. In obstacle handling, traditional methods have vague criteria for distinguishing between static and dynamic obstacles, lack quantitative judgment basis, and have insufficient accuracy in predicting the trajectory of dynamic obstacles, which can easily lead to path failure due to obstacle movement.

[0004] In the path search process, the limitations of a single algorithm are significant. Although the improved A* algorithm can obtain the shortest path, the generated path has many redundant turns, which increases energy consumption and control complexity. Although the RRT algorithm can generate multiple feasible paths, the lack of a goal-oriented strategy leads to low search efficiency. At the same time, existing technologies often ignore the impact of the time dimension on path feasibility and fail to establish spatiotemporal connectivity constraints. The planned path may conflict with dynamic obstacles in the future.

[0005] In the path selection stage, existing technologies mostly rely on path length or number of turns for judgment, without considering the standardized quantification of path tortuosity. This makes it difficult to compare the smoothness of different paths horizontally, and it does not combine load swing characteristics to analyze collision risks. Accidents may be caused by load swing exceeding the safe range. These problems make it difficult for existing path planning methods to meet the needs of complex industrial scenarios in terms of efficiency, safety and adaptability. There is an urgent need for a comprehensive solution that integrates dynamic environment modeling, spatiotemporal path search and risk quantification analysis.

[0006] To address the above problems, this invention proposes a method and system for planning the motion path of a bridge crane. Summary of the Invention

[0007] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0008] The technical solution adopted by this invention to solve the technical problem is: a method for planning the motion path of a bridge crane, comprising:

[0009] The motion environment of the bridge crane is uniformly divided into grid cells, dynamic obstacles are identified and a state value is assigned to each grid cell, the trajectory of the dynamic obstacles is predicted, and a time grid state matrix including the state value of each grid cell at the corresponding time is generated.

[0010] Based on the time grid state matrix, a temporal access grid set sequence including the access grid set is constructed. An improved A* algorithm is used to explore the shortest path, and the RRT algorithm is used to explore multiple feasible paths. Change nodes are marked and multiple feasible paths are filtered based on the shortest path. The initial path is marked and the initial path set and path node sequence are organized.

[0011] For the initial path, the path tortuosity is quantitatively calculated, and the path tortuosity is used for screening, marking the preferred paths and organizing them into a set of preferred paths;

[0012] For the preferred path, analyze the load swing at the change node, calculate the load swing envelope range, mark the risk change node, and filter to obtain the optimal path;

[0013] The state value is obtained in the following way:

[0014] A three-dimensional coordinate system is established for the motion environment of the bridge crane. The motion environment is uniformly divided into grid cells. Motion environment data is collected in real time and obstacles are identified. An obstacle analysis period is set with the current time as the endpoint. During the obstacle analysis period, the distance between the coordinates of the grid cells covered by the obstacle at adjacent times is calculated. If the distance between any two adjacent times meets the preset conditions, the obstacle is judged as a static obstacle; otherwise, the obstacle is judged as a dynamic obstacle.

[0015] Assign a state value to each grid cell: a grid cell not covered by any obstacle has a state value of 0, a grid cell covered by a static obstacle has a state value of 1, and a grid cell covered by a dynamic obstacle has a state value of 2.

[0016] The time grid state matrix is ​​obtained as follows:

[0017] Set a future time period starting from the current time, obtain the coordinates of the grid cells covered by the dynamic obstacle at each time point within the obstacle analysis time period, use the Kalman filter algorithm to predict the motion trajectory of the dynamic obstacle in the future time period, obtain the coordinates of the grid cells covered by the dynamic obstacle at each time point within the future time period, organize and transform to obtain the state value of each grid cell at each time point within the future time period, and generate a time grid state matrix.

[0018] The method for obtaining the temporal access grid set sequence is as follows:

[0019] The path planning is defined by the starting grid cell, the ending grid cell, and the obstacle safety distance. At each time point in the future time period, all grid cells with a state value of 0 are obtained and marked as empty grids. Grid cells with a state value of non-zero are marked as obstacle grids. The minimum Euclidean distance between empty grids and obstacle grids is calculated. If it is greater than the obstacle safety distance, the empty grid is marked as a passable grid and added to the passable grid set.

[0020] Preset the desired start time for the bridge crane, clear the passing grid set before the desired start time, and sort it to obtain the time sequence passing grid set sequence;

[0021] The method for obtaining the shortest path is as follows:

[0022] Each grid cell center point is considered as a node, where the starting grid cell and the ending grid cell are the starting node and the ending node, respectively. The starting grid cell of the temporal access grid set sequence is obtained. The improved A* algorithm is used to perform path search within the starting grid cell to obtain the basic shortest path. The line segments between the nodes in the shortest path and the non-adjacent subsequent nodes are analyzed to see if they are within the starting grid cell. Based on the analysis results, the changed nodes are marked to obtain the shortest path.

[0023] Set the desired average speed for the bridge crane, combine the shortest path with the desired start time, mark the grid cell where the spreader center is located as the instantaneous position, calculate the instantaneous position of the bridge crane at each moment in the time-series travel grid set sequence, if there is an instantaneous position that is not in the travel grid set at the corresponding moment, then clear the node corresponding to the instantaneous position in the starting travel grid set and perform path search again, otherwise obtain the shortest path;

[0024] The path node sequence is obtained as follows:

[0025] Obtain the starting grid set of the temporal grid set sequence, and use the RRT algorithm to run independently multiple times to search for paths within the starting grid set. This will yield multiple feasible paths. Analyze whether the line segments between nodes in the feasible paths and non-adjacent subsequent nodes are within the starting grid set. Based on the analysis results, mark the changed nodes and simplify the update of feasible paths.

[0026] For any feasible path, combine the expected average speed and the expected start time to determine whether the instantaneous position is within the travel grid set at the corresponding time. If there is an instantaneous position that is not within the travel grid set at the corresponding time, the feasible path is cleared.

[0027] Calculate the actual lengths of feasible paths and shortest paths and perform data processing to obtain the length deviation of feasible paths. If the length deviation meets the preset conditions, mark the feasible path as the initial path. After traversing all the obtained feasible paths, mark the shortest path as the initial path and integrate the nodes of the initial path to obtain the path node sequence of each initial path.

[0028] The method for obtaining the path tortuosity is as follows:

[0029] For each initial path, the sequence of path nodes of the initial path is obtained. The vector calculation method is used to calculate the turning angle of the changing nodes, calculate the actual length of each initial path, and combine the data processing with the number of changing nodes in the initial path to calculate the path tortuosity of the initial path.

[0030] The preferred path is obtained as follows:

[0031] The mean path tortuosity of each initial path in the initial path set is calculated as the tortuosity threshold. The initial path set is then filtered based on the tortuosity threshold. For each initial path, if the path tortuosity of the initial path is less than the tortuosity threshold, the initial path is marked as the preferred path.

[0032] The optimal path is obtained as follows:

[0033] The motion equations of the load pendulum model are established based on the Lagrange equation. Each change node in the preferred path is obtained. Based on the expected average speed and the turning angle of the change node, the acceleration of the control load of the bridge crane in the x and y directions of the horizontal plane is calculated respectively. The load swing angle is solved by numerical integration. The swing envelope range of the load is obtained by combining the length of the wire rope lowered by the bridge crane.

[0034] By combining the preferred path, expected average speed, and expected start time, the instantaneous position of the bridge crane reaches the change node at the moment of change and marks it as the change time. The change window is taken with the change time as the midpoint. The overlapping part of the passing grid set within the change window is organized into an overlapping grid set. If the swing envelope range of the load on the change node exceeds the overlapping grid set, the change node is designated as a risk change node. The preferred path with the minimum number of risk change nodes is the optimal path.

[0035] A bridge crane motion path planning system includes the following modules:

[0036] Environment Modeling Module: The motion environment of the bridge crane is uniformly divided into grid cells, dynamic obstacles are identified and a state value is assigned to each grid cell, the trajectory of the dynamic obstacles is predicted, and a time grid state matrix including the state value of each grid cell at the corresponding time is generated.

[0037] Basic planning module: Based on the time grid state matrix, construct a temporal access grid set sequence including the access grid set, and use the improved A* algorithm to explore the shortest path, use the RRT algorithm to explore multiple feasible paths, mark change nodes and filter multiple feasible paths based on the shortest path, mark the initial path and organize the initial path set and path node sequence;

[0038] Preliminary screening module: For the initial path, the path tortuosity is quantitatively calculated, and the path tortuosity is screened based on the path tortuosity, the preferred paths are marked and organized into a set of preferred paths;

[0039] Load selection module: For the preferred path, analyze the load swing at the change node, calculate the load swing envelope range, mark the risk change node, and filter to obtain the optimal path.

[0040] The beneficial effects of this invention are as follows:

[0041] 1. This invention, by uniformly dividing the motion environment into grid cells and generating a time-grid state matrix, can accurately grasp dynamic obstacle information, providing a reliable basis for path planning. Constructing a temporal passage grid set sequence, combined with an improved A* algorithm and RRT algorithm, ensures that the shortest path can be found while exploring multiple feasible paths. After screening, a more reasonable initial path is obtained, effectively improving the comprehensiveness and accuracy of path planning and reducing the operational risks of cranes.

[0042] 2. This invention, by quantifying and calculating the tortuosity of the initial path and filtering it, can remove unreasonable paths and obtain a straighter, preferred path, thereby reducing energy consumption and time costs during crane operation. It also analyzes the swaying of the load at changing nodes and marks risky changing nodes, filtering out the optimal path. This can greatly reduce the safety hazards caused by load swaying, ensure the safe and stable operation of the bridge crane, and improve the overall work efficiency and quality, thus having high practical value. Attached Figure Description

[0043] The invention will now be further described with reference to the accompanying drawings.

[0044] Figure 1 This is a flowchart illustrating the steps of a bridge crane motion path planning method according to an embodiment of the present invention.

[0045] Figure 2 This is a module architecture diagram of a bridge crane motion path planning system according to an embodiment of the present invention. Detailed Implementation

[0046] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0047] Example 1:

[0048] Please see Figure 1 As shown in the figure, a method for planning the motion path of a bridge crane according to an embodiment of the present invention includes the following steps:

[0049] S1: Divide the motion environment of the bridge crane into grid cells, identify dynamic obstacles and assign state values ​​to each grid cell, predict the trajectory of the dynamic obstacles, and generate a time grid state matrix including the state values ​​of each grid cell at the corresponding time.

[0050] The motion environment of the bridge crane is modeled in a grid. Specifically, the three-dimensional spatial boundary of the motion environment is determined based on the physical motion limit of the bridge crane. The three-dimensional spatial boundary includes the length boundary, the width boundary, and the height boundary. The length boundary is the extreme positions at both ends of the main track, the width boundary is the motion limit of the trolley on the main track, and the height boundary is the lowest and highest positions of the lifting device. The three-dimensional spatial boundary is obtained by the motion limit parameters in the bridge crane design manual and the actual spatial dimensions obtained by laser scanning on site.

[0051] Establish a three-dimensional coordinate system for the motion environment, and uniformly divide the motion environment into several grid cells. The coordinates of the center point of each grid cell in the three-dimensional coordinate system are used as the coordinates of the grid cell, and are labeled as follows: ;

[0052] The motion environment data is collected in real time by fusing LiDAR point cloud data with high-definition camera data. The motion environment data includes LiDAR point cloud data and motion environment images. Based on the motion environment data, obstacles are identified in the motion environment, and the three-dimensional contours of the identified obstacles are projected into the corresponding grid cells.

[0053] An obstacle analysis period is set with the current time as the endpoint. During the obstacle analysis period, the obstacles identified based on the motion environment data are analyzed. For any obstacle during the obstacle analysis period, the distance between the coordinates of the grid cells covered by the obstacle at adjacent times is calculated and marked as the change in the position of the obstacle.

[0054] If the position change of an obstacle is less than or equal to a preset position change threshold at any two adjacent moments within the obstacle analysis period, the obstacle is determined to be a static obstacle; otherwise, the obstacle is determined to be a dynamic obstacle.

[0055] For example, static obstacles include factory columns, fixed shelves, ground equipment, etc., while dynamic obstacles include other cranes, forklifts, personnel, etc. It should be noted that neither static nor dynamic obstacles include the bridge crane itself and the load it transports.

[0056] Assign a state value to each grid cell ∈{0,1,2}, Indicates coordinates as The state value of the grid cell; for a grid cell not covered by any obstacle, the state value of the grid cell. =0, for a grid cell covered by a static obstacle, the state value of the grid cell is... For a grid cell covered by dynamic obstacles, the state value of the grid cell is... The state value of a grid cell can change as a dynamic obstacle moves over time;

[0057] For dynamic obstacles, a future time period starting from the current time is set, and the duration of the future time period is the preset maximum expected passage time of the bridge crane in the motion environment. The coordinates of the grid cells covered by the dynamic obstacle at each moment in the obstacle analysis time period are obtained. The Kalman filter algorithm is used to predict the motion trajectory of the dynamic obstacle in the future time period, and the coordinates of the grid cells covered by the dynamic obstacle at each moment in the future time period are obtained.

[0058] The state values ​​of each grid cell at each time point within the future time period are obtained by sorting and transforming the data, generating a time grid state matrix M(t,g), where t represents the t-th time point within the future time period, t∈[1,T], T represents the total number of time points within the future time period, and the matrix elements represent the state value of grid cell g at the t-th time point within the future time period. The coordinates of grid cell g are... ;

[0059] It should be noted that the purpose of this step is to construct a spatiotemporal dynamic model of the bridge crane's motion environment, providing accurate environmental data support for subsequent path planning, introducing Kalman filtering to predict dynamic obstacle trajectories, and generating a time grid state matrix to achieve integrated modeling of spatial grid and time series, breaking through the limitations of traditional static environment modeling and laying the foundation for path planning.

[0060] S2: Construct a temporal access grid set sequence including the access grid set based on the time grid state matrix, and use the improved A* algorithm to explore the shortest path, use the RRT algorithm to explore multiple feasible paths, mark change nodes and filter multiple feasible paths based on the shortest path, mark the initial path and organize the initial path set and path node sequence;

[0061] Based on the obtained time grid state matrix, preliminary path planning is performed for the bridge crane;

[0062] Specifically, the basic parameters for path planning are set, including the starting grid cell, the ending grid cell, and the obstacle safety distance d. The starting grid cell is the grid cell where the center of the spreader is located at the current moment on the bridge crane, and the ending grid cell is the grid cell where the center of the spreader is located when the bridge crane transports the load to the designated location. During the load transportation process, the grid cell where the center of the spreader is located is marked as the real-time position. The obstacle safety distance d is preset by those skilled in the art based on the bridge crane's own parameters and the load volume.

[0063] Based on the state values ​​of grid cells and the obstacle safety distance d, a temporal access grid set sequence is constructed;

[0064] Specifically, at each time point in the future time period, all grid cells with a state value of 0 are obtained and marked as empty grids, and grid cells with a state value of non-zero are marked as obstacle grids. The minimum Euclidean distance between empty grids and obstacle grids is calculated. If the minimum Euclidean distance is greater than the obstacle safety distance, it is determined that the empty grid is passable, and the empty grid is marked as a passable grid and added to the passable grid set.

[0065] The access grid set at each moment in the future time period is organized according to the time sequence to obtain the time sequence access grid set sequence. The expected start time of the bridge crane is preset. The access grid set before the expected start time is cleared in the time sequence access grid set sequence. After clearing, the time sequence access grid set sequence starts from the access grid set at the expected start time.

[0066] Obtain the starting access grid set of the temporal access grid set sequence, and use the improved A* algorithm to perform path search within the starting access grid set;

[0067] Specifically, the center point of each grid cell is considered as a node, with the starting grid cell and the ending grid cell being the starting node and the ending node, respectively. A cost function is then set. :

[0068] ;

[0069] Where n represents a node, This represents the actual path length from the starting node to the current node n, which is obtained by summing the Euclidean distances between each adjacent node along the path from the starting node to the current node n. Let n be the heuristic function, representing the estimated cost from the current node n to the destination node, which is obtained by calculating the Euclidean distance between node n and the destination node.

[0070] Create an open list and a closed list. The open list is used to store nodes to be explored, and initially only contains the starting node. The closed list is used to store nodes that have been explored, and is initially empty. Create a predecessor node matrix prev to record the node before each node on the basic shortest path. Initially, all elements are set to null.

[0071] Select a cost function from the open list. The grid cell with the smallest value is selected as the current node and moved from the open list to the closed list. For all adjacent nodes m of the current node n, if node m does not belong to the starting passable grid set or is already in the closed list, node m is skipped. If node m is not skipped, when node m is not in the open list, the cost function of node m is calculated and m is added to the open list. After traversing all adjacent nodes m, the node m with the smallest cost function is selected from the open list as the current node and moved from the open list to the closed list.

[0072] If the current node is the destination node, the algorithm ends by backtracking the path through the predecessor node matrix prev, and the basic shortest path is obtained.

[0073] The algorithm sequentially checks whether the line segment between the starting node and non-adjacent subsequent nodes in the basic shortest path is within the starting grid set. For any non-adjacent subsequent node, if all are within the grid set, the node between the starting node and non-adjacent subsequent nodes is removed from the basic shortest path. The algorithm then continues to analyze the next non-adjacent subsequent node until a line segment between the starting node and non-adjacent subsequent node exceeds the starting grid set. This non-adjacent subsequent node is marked as a changed node. Starting from the changed node, the algorithm continues to analyze the non-adjacent subsequent nodes of the changed node to find subsequent changed nodes until the ending node is reached. This simplifies the basic shortest path structure and yields the shortest path.

[0074] It should be noted that the change node refers to the node in the planned path where the angle needs to be changed;

[0075] Set the desired average speed for the bridge crane. Combine the obtained shortest path with the desired start time to calculate the instantaneous position of the bridge crane at each moment in the time-series travel grid set. Determine whether the instantaneous position is in the travel grid set at the corresponding moment. If not, clear the node corresponding to the instantaneous position in the starting travel grid set of the time-series travel grid set. Re-search the path based on the cleared starting travel grid set. If all instantaneous positions are in the travel grid set at the corresponding moment, the shortest path is obtained, and the actual length of the shortest path is calculated.

[0076] To obtain the starting point of the temporal access grid set sequence, the Fast Exploratory Random Tree (RRT) algorithm is used to perform path search within the starting point access grid set. Specifically, the starting point node is used as the root node, and a random tree T is initialized. The tree only contains the starting point node and the index of the parent node. A node is randomly sampled from all the outermost nodes in the starting point access grid set and marked as the sampling node. To improve the efficiency of expanding to the destination, a target bias strategy is introduced, that is, the destination node is directly used as the sampling point with a probability of p.

[0077] Find the node in the random tree that is closest to the sampling point in Euclidean distance and mark it as the nearest node. Extend a new node from the nearest node toward the sampling point. The extension step size is set according to the grid size. Check whether there are any nodes between the line segments from the nearest node to the new node that do not belong to the starting point's passable grid set. If there are none, the new node is determined to be a feasible node. Otherwise, discard the new node, add the feasible node to the random tree, and record its parent node as the nearest node.

[0078] If the new node is the destination node, it is considered that the destination has been reached. By backtracking the parent node, a feasible path is obtained. The RRT algorithm is run independently multiple times, with different sampling points each time, resulting in different feasible paths.

[0079] Similarly, we sequentially determine whether the line segments between the starting node and non-adjacent subsequent nodes in the feasible path are within the starting point's passable grid set. For any non-adjacent subsequent node, if they are all within the starting point's passable grid set, we remove the nodes between the starting node and non-adjacent subsequent nodes in the feasible path and continue analyzing the next non-adjacent subsequent node until a line segment between the starting node and non-adjacent subsequent node exceeds the starting point's passable grid set. We mark this non-adjacent subsequent node as a changed node. Starting from the changed node, we continue to analyze the non-adjacent subsequent nodes of the changed node to find subsequent changed nodes until we reach the ending node, simplifying the feasible path structure and obtaining multiple feasible paths.

[0080] For any feasible path, combining the obtained feasible path, expected average speed and expected start time, calculate the instantaneous position of the bridge crane at each moment in the time-series traffic grid set sequence, determine whether the instantaneous position is in the traffic grid set at the corresponding moment, if all instantaneous positions are in the traffic grid set at the corresponding moment, determine that the feasible path can be applied and retain the feasible path; otherwise, clear the feasible path.

[0081] Calculate the actual length of the feasible path and take the absolute value of the difference between it and the actual length of the shortest path to obtain the length deviation of the feasible path. If the length deviation is less than the preset length deviation threshold, the feasible path is marked as the initial path and added to the initial path set; otherwise, the feasible path is cleared.

[0082] After traversing all the feasible paths obtained, the shortest path is marked as the initial path and added to the initial path set. Each initial path contains only a start node, a change node, and an end node. The start node, change node, and end node can be uniformly classified as path nodes. The path nodes of each initial path are integrated in sequence to obtain the path node sequence of each initial path.

[0083] It should be noted that the purpose of this step is to search and select an initial path set that meets the requirements of safety and efficiency based on the time grid state matrix and dynamic environmental constraints. The improved A* algorithm simplifies the path structure by identifying changing nodes. The construction of the time-series access grid set sequence incorporates the time dimension into the path search, upgrading path planning from static spatial obstacle avoidance to spatiotemporal dynamic obstacle avoidance, ensuring that the path is in the safe grid at every moment. The integration of the improved A* algorithm and the RRT algorithm balances path efficiency and robustness, overcoming the limitations of a single algorithm.

[0084] S3: For the initial path, the path tortuosity is quantitatively calculated, and the path tortuosity is used for screening, marking the preferred paths and organizing them into a set of preferred paths;

[0085] Specifically, for each initial path, the sequence of path nodes of the initial path is obtained. A vector calculation method is used to calculate the sequence of three consecutive path nodes on the initial path. , , Calculate the vector between any two adjacent nodes, where the vector direction is from the previous node to the next node. The cosine of the angle between the two vectors is calculated using the dot product of these vectors. Based on the obtained cosine value of the included angle, the included angle between the two vectors is calculated. The turning angle of the changed node is marked;

[0086] For each initial path, count the number of changed nodes (num) contained in the initial path, and calculate the turning angle of each changed node. Within the initial path set, all turning angles of each initial path are normalized to obtain the normalized turning angle of each changed node. , This represents the normalized turning angle of the changed node at the poi-th changed node in the corresponding initial path. The actual length of each initial path is calculated, and then normalized within the initial path set to obtain the normalized actual length of each initial path. , where rou represents the initial path number;

[0087] Calculate the path tortuosity of the initial path The formula is:

[0088] ;

[0089] It should be noted that the purpose of path tortuosity is to standardize the degree of tortuosity of paths of different lengths, so that the degree of tortuosity between initial paths is comparable laterally.

[0090] The mean path tortuosity of each initial path in the initial path set is calculated as the tortuosity threshold. The initial path set is then filtered based on the tortuosity threshold. For each initial path, if the path tortuosity of the initial path is less than the tortuosity threshold, the initial path is marked as a preferred path and included in the preferred path set.

[0091] It should be noted that the purpose of this step is to select paths with lower tortuosity from the initial path set, thereby reducing the steering energy consumption and load sway risk during crane operation. By quantifying the steering angle and length of the path, a path tortuosity index is constructed to achieve lateral comparability and scientific selection of different paths. The steering angle of continuous path nodes is calculated using vector calculation method to accurately quantify the local tortuosity characteristics of the path. Through normalization processing and path tortuosity formula, standardized comparison of paths with different lengths and steering characteristics is achieved, breaking through the one-sidedness of traditional selection based solely on length or number of turns.

[0092] S4: For the preferred path, analyze the load swing at the change node, calculate the load swing envelope range, mark the risk change node, and filter to obtain the optimal path;

[0093] A model of a loaded pendulum is established based on the Lagrange equations. The equations of motion for the loaded pendulum model are as follows:

[0094] ;

[0095] in, Indicates the load swing angle. Let represent the second derivative of the load swing angle, represent the oscillating inertial force, and LG represent the length of the wire rope lowered by the bridge crane. Indicates the load mass, and c represents the air damping coefficient. and These represent the accelerations of the control load of the bridge crane in the horizontal x and y directions, respectively.

[0096] Based on any preferred path, obtain the acceleration of each changed node within the preferred path. Based on the expected average speed and the turning angle of the changed node, calculate the acceleration of the control load of the bridge crane in the x and y directions of the horizontal plane. and The load swing angle is solved by numerical integration. By combining the length LG of the wire rope lowered by the bridge crane, the swing envelope range of the load is obtained;

[0097] Combining the preferred path, expected average speed, and expected start time, the instantaneous position of the bridge crane arrives at the change node at the moment of change and marks it as the change time. Taking the change time as the midpoint, a change window is taken. Based on the temporal access grid set sequence, the access grid set at each moment in the change window is obtained. The overlapping part of the access grid sets in the change window is organized into an overlapping grid set. It is determined whether the swing envelope range of the load on the change node exceeds the overlapping grid set. If it does, it is determined that there is a collision risk and the change node is designated as a risk change node.

[0098] For each preferred path, the number of risk change nodes within the preferred path is obtained, and the preferred paths are sorted from smallest to largest according to the number of risk change nodes to obtain a preferred path sequence. The first preferred path in the preferred path sequence is the optimal path found in the path planning, thus completing the planning of the movement path of the bridge crane.

[0099] It should be noted that the purpose of this step is to analyze the collision risk between the load and obstacles in the optimal path by combining the dynamic swing characteristics of the load, and finally select the path with the lowest risk. By establishing a load pendulum model, the swing range is calculated, and the safety is verified by combining the overlapping grid set of the change window, ensuring that the path is collision-free in actual operation. By introducing the change window and the overlapping grid set, the load swing range is spatiotemporally matched with the obstacle grid at different times, realizing the dual collision risk analysis of dynamic load swing and dynamic obstacles, upgrading from the safety of the path itself to the overall system safety including the load.

[0100] The technical solution of this invention is as follows: the motion environment of the bridge crane is uniformly divided into grid cells, dynamic obstacles are identified and a state value is assigned to each grid cell, the trajectory of the dynamic obstacles is predicted, a time grid state matrix including the state value of each grid cell at the corresponding time is generated, a time-series passable grid set sequence including the passable grid set is constructed based on the time grid state matrix, and the improved A* algorithm is used to explore the shortest path, the RRT algorithm is used to explore multiple feasible paths, change nodes are marked and multiple feasible paths are filtered based on the shortest path, the initial path is marked and the initial path set and path node sequence are organized, for the initial path, the path tortuosity of the initial path is quantitatively calculated and filtered based on the path tortuosity, the preferred path is marked and organized into a preferred path set, for the preferred path, the load swing at the change node is analyzed, the swing envelope range of the load is calculated, the risk change node is marked, and the optimal path is obtained by filtering.

[0101] Example 2:

[0102] like Figure 2 As shown in the embodiment of the present invention, a bridge crane motion path planning system includes the following modules:

[0103] Environment Modeling Module: The motion environment of the bridge crane is uniformly divided into grid cells, dynamic obstacles are identified and a state value is assigned to each grid cell, the trajectory of the dynamic obstacles is predicted, and a time grid state matrix including the state value of each grid cell at the corresponding time is generated.

[0104] Basic planning module: Based on the time grid state matrix, construct a temporal access grid set sequence including the access grid set, and use the improved A* algorithm to explore the shortest path, use the RRT algorithm to explore multiple feasible paths, mark change nodes and filter multiple feasible paths based on the shortest path, mark the initial path and organize the initial path set and path node sequence;

[0105] Preliminary screening module: For the initial path, the path tortuosity is quantitatively calculated, and the path tortuosity is screened based on the path tortuosity, the preferred paths are marked and organized into a set of preferred paths;

[0106] Load selection module: For the preferred path, analyze the load swing at the change node, calculate the load swing envelope range, mark the risk change node, and filter to obtain the optimal path.

[0107] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method of motion path planning for a bridge crane, characterized by: The application relates to a bridge crane motion environment planning method. The bridge crane motion environment is evenly divided into grid units, dynamic obstacles are identified, and a state value is given to each grid unit; the trajectory of the dynamic obstacle is predicted, and a time grid state matrix including the state value of each grid unit at a corresponding moment is generated; A time sequence passing grid set sequence including passing grid sets is constructed based on the time grid state matrix, an improved A* algorithm is used to explore the shortest path, an RRT algorithm is used to explore multiple feasible paths, a change node is marked, and the multiple feasible paths are screened based on the shortest path, an initial path is marked, and an initial path set and a path node sequence are arranged; For the initial path, the path tortuosity of the initial path is quantitatively calculated, and the path tortuosity is screened based on the path tortuosity, an optimal path is marked, and the optimal path set is arranged; For the optimal path, the load swing of the load at the change node is analyzed, the swing envelope range of the load is calculated, the risk change node is marked, and the optimal path is screened.

2. A method of motion path planning for a bridge crane as claimed in claim 1, characterized in that: The state value is obtained in the following manner: A three-dimensional coordinate system is established for the bridge crane motion environment, the motion environment is evenly divided into grid units, motion environment data are collected in real time, and obstacles are identified; a current moment is taken as an end point to set an obstacle analysis period; in the obstacle analysis period, the distance between the coordinates of the grid units covered by the obstacles at adjacent moments is calculated; if the distance at each two adjacent moments satisfies a preset condition, the obstacle is determined to be a static obstacle; otherwise, the obstacle is determined to be a dynamic obstacle; A state value is given to each grid unit; the state value of a grid unit not covered by any obstacle is 0; the state value of a grid unit covered by a static obstacle is 1; and the state value of a grid unit covered by a dynamic obstacle is 2.

3. A bridge crane motion path planning method according to claim 2, characterized in that: The time grid state matrix is obtained in the following manner: A future period with the current moment as a starting point is set, the coordinates of the grid units covered by the dynamic obstacles at each moment in the obstacle analysis period are obtained, the motion trajectory of the dynamic obstacle in the future period is predicted by using a Kalman filtering algorithm, the coordinates of the grid units covered by the dynamic obstacle at each moment in the future period are obtained, the state values of the grid units at each moment in the future period are arranged and converted to obtain a time grid state matrix.

4. The method of claim 1, wherein: The time sequence passing grid set sequence is obtained in the following manner: A path planning including a starting point grid unit, an end point grid unit and an obstacle safety distance is set; at each moment in the future period, all grid units with a state value of 0 are obtained, and the grid units are marked as empty grid units; the grid units with a state value not equal to 0 are marked as obstacle grid units, and the minimum Euclidean distance between the empty grid units and the obstacle grid units is calculated; if the distance is greater than the obstacle safety distance, the empty grid unit is marked as a passing grid unit and is included in the passing grid set; The bridge crane is preset with an expected starting moment, the passing grid set before the expected starting moment is cleared, and a time sequence passing grid set sequence is arranged.

5. A bridge crane motion path planning method according to claim 4, characterized in that: The shortest path is obtained in the following manner: Each grid cell center point is regarded as a node, wherein the start grid cell and the end grid cell are the start node and the end node respectively, the start passing grid set of the time sequence passing grid set sequence is obtained, the improved A* algorithm is used for path searching in the start passing grid set, a basic shortest path is obtained, whether a line segment between the nodes and non-adjacent subsequent nodes in the shortest path is in the start passing grid set is analyzed, and a change node is marked based on an analysis result, so as to obtain the shortest path; The shortest path is obtained by setting an expected average speed for the bridge crane, combining the shortest path and an expected starting time, marking a grid cell in which a spreader center is located as an instant position, and calculating the instant position of the bridge crane at each time in the time sequence passing grid set sequence. If the instant position is not in the passing grid set at the corresponding time, the node corresponding to the instant position is removed in the start passing grid set, and path searching is performed again. Otherwise, the shortest path is obtained.

6. A bridge crane motion path planning method according to claim 5, characterized in that: The path node sequence is obtained in the following manner: The start passing grid set of the time sequence passing grid set sequence is obtained, the RRT algorithm is used for path searching in the start passing grid set by multiple independent running, a plurality of feasible paths are obtained, whether a line segment between the nodes and non-adjacent subsequent nodes in the feasible path is in the start passing grid set is analyzed, and the change node is marked and the feasible path is simplified and updated based on the analysis result; For any feasible path, whether the instant position is in the passing grid set at the corresponding time is judged by combining the expected average speed and the expected starting time. If the instant position is not in the passing grid set at the corresponding time, the feasible path is removed. The actual lengths of the feasible path and the shortest path are calculated and data processing is performed, so as to obtain the length deviation of the feasible path. If the length deviation meets a preset condition, the feasible path is marked as an initial path. After all the obtained feasible paths are traversed, the shortest path is marked as the initial path. The nodes of the initial path are integrated, so as to obtain the path node sequence of each initial path.

7. The method of motion path planning for a bridge crane of claim 1, wherein: The path tortuosity is obtained in the following manner: For each initial path, the path node sequence of the initial path is obtained. The turning angle of the change node is calculated by using a vector calculation method. The actual length of each initial path is calculated. Data processing is performed by combining the number of change nodes in the initial path. The path tortuosity of the initial path is calculated.

8. A bridge crane motion path planning method according to claim 7, characterized in that: The preferred path is obtained in the following manner: The mean value of the path tortuosity of each initial path in the initial path set is calculated as a tortuosity threshold. Based on the tortuosity threshold, the initial path set is screened. For each initial path, if the path tortuosity of the initial path is less than the tortuosity threshold, the initial path is marked as a preferred path.

9. A method of motion path planning for a bridge crane as claimed in claim 5, characterized in that: The optimal path is obtained in the following manner: The motion equation of the load pendulum model is established based on the Lagrange equation. Each change node in the preferred path is obtained. The acceleration of the bridge crane control load in the x direction and the y direction of the horizontal plane is respectively calculated based on the expected average speed and the turning angle of the change node. The load swing angle is solved by numerical integration. The length of the steel wire rope dropped by the bridge crane is obtained and combined, so as to obtain the swing envelope range of the load. The time when the instant position of the bridge crane reaches the change node is calculated in combination with the preferred path, the expected average speed, and the expected starting time, and is marked as a change time. A change window is taken with the change time as a middle point, the overlapping part of the passing grid set in the change window is arranged into an overlapping grid set, if the swing envelope range of the load on the change node exceeds the overlapping grid set, the change node is a risk change node, and the preferred path with the minimum number of risk change nodes is the optimal path.

10. A bridge crane motion path planning system characterized by: The method comprises the following modules: An environment modeling module: uniformly divides the moving environment of the bridge crane into grid units, identifies dynamic obstacles, assigns a state value to each grid unit, performs trajectory prediction on the dynamic obstacles, and generates a time grid state matrix comprising the state value of each grid unit at a corresponding time; A basic planning module: constructs a time sequence passing grid set sequence comprising passing grid sets based on the time grid state matrix, explores a shortest path using an improved A* algorithm, explores multiple feasible paths using an RRT algorithm, marks change nodes, and screens the multiple feasible paths based on the shortest path; A preliminary screening module: quantitatively calculates the path tortuosity of the initial path, screens based on the path tortuosity, marks a preferred path, and arranges the preferred path into a preferred path set; A load selection module: analyzes the load swing of the load at the change node for the preferred path, calculates the swing envelope range of the load, marks risk change nodes, and screens to obtain an optimal path.

Images