Device, method, equipment and medium for detecting manufacturing anomalies of a marine propulsion device

Abstract

The application discloses a kind of detection device, method, equipment and medium of ship propulsion device manufacturing exception, and the application belongs to the technical field of the propulsion device of ship.The detection device includes: point cloud data acquisition module, for obtaining the point cloud data of propeller blade;Blade angle identification module is used to identify the angle between the multiple blades of propeller according to point cloud data;Point cloud data comparison module is used to determine comparison point cloud data from each blade according to angle, and comparison result is obtained by comparison;Consistency determination module is used to determine whether blade parameter has consistency according to comparison result between comparison point cloud data;Report information generation module is used to generate the report information of consistency exception when blade parameter does not have consistency.The scheme can detect the manufacturing exception of marine propeller by comparing the measured propeller and standard point cloud data, improve the efficiency and precision of marine propeller detection, save manpower and time.

Description

Technical Field

[0001] This application belongs to the field of ship propulsion technology, specifically relating to a detection device, method, equipment, and medium for manufacturing abnormalities in ship propulsion devices. Background Technology

[0002] As the most important propulsion device for ships, the propeller converts the rotational power of the engine into propulsion force. It is an indispensable part of the ship's power system and has high quality requirements. Therefore, marine propellers need to undergo rigorous qualification testing to ensure their normal operation.

[0003] Currently, the inspection of marine propellers often involves manual checks of the length, angle, and curvature of each blade. This method requires highly skilled personnel and is inefficient. Furthermore, if this inspection is time-consuming during the propeller manufacturing process, it can significantly deplete the overall manufacturing time. Therefore, how to perform high-precision and efficient inspection of manufacturing anomalies in marine propellers is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this application is to provide a detection device, method, equipment, and medium for manufacturing anomalies in marine propulsion devices, which can improve the efficiency and accuracy of testing marine propellers and reduce the requirements for personnel.

[0005] In a first aspect, embodiments of this application provide a detection device for manufacturing anomalies in a ship propulsion device, the device comprising:

[0006] The point cloud data acquisition module is used to acquire point cloud data of the propeller blades;

[0007] The blade angle recognition module is used to identify the angle between multiple blades of the propeller based on the point cloud data.

[0008] The point cloud data comparison module is used to determine the comparison point cloud data from each blade according to the included angle, and to compare them to obtain the comparison result;

[0009] The consistency determination module is used to determine whether the blade parameters are consistent based on the comparison results between the comparison point cloud data.

[0010] The report information generation module is used to generate a report information of consistency anomaly when the blade parameters are inconsistent.

[0011] Furthermore, the device also includes:

[0012] The blade parameter identification module is used to identify the blade curvature and blade length of multiple blades of the propeller based on the point cloud data.

[0013] The extraction boundary determination module is used to determine the extraction boundary of the comparative point cloud data from each blade based on the blade curvature and the blade length.

[0014] Furthermore, the extraction boundary determination module is also used for:

[0015] The width boundary of the comparison point cloud data is determined based on the blade curvature.

[0016] as well as,

[0017] The length boundary of the comparison point cloud data is determined based on the blade length.

[0018] Furthermore, the point cloud data comparison module is specifically used for:

[0019] Based on the included angle, comparative point cloud data is determined from each blade;

[0020] The comparison point cloud data is labeled according to the blade to which it belongs;

[0021] Rotate the comparison point cloud data according to the included angle to obtain point cloud information with corresponding relationships;

[0022] The comparison results are determined based on the distance information between points with corresponding point cloud information.

[0023] Furthermore, the point cloud data comparison module is also used for:

[0024] Determine the comparison cells for the comparison point cloud data on each blade;

[0025] Based on the included angle and the location information of the comparison cell, the points in the comparison cell are spatially transformed to obtain the distance information between each point, and the comparison result is determined.

[0026] Furthermore, the device also includes:

[0027] A hub point cloud data recognition module is used to recognize hub point cloud data in the point cloud data;

[0028] The rotor hub comparison result acquisition module is used to compare the rotor hub point cloud data with the pre-determined rotor hub model data to obtain the rotor hub comparison result.

[0029] The propeller hub manufacturing standard judgment module is used to determine whether the propeller hub meets the manufacturing standards based on the comparison results of the propeller hubs.

[0030] Furthermore, the propeller hub comparison result acquisition module is also used for:

[0031] The rotor hub model data is spatially transformed to obtain a spatial pose that matches the rotor hub point cloud data;

[0032] Identify the planar and surface data of the propeller hub model in the current spatial pose;

[0033] The planar point cloud data and the curved point cloud data of the rotor hub are compared with the planar point cloud data and the curved point cloud data of the rotor hub to obtain the rotor hub comparison result.

[0034] Secondly, embodiments of this application provide a method for detecting manufacturing anomalies in a ship propulsion device, the method comprising:

[0035] Obtain point cloud data of the propeller blades;

[0036] Based on the point cloud data, the included angle between the multiple blades of the propeller is identified;

[0037] Based on the included angle, the comparison point cloud data is determined from each blade, and the comparison results are obtained.

[0038] Based on the comparison results between the point cloud data, it is determined whether the blade parameters are consistent.

[0039] If not, a consistency anomaly report will be generated.

[0040] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0041] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0042] Fifthly, embodiments of this application provide a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the method as described in the first aspect.

[0043] In this embodiment, point cloud data of the propeller blades is acquired; based on the point cloud data, the included angles between multiple blades of the propeller are identified; comparative point cloud data is determined from each blade based on the included angles, and a comparison result is obtained; based on the comparison result between the comparative point cloud data, it is determined whether the blade parameters are consistent; if not, a report of consistency anomaly is generated. Through the above-described method for detecting manufacturing anomalies in marine propulsion devices, consistency detection of manufacturing anomalies in marine propellers can be performed by establishing and comparing point cloud data of the tested propeller and a standard propeller, thereby improving the efficiency and accuracy of marine propeller inspection and saving manpower and time. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities of ship propulsion devices provided in Embodiment 1 of this application;

[0045] Figure 2 This is a schematic diagram of the structure of the marine propeller provided in Embodiment 1 of this application;

[0046] Figure 3 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities of ship propulsion devices provided in Embodiment 2 of this application;

[0047] Figure 4 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities of ship propulsion devices provided in Embodiment 3 of this application;

[0048] Figure 5 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities of ship propulsion devices provided in Embodiment 4 of this application;

[0049] Figure 6 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities of ship propulsion device provided in Embodiment 5 of this application;

[0050] Figure 7 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities of ship propulsion devices provided in Embodiment Six of this application;

[0051] Figure 8 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities of ship propulsion devices provided in Embodiment 7 of this application;

[0052] Figure 9 This is a flowchart illustrating the method for detecting manufacturing abnormalities in a ship propulsion device provided in Embodiment 8 of this application;

[0053] Figure 10 This is a schematic diagram of the structure of the electronic device provided in Embodiment 9 of this application. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0055] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0056] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0057] The following description, in conjunction with the accompanying drawings, details the detection device, method, equipment, and medium for manufacturing anomalies of ship propulsion devices provided in this application, through specific embodiments and application scenarios.

[0058] Example 1

[0059] Figure 1 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities in ship propulsion devices provided in Embodiment 1 of this application. Figure 1 As shown, the specific steps include the following:

[0060] Point cloud data acquisition module 110 is used to acquire point cloud data of propeller blades;

[0061] The blade angle recognition module 120 is used to identify the angle between multiple blades of the propeller based on the point cloud data.

[0062] The point cloud data comparison module 130 is used to determine the comparison point cloud data from each blade according to the included angle, and compare them to obtain the comparison result;

[0063] The consistency determination module 140 is used to determine whether the blade parameters are consistent based on the comparison results between the comparison point cloud data.

[0064] The report information generation module 150 is used to generate a report information of consistency anomaly when the blade parameters are inconsistent.

[0065] First, the application scenario of this solution can be that a smart terminal or testing platform compares the point cloud data of the tested marine propeller with the point cloud data of a standard propeller, or compares the consistency of each blade based on the point cloud data of the tested marine propeller, in order to determine whether the tested marine propeller has manufacturing abnormalities.

[0066] Based on the above application scenarios, it is understood that the executing entity of this application can be the smart terminal or the testing platform, or it can be the testing software running in the smart terminal, without further limitations here.

[0067] In this design, a marine propeller can be used to convert engine rotational power into propulsion. The marine propeller can be structurally divided into a blade section and a hub section. The number of blades can include two or more, and each blade can be connected to the hub. The rearward-facing surface of each blade is a helical surface or a near-helical surface, making it a marine propulsion device.

[0068] The point cloud data acquisition module 110 can be a program designed for a smart terminal to acquire point cloud data of propeller blades. Point cloud data can be a set of point data obtained by measuring the surface of a product using measuring instruments during reverse engineering, after acquiring the spatial coordinates of each sampling point on the object's surface. Reverse engineering can be a method that uses a 3D laser scanner or coordinate measuring machine to perform 3D scanning or measurement on the object's surface to obtain the object's 3D point data, then uses reverse engineering software to organize, edit, and obtain the required 3D feature curves, ultimately expressing the object's shape through a 3D surface.

[0069] In this solution, the point cloud data of the propeller blades can be obtained through 3D laser scanning technology. Specifically, it can utilize the principle of laser ranging. By emitting a laser towards the propeller under test, the distance to each data point on the propeller is calculated based on the laser's propagation speed and the time it takes to receive the laser. Based on the distance and time information, a large number of dense points on the surface of the propeller under test, along with their 3D coordinates, reflectivity, and texture information, are obtained. After acquiring the information of the points on the surface of the propeller under test, a smart terminal filters these dense points, removing redundant stray points to obtain the point cloud data.

[0070] The blade angle recognition module 120 can be a program designed by a smart terminal to recognize the blade angle. The angle between blades can be the smallest angle formed by the intersection of two adjacent blades of a marine propeller at the hub. In the calculation process, it can be replaced by the smallest angle formed by the intersection of two propeller blades at the same position, such as the centerline position, at the hub.

[0071] Figure 2 This is a schematic diagram of the structure of the marine propeller provided in Embodiment 1 of this application, as shown below. Figure 2 As shown, angle α is the included angle between the blades. The hub, which is the part where the blades are mounted and joined, is located at the center of the propeller and is usually conical or frustum-shaped, driving the propeller to rotate.

[0072] One way to identify the angle between propeller blades is to first determine the extraction boundary of the point cloud data, distinguish the position of each propeller blade according to the boundary, and then select a point at the same position of the two propeller blades to be identified on the plane where the hub section is located. Take the line connecting the center point of the hub section and the two propeller blade points, and calculate the minimum angle formed by the two lines relative to the center point of the hub as the angle of the blade to be identified.

[0073] In this scheme, the propeller blade angle recognition module can use a pre-built 3D convolutional neural network (3DCNN) model. Specifically, high-precision point cloud data of the propeller blades can be obtained using devices such as LiDAR and structured light scanning. To improve the model's generalization ability, data needs to be collected from various aspects, including different angles, different lighting conditions, and different propeller models. The collected point cloud data is preprocessed, including noise removal, data normalization (normalizing the coordinate range of the point cloud data to a certain interval, such as [0, 1]), and data augmentation (expanding the dataset by random rotation, translation, scaling, etc., to increase data diversity). The true values ​​of the angles between multiple blades in each propeller point cloud data are manually or using semi-automatic tools as supervision information for model training. The structure of the 3D convolutional neural network model can include an input layer that takes the preprocessed point cloud data as input. Since it is 3D data, the input format can be an array of point coordinates (x, y, z). Convolutional layers use 3D convolutional kernels to perform convolution operations on point cloud data to extract local features at different scales. For example, multiple convolutional layers can be set, gradually increasing the number and size of the convolutional kernels, enabling the model to learn more complex features. During convolution, the stride controls the movement of the convolutional kernel to adjust the size of the feature map. Pooling layers use operations such as max pooling or average pooling to downsample the feature map output by the convolutional layers, reducing the amount of data while retaining important feature information. Pooling operations can reduce the computational complexity of the model and prevent overfitting. Fully connected layers unfold the feature map after convolution and pooling into a one-dimensional vector and input it into the fully connected layer. The fully connected layer further transforms and combines the features through matrix multiplication and activation functions to extract global features. The output layer, according to the task requirements of angle recognition, outputs the predicted angle values ​​between multiple blades. The output layer can use linear activation functions to directly output continuous angle values. Specifically, an appropriate loss function can be chosen to measure the difference between the model's predicted values ​​and the true values, such as the mean squared error (MSE) loss function, used in regression tasks to calculate the average squared error between the predicted and true angles. Optimizers such as stochastic gradient descent (SGD) and Adam are used to update the model's parameters based on the gradient of the loss function to minimize it. During training, appropriate hyperparameters such as the learning rate and number of iterations can be set. During training, the labeled point cloud data can be divided into training, validation, and test sets. The model is iteratively trained on the training set, with a batch of point cloud data input for each training iteration, calculating the loss function and updating the model parameters. The model's performance is evaluated on the validation set, and hyperparameters are adjusted based on the validation results to prevent overfitting. Once the model's performance on the validation set is stable, a final evaluation is performed on the test set. Metrics such as mean squared error (MSE) and mean absolute error (MAE) are used to evaluate the model's prediction accuracy.By calculating these metrics, we can understand the average error between the model's predicted angle value and the actual value. The trained model is then applied to a real-world propeller blade angle recognition task. Newly acquired point cloud data is input, and the model outputs the predicted blade angle value, providing a basis for subsequent analysis and decision-making.

[0074] The point cloud data comparison module 130 can be a program designed for a smart terminal to compare point cloud data. The point cloud data comparison can be based on reference point cloud data selected from a specific area at the same position on each blade, used to determine whether the tested propeller blade is consistent with a standard propeller blade. Specifically, the point cloud data contained in the entire blade can be compared with the point cloud data of the standard propeller blade. However, because the point cloud data of the entire blade is too dense, the computational load is large and the computation time is long when performing consistency comparison detection. Therefore, in this solution, a specific area of ​​the blade can also be selected, and the area can be rotated multiple times according to the blade angle to obtain the regional point cloud data of each blade of the propeller, forming the comparison point cloud data for calculation to reduce the computational load. For example, Figure 2 Region 1 is selected, and it is rotated four times according to the blade angle to obtain regions 2, 3, and 4 at the same positions on each blade. The point cloud data in regions 1, 2, 3, and 4 constitute the comparative point cloud data of the propeller blade. The comparison result can be the similarity between the reference point cloud data of the tested propeller blade and the reference point cloud data of the standard propeller blade.

[0075] Another comparison method is to compare the consistency between the individual blades of the propeller. Specifically, a region can be selected at the same position on each blade, the normal vector of the plane containing each region can be calculated, and the direction of the normal vector of each region can be compared to see if there is any deviation. In this embodiment, the comparison method can also be to select a region at the same position on each blade, rotate the selected region of one blade according to the angle between the blades, and compare the similarity between that region and the region at the same position on the adjacent blade.

[0076] The method for determining the comparison point cloud data can be as follows: Select a propeller blade in the measurement plane and acquire point cloud data of a random area on the blade. After rotating the propeller under test and measuring the angle between the blades, acquire point cloud data of the original area again. Repeat this process the same number of times as the number of blades to form a set of comparison point cloud data that can be compared with the point cloud data of the same position of a qualified propeller. The method for determining the measurement plane can be as follows: Randomly select three non-collinear points in the propeller's point cloud data. Use these three points to define a plane, and find points in the propeller's point cloud data whose distance to this plane is less than a preset distance. Repeat this process, updating the selected three points to define a new plane. The plane with the most points whose distance to this plane is less than the preset distance is selected as the measurement plane. The method for determining the comparison result can be as follows: Compare the reference point cloud data of the tested propeller blade with the reference point cloud data of the standard propeller blade, calculate the similarity between the two, and use the similarity score to represent the comparison result.

[0077] In this solution, the point cloud data comparison module can use local feature descriptors (such as feature descriptors in PointNet++) to extract local features of each blade point cloud data.

[0078] PointNet++ is a deep learning model developed based on PointNet for processing point cloud data. Through multi-level local feature extraction, it can better capture the local geometric structure and semantic information of point cloud data. PointNet++ can acquire point cloud data for each propeller blade. Similar to the data preprocessing of the blade angle recognition module, it performs noise removal and normalization operations to ensure data quality and consistency. The point cloud data for each blade is organized into a suitable format for input into the PointNet++ model. Point cloud data can be represented as a set of point coordinates, with each point containing its position information (x, y, z) in three-dimensional space. This scheme uses the farthest point sampling (FPS) algorithm to select a certain number of representative points from the input point cloud data, reducing the amount of point cloud data while maintaining the data distribution characteristics. These sampled points will serve as the center for subsequent local feature extraction. Centered on the sampled points, the surrounding points are divided into different local regions (neighborhoods) by setting a radius or the number of neighboring points. The point cloud data within each local region will be used to extract local features. For each local region of point cloud data, the PointNet network structure (including a multilayer perceptron and max pooling operation) is used to extract local features. The PointNet network transforms and combines the features of each point using the multilayer perceptron, and then uses max pooling to extract the global features of that local region, serving as the feature descriptor for that region. The extracted local features are then mapped back to the original point cloud data through feature propagation, ensuring that each point possesses a feature descriptor for its local region. In this way, each point in each blade point cloud contains both geometric and semantic information about its surrounding local region.

[0079] Specifically, based on the local feature descriptors extracted from each blade's point cloud data, feature matching algorithms (such as nearest neighbor matching, distance-based matching, etc.) can be used to find the corresponding local regions on different blades. By comparing the feature descriptors of these corresponding local regions, the similarities and differences between different blades can be determined. Comparative analysis is performed on the feature descriptors of the matched local regions, and difference metrics (such as Euclidean distance, cosine similarity, etc.) are calculated to obtain point cloud data comparison results. These comparison results can be used to determine the consistency of blade parameters, providing a basis for subsequent decision-making. This approach effectively constructs a 3D convolutional neural network model for blade angle recognition and utilizes feature descriptors in PointNet++ to extract local features from point cloud data for comparative analysis.

[0080] This embodiment provides the following specific methods for point cloud comparison for reference:

[0081] Calculate the feature descriptor factors for both the original and rotated blade point cloud data. Based on these feature descriptor factors, compare the points in both point clouds, randomly selecting three or more pairs of comparison points. Calculate the rotation and translation matrices for each comparison and then calculate the corresponding error. Repeat the steps of selecting comparison points, calculating the transformation matrix, and calculating the error. Use the rotation and translation with the smallest error as the final result and calculate the degree of overlap between the original and rotated blade point cloud data. The core code for this method can be:

[0082] pcl::SampleConsensusInitialAlignment<pcl::PointXYZ,pcl::PointXYZ,pcl::FPF HSignature33> sac_ia;

[0083] sac_ia.setInputSource(source);

[0084] sac_ia.setSourceFeatures(source_fpfh);

[0085] sac_ia.setInputTarget(target);

[0086] sac_ia.setTargetFeatures(target_fpfh);

[0087] sac_ia.setMinSampleDistance(0.1);

[0088] sac_ia.setCorrespondenceRandomness(6);

[0089] pointcloud::Ptr align(new pointcloud);

[0090] sac_ia.align(*align);

[0091] In this scheme, principal component analysis can also be used to compare point cloud data. For example, the principal axis directions of the original blade point cloud data and the rotated blade point cloud data can be compared. First, the covariance matrix of the two sets of point cloud data is calculated. Based on the covariance matrix, the main eigencomponents, i.e., the principal axis directions of the point cloud data, are calculated. Then, the rotation matrix is ​​obtained from the principal axis directions, and the translation vector is directly obtained by calculating the offset of the center coordinates of the two sets of point clouds. Finally, the overlap between the original blade point cloud data and the rotated point cloud data is calculated after matrix transformation. The core code of this method can be:

[0092]

[0093] Using a method similar to the one described above, the blade point cloud data can be compared.

[0094] The consistency determination module 140 can be a program designed by a smart terminal to determine whether the blade parameters are consistent. Blade parameters can be the geometric parameters of the propeller blades, including the number of blades, the angle between blades, and the blade length. Consistency can be a criterion describing whether the parameters of the tested propeller are consistent with those of a qualified propeller. When the consistency requirement is not met, the tested propeller fails the qualification test.

[0095] One way to determine whether the blade parameters are consistent is to pre-set a condition where the comparison result reaches a preset value. When the smart terminal obtains the comparison result between the comparison point cloud data, it compares this result with the preset value to determine whether the tested propeller parameters are consistent. For example, the comparison result is set to determine that the blade parameters are consistent when it reaches 97%. If the current comparison result is 93%, which does not reach the preset value, then the blade parameters are not consistent. If the comparison result is between 93% and 97%, a manual re-inspection instruction can be generated for staff to re-inspect it.

[0096] The report information generation module 150 is a program designed for use on a smart terminal to generate consistency anomaly reports. The report information can include notifications to staff that the tested propeller is substandard. It may include the propeller's serial number and whether the propeller exhibits consistency anomalies. For example, a marine propeller, number 399, with a blade angle of 43°, exhibits a consistency anomaly and fails the inspection.

[0097] One way to generate a consistency anomaly is for the testing platform to receive a signal indicating inconsistency and then send a report of the inconsistency to the worker's smart terminal device. This smart terminal device can be a multimedia device connected to the testing platform for monitoring the testing process and receiving test results, such as a mobile phone, computer, or tablet.

[0098] Understandably, in another embodiment, if the comparison results show consistency, a consistency report can also be generated to inform staff of the final results of the current marine propeller inspection.

[0099] In this embodiment, a point cloud data acquisition module is used to acquire point cloud data of the propeller blades; a blade angle identification module is used to identify the angles between multiple blades of the propeller based on the point cloud data; a point cloud data comparison module is used to determine comparison point cloud data from each blade based on the angles and compare them to obtain a comparison result; a consistency determination module is used to determine whether the blade parameters are consistent based on the comparison result between the comparison point cloud data; and a report information generation module is used to generate a consistency anomaly report when the blade parameters are inconsistent. Through the above-described detection device for manufacturing anomalies in marine propulsion devices, consistency detection of manufacturing anomalies in marine propellers can be performed by establishing and comparing point cloud data of the tested propeller and a standard propeller, thereby improving the efficiency and accuracy of marine propeller inspection and saving manpower and time.

[0100] Example 2

[0101] Figure 3 This is a schematic diagram of the detection device for manufacturing anomalies in a ship propulsion device provided in Embodiment 2 of this application. This solution makes a further improvement on the above embodiment, specifically: the device further includes: a blade parameter identification module, used to identify the blade curvature and blade length of multiple blades of the propeller based on the point cloud data; and an extraction boundary determination module, used to determine the extraction boundary of the comparative point cloud data from each blade based on the blade curvature and blade length.

[0102] like Figure 3 As shown, it specifically includes the following:

[0103] Point cloud data acquisition module 310 is used to acquire point cloud data of propeller blades;

[0104] The blade angle recognition module 320 is used to identify the angle between multiple blades of the propeller based on the point cloud data.

[0105] The point cloud data comparison module 330 is used to determine the comparison point cloud data from each blade according to the included angle, and to compare them to obtain the comparison result.

[0106] The consistency determination module 340 is used to determine whether the blade parameters are consistent based on the comparison results between the comparison point cloud data.

[0107] The report information generation module 350 is used to generate a report information of consistency anomaly when the blade parameters are inconsistent.

[0108] The blade parameter identification module 360 ​​is used to identify the blade curvature and blade length of multiple blades of the propeller based on the point cloud data.

[0109] The extraction boundary determination module 370 is used to determine the extraction boundary of the comparative point cloud data from each blade based on the blade curvature and the blade length.

[0110] The blade parameter recognition module 360 ​​can be a program designed for a smart terminal to identify blade curvature and blade parameters. Blade curvature can be the rate of rotation of the tangent angle at a point on the curved edge of the blade away from the hub with respect to the arc length; it is defined by differentiation and indicates the degree to which the curve deviates from a straight line. Blade length can be the distance between the edge curve point of the blade away from the hub and the center point of the hub in the blade point cloud data on the plane containing the hub section.

[0111] One method to identify blade curvature is to select a random point on the curved edge of the blade away from the hub, calculate the rotation rate of the tangent direction angle with respect to the arc length at that point, and use this as the blade curvature at that point. Another method to identify blade length is to select a random point on the edge curve of the blade away from the hub from the blade point cloud data on the plane containing the hub section, and measure the distance of this point from the hub center point. This process is repeated, updating the selected measurement points, and choosing the point farthest from the hub center point; this distance is then used as the blade length.

[0112] The extraction boundary determination module 370 can be a program design for a smart terminal to determine the extraction boundary of the comparison point cloud data. The extraction boundary of the comparison point cloud data can be the boundary of the selected comparison point cloud data, which can be the length and width boundaries of the extracted comparison point cloud rectangle. It is understood that the shape of the extraction boundary of the comparison point cloud data is not limited to a rectangle, but can also be a square, a circle, or any other arbitrary shape.

[0113] One method to determine the extraction boundaries of point cloud data for comparison is to determine the width boundary of the point cloud data based on the blade curvature and the length boundary based on the blade length. The width boundary of the point cloud data can be the length of the line segment on the shorter side of the extraction boundary. The length boundary of the point cloud data can be the length of the line segment on the longer side of the extraction boundary.

[0114] For example, if the measured blade length is 2m, then 2m is used as the length boundary. Regarding the determination of the width boundary, the curvature may differ at different positions on a blade. One approach is to use the same width boundary for all blades, for example, uniformly setting the width boundary to 3cm; alternatively, one approach is to calculate the width boundary separately for each region, for example, setting it to be less than 0.5m when the curvature is less than 0.5m. -1 If 5cm is defined as the width boundary, then when the curvature reaches 0.5m... -1 In this case, 2.5cm is used as the width boundary.

[0115] The advantage of this scheme is that the extraction boundary of point cloud data can be determined based on the curvature and length of the propeller blades, which limits the selection range of point cloud data. It can remove some stray points outside the propeller, so that the point cloud data of the propeller can more realistically and accurately reflect the blade parameters of the propeller being tested.

[0116] Example 3

[0117] Figure 4 This is a schematic diagram of the structure of the detection device for manufacturing anomalies in a ship propulsion device provided in Embodiment 3 of this application. This solution makes a further improvement to the above embodiment, specifically: the extraction boundary determination module is further configured to: determine the width boundary of the comparison point cloud data based on the blade curvature; and determine the length boundary of the comparison point cloud data based on the blade length.

[0118] like Figure 4 As shown, it specifically includes the following:

[0119] Point cloud data acquisition module 410 is used to acquire point cloud data of propeller blades;

[0120] The blade angle recognition module 420 is used to identify the angle between multiple blades of the propeller based on the point cloud data.

[0121] The point cloud data comparison module 430 is used to determine the comparison point cloud data from each blade according to the included angle, and compare them to obtain the comparison result;

[0122] The consistency determination module 440 is used to determine whether the blade parameters are consistent based on the comparison results between the comparison point cloud data.

[0123] The report information generation module 450 is used to generate a report information of consistency anomaly when the blade parameters are inconsistent.

[0124] The blade parameter identification module 460 is used to identify the blade curvature and blade length of multiple blades of the propeller based on the point cloud data.

[0125] The extraction boundary determination module 470 is used to determine the extraction boundary of the comparative point cloud data from each blade based on the blade curvature and the blade length.

[0126] The boundary determination module 470 is further configured to determine the width boundary of the comparison point cloud data based on the blade curvature.

[0127] as well as,

[0128] The length boundary of the comparison point cloud data is determined based on the blade length.

[0129] One way to determine the width boundary of the point cloud data for comparison is to obtain parameters such as the curvature circle and radius of curvature based on the curvature. Then, the arc length of the portion where the curvature circle coincides with the blade edge curve is calculated using these parameters. The chord length corresponding to this arc length within the curvature circle is used as the width boundary of the point cloud data. Understandably, when the blade curvature is large, the selected point cloud data width boundary is smaller; when the blade curvature is small, the selected point cloud data width boundary is larger. For example, if the smart terminal detects that the curvature of the propeller being tested is 0.2, then... Figure 2 As shown, a curvature circle with a radius of curvature of 5 cm can be obtained. The chord length corresponding to the arc length of the part of the curvature circle that coincides with the blade edge curve is 5 cm. Therefore, the width boundary of the point cloud data is 5 cm.

[0130] One way to determine the length boundary of the comparison point cloud data is to obtain the length value of the propeller blades. For example, if the smart terminal detects that the length of the propeller blades is 1 meter, then the length boundary of the comparison point cloud data is 1 meter. Figure 2 The length boundary is shown.

[0131] The advantage of this scheme is that it limits the selection range of point cloud data, enabling the propeller's point cloud data to more realistically and accurately reflect the blade parameters of the propeller being tested.

[0132] Example 4

[0133] Figure 5 This is a schematic diagram of the detection device for manufacturing anomalies in a ship propulsion device provided in Embodiment 4 of this application. This solution makes a further improvement on the above embodiment, specifically: the point cloud data comparison module is specifically used for: determining comparison point cloud data from each blade according to the included angle; labeling the comparison point cloud data according to the blade to which it belongs; rotating the comparison point cloud data according to the included angle to obtain point cloud information with corresponding relationships; and determining the comparison result based on the distance information between the points with corresponding relationship point cloud information.

[0134] like Figure 5 As shown, it specifically includes the following:

[0135] The point cloud data acquisition module 510 is used to acquire point cloud data of the propeller blades;

[0136] The blade angle recognition module 520 is used to identify the angle between multiple blades of the propeller based on the point cloud data.

[0137] The point cloud data comparison module 530 is used to determine the comparison point cloud data from each blade according to the included angle, and compare them to obtain the comparison result;

[0138] The consistency determination module 540 is used to determine whether the blade parameters are consistent based on the comparison results between the comparison point cloud data.

[0139] The report information generation module 550 is used to generate a report information of consistency anomaly when the blade parameters are inconsistent.

[0140] Among them, the point cloud data comparison module 530 is specifically used to determine the comparison point cloud data from each blade according to the included angle;

[0141] The comparison point cloud data is labeled according to the blade to which it belongs;

[0142] Rotate the comparison point cloud data according to the included angle to obtain point cloud information with corresponding relationships;

[0143] The comparison results are determined based on the distance information between points with corresponding point cloud information.

[0144] One way to determine the comparative point cloud data from each blade is to determine the approximate outline of the propeller based on the angle between the blades and the extraction boundary of each blade, and to distinguish the region where each blade of the propeller is located and the point cloud data within that region.

[0145] The blade to which the data belongs can be the blade on which each point cloud data point belongs. Blade labeling can be the division of blades on a propeller, indicating which blade each point cloud data point belongs to.

[0146] One method for labeling propeller blades is to assign a number to each blade and the region it belongs to, and then label the point cloud data within that region with the blade's number. For example, if the four blades of a propeller are labeled with numbers 1, 2, 3, and 4, then all point cloud data within the region containing blade 1, such as points x1(1,1,1) and x2(1,1,2), would be labeled as 1.

[0147] Point cloud information with corresponding relationships can be point cloud data with the same blade number label before and after rotation.

[0148] One way to obtain corresponding point cloud information is to rotate the comparison point cloud data according to the angle between the blades, and then obtain point cloud data with the same labels before and after the rotation. For example, if the angle between the blades is 72°, rotating the comparison point cloud data by 72° will yield rotated comparison point cloud data. Point cloud data with the same labels and the same position in their respective blades before and after the rotation will be used as corresponding point cloud information.

[0149] The distance information between points in a corresponding point cloud can be calculated by determining the spatial distance between the points. For example, if the spatial coordinates of points in a pair of corresponding point clouds before and after rotation are (x1, y1, z1) and (x2, y2, z2), then the distance information for those points is:

[0150]

[0151] Using an algorithm similar to the one described above, distance information between points with corresponding point cloud information can be obtained.

[0152] One way to determine the comparison results is to compare the distance information between points with corresponding point cloud information with the distance information between points with corresponding point cloud information in qualified propeller point cloud data, and analyze the similarity between the two.

[0153] In addition, this scheme can also use the ICP (Iterative Closest Point) algorithm. The basic idea of ​​ICP is:

[0154] Given two point cloud sets:

[0155] X=left{x_{1},x_{2},cdots,x_{N_{x}}right};

[0156] P=left{p_{1},p_{2},cdots,p_{N_{p}}right};

[0157] Where x_{i} and p_{i} represent point cloud coordinates, and N_{x} and N_{p} represent the number of point clouds.

[0158] Find the rotation matrix R and translation vector t to minimize the following result.

[0159] E(R,t)=frac{1}{N_{p}}sum_{i=1}^{N_{p}}left|x_{i}-R p_{i}-tright|^{2};

[0160] In practical engineering, it is impossible to know how the points of two point clouds are paired. The only way is to reduce the error step by step through iterative solution, and finally obtain the rotation matrix R and translation matrix t that minimize the error equation.

[0161] Algorithm flow:

[0162] 1) Find the corresponding point

[0163] Typically, odometry data from the encoder disk is used to obtain the pose difference, which represents the robot's current pose in the previous robot coordinate system. This R and t are used as the first guess in the ICP algorithm to help the algorithm find the corresponding points in the point cloud.

[0164] It should be noted here that if the laser sensor is not installed at the center of the robot's coordinate system, there is a coordinate transformation relationship between the pose obtained from the odometry and the pose of the laser sensor.

[0165] 2) Calculate R and t based on the corresponding points.

[0166] This step involves constructing the error equation based on the identified corresponding points. Ordinary ICP uses the point-to-point distance as the error. The error equation is as follows:

[0167] E(R,t)=frac{1}{N_{p}}sum_{i=1}^{N_{p}}left|x_{i}-R p_{i}-tright|^{2};

[0168] Then R and t are solved. The specific derivation is as follows:

[0169] begin{aligned}E(R,t)&=frac{1}{N_{p}}sum_{i=1}^{N_{p}}left|x_{i}-Rp_{i}-tright|^{2}\E(R,t)&=frac{1}{N_{p}}sum_{i=1}^{N_{p}}le ft | ht)+left(u_{x}-Ru_{p}-tright)right|^{2}\=&frac{1}{N_{p}}sum_{i=1}^{N_{p}}left|x_{i}-u_{x}-Rleft(p_{i}-u_{p}right)right|^{2}+ left|u_{x}-Ru_{p}-tright|^{2}\&+2left(x_{i}-u_{x}-Rleft(p_{i}-u_{p}right)right)^{T}left(u_{x}-Ru_{p}-tright)\&=frac{1}{N_{p}}

[0170] sum_{i=1}^{N_{p}}left|x_{i}-u_{x}-Rleft(p_{i}-u_{p}right)right|^{2}+left|u_{x}-Ru_{p}-tright|^{2}end{aligned}

[0171] in,

[0172] u_{x}=frac{1}{N_{p}}sum_{i=1}^{N_{p}}x_{i}quad u_{p}=frac{1}{N_{p}}sum_{i=1}^{N_{p}}p_{i};

[0173] The sum of the terms left(x_{i}-u_{x}-Rleft(p_{i}-u_{p}right)right) will equal 0 after being accumulated N_{p} times.

[0174] In the final equation, only the term left|x_{i}-u_{x}-Rleft(p_{i}-u_{p}right)right|^{2} is related to R.

[0175] We can first set left|u_{x}-Ru_{p}-tright|^{2} equal to 0. Then we only need to minimize the term left|x_{i}-u_{x}-Rleft(p_{i}-u_{p}right)right|^{2}. After finding R, we substitute it into left|u_{x}-Ru_{p}-tright|^{2}=0 to solve for t. Therefore, the error equation can be simplified to the following:

[0176] min E(R,t)=frac{1}{N_{p}}sum_{i=1}^{N_{p}}left|x_{i}-u_{x}-Rleft(p_{i}-u_{p}right)right|^{2}=frac{1}{N_{p}} sum_{i=1}^{N_{p}}left|x_{i}^{prime}-Rp_{i}^{prime}right|^{2}=frac{1}{N_{p}} sum_{i=1}^{N_{p}} p_{i}^{prime}-2x_{i}^{T}Rp_{i}^{prime}=sum_{i=1}^{N_{p}}-2x_{i}^{prime T}R p_{i}^{prime};

[0177] Since the term x_{i}^{T}x_{i}^{prime} is unrelated to matrix R, it can be ignored. Because R is an orthogonal matrix, R^TR = I, and therefore p_{i}^{T}R^{T}R p_{i}^{prime} = p_{i}^{T}p_{i}^{prime}. This term is also unrelated to R, so it is also ignored. Finally, only one term remains that is related to R.

[0178] To minimize, we need to maximize, the following expression:

[0179] max sum_{i=1}^{N_{p}}x_{i}^{prime T}{{R}p_{i}^{prime}}=sum_{i=1}^{N_{p}}operatorname{Trace}left(R x_{i}^{prime}p_{i}^{T}right)=operatorname{Trace}(RH);

[0180] Where H = sum_{i=1}^{N_{p}}x_{i}^{prime}p_{i}^{prime T};

[0181] Using the theorem, assuming that matrix A is a positive definite symmetric matrix, then for any orthogonal matrix B, we have operatorname{Trace}(A)geq operatorname{Trace}(BA);

[0182] Perform SVD decomposition on H: H = U Lambda V^{T};

[0183] Let X = VU^{T}, then X is an orthogonal matrix.

[0184] XH=VU^{T}U Lambda V^{T}=V Lambda V^{T} yields a positive definite symmetric matrix.

[0185] but:

[0186] operatorname{Tr}operatorname{ace}(XH)geq operatorname{Tr}operatorname{ace}(BXH);

[0187] B is any orthogonal matrix, and X is also an orthogonal matrix; therefore, BX can take all orthogonal matrices. Thus, BX also includes the rotation matrix T that needs to be solved for.

[0188] operatorname{Trace}(RH)leq operatorname{Trace}(XH);

[0189] The equation holds true when R = X. Therefore:

[0190] R = X = VU^{T}

[0191] Then mathrm{t}=u_{x}-Ru_{p}

[0192] 3) Use the calculated R and t for the next iteration calculation until the error value is less than the set threshold.

[0193] This points out a significant flaw in ICP:

[0194] Points in two frames of laser point cloud data cannot represent the same location in space. Therefore, using the point-to-point distance as the error equation will inevitably introduce random errors.

[0195] Point cloud comparison algorithms are used to compare two frames of point cloud data to obtain the pose difference of the sensor (LiDAR or camera) before and after the comparison, i.e., odometer data. Comparison algorithms have evolved from the initial ICP method into several improved algorithms. These improvements focus on aspects such as registration point finding and error equations.

[0196] The advantage of this scheme is that the blades can be rotated according to the angle between them, and the distance information between points with corresponding point cloud information can be analyzed. This can detect whether the distribution of each blade of the propeller is reasonable and increase the accuracy of the detection results.

[0197] Example 5

[0198] Figure 6 This is a schematic diagram of the structure of the detection device for manufacturing anomalies in a ship propulsion device provided in Embodiment 5 of this application. This solution makes a further improvement to the above embodiment, specifically: the point cloud data comparison module is further used to: determine the comparison cell for the comparison point cloud data on each blade; according to the included angle and the location information of the comparison cell, spatially transform the points in the comparison cell to obtain the distance information between each point, and determine the comparison result.

[0199] like Figure 6 As shown, it specifically includes the following:

[0200] Point cloud data acquisition module 610 is used to acquire point cloud data of propeller blades;

[0201] The blade angle recognition module 620 is used to identify the angle between multiple blades of the propeller based on the point cloud data.

[0202] The point cloud data comparison module 630 is used to determine the comparison point cloud data from each blade according to the included angle, and to compare them to obtain the comparison result.

[0203] The consistency determination module 640 is used to determine whether the blade parameters are consistent based on the comparison results between the comparison point cloud data.

[0204] The report information generation module 650 is used to generate a report information of consistency anomaly when the blade parameters are inconsistent.

[0205] Among them, the point cloud data comparison module 630 is specifically used to determine the comparison point cloud data from each blade according to the included angle;

[0206] The comparison point cloud data is labeled according to the blade to which it belongs;

[0207] Rotate the comparison point cloud data according to the included angle to obtain point cloud information with corresponding relationships;

[0208] The comparison results are determined based on the distance information between points with corresponding point cloud information.

[0209] Among them, the cloud data comparison module 630 is also used to determine the comparison cell of the cloud data on each blade.

[0210] Based on the included angle and the location information of the comparison cell, the points in the comparison cell are spatially transformed to obtain the distance information between each point, and the comparison result is determined.

[0211] The comparison cells for point cloud data can be used to divide the point cloud data into regions, with each cell being the smallest unit of the point cloud data. The size of the cells can be manually defined, or the appropriate size of the cells can be determined by analysis by a smart terminal.

[0212] The location information of the comparison cells can be their spatial coordinates. Spatial transformation can involve changing the coordinate system of points within the rotated comparison cells to the spatial coordinate system of points within the original comparison cells, to facilitate subsequent distance calculations.

[0213] One way to obtain the distance information between points is to rotate the point cloud data in each blade according to the blade angle, and then perform a spatial transformation on the points in the comparison cell after rotation, transforming them to the spatial coordinate system of the points in the comparison cell before rotation. Record the corresponding spatial coordinates after spatial transformation, and calculate the distance information with the point cloud data with the same label and position information before rotation.

[0214] The advantage of this approach is that it allows us to determine the positional information of the point cloud data before and after rotation by using cells. Placing the points before and after rotation in the same space makes it easier to quickly find the distance information of point cloud data with corresponding relationships.

[0215] Example 6

[0216] Figure 7 This is a schematic diagram of the detection device for manufacturing anomalies in ship propulsion devices provided in Embodiment Six of this application. This solution makes a further improvement on the above embodiment, specifically: the device further includes: a hub point cloud data identification module, used to identify hub point cloud data in the point cloud data; a hub comparison result acquisition module, used to compare the hub point cloud data with pre-determined hub model data to obtain a hub comparison result; and a hub manufacturing standard judgment module, used to determine whether the propeller hub meets manufacturing standards based on the hub comparison result.

[0217] like Figure 7 As shown, it specifically includes the following:

[0218] The point cloud data acquisition module 710 is used to acquire point cloud data of the propeller blades;

[0219] The hub point cloud data recognition module 720 is used to recognize the hub point cloud data in the point cloud data.

[0220] The rotor hub comparison result acquisition module 730 is used to compare the rotor hub point cloud data with the predetermined rotor hub model data to obtain the rotor hub comparison result.

[0221] The propeller hub manufacturing standard judgment module 740 is used to determine whether the propeller hub meets the manufacturing standards based on the propeller hub comparison results.

[0222] The report information generation module 750 is used to generate a consistency anomaly report information when the propeller hub does not meet the manufacturing standards.

[0223] The propeller hub point cloud data recognition module 720 can be a program designed for a smart terminal to recognize propeller hub point cloud data. The propeller hub point cloud data can be point cloud data information of the propeller hub part.

[0224] The method for identifying the rotor hub point cloud data in the point cloud data can be to determine the extraction boundary of the rotor hub based on the length, width and other dimensional parameters of the rotor hub, and then obtain the point cloud data within the extraction boundary.

[0225] The propeller hub comparison result acquisition module 730 can be a program designed by a smart terminal to acquire comparison results of propeller hub point cloud data. The pre-determined propeller hub model data can be obtained by staff scanning qualified propellers in advance using 3D laser scanning technology, and then analyzing and extracting the point cloud data of their propeller hubs. This data is used to provide a comparison standard for the propeller hubs being inspected.

[0226] One way to obtain the propeller hub comparison results is to compare the point cloud data of the propeller hub to be detected with the pre-determined propeller hub model data, and analyze and determine the similarity between the two point cloud data.

[0227] The propeller hub manufacturing standard judgment module 740 can be a program designed by a smart terminal to determine whether the propeller hub meets manufacturing standards. The method for determining whether the propeller hub meets manufacturing standards can be to analyze whether the similarity between the point cloud data of the tested propeller hub and the pre-determined propeller hub model data reaches a specified value. For example, if the similarity between the point cloud data of the tested propeller hub and the pre-determined propeller hub model data reaches 98%, then the propeller hub is determined to meet the manufacturing standards. If the similarity is 96%, then the propeller hub does not meet the manufacturing standards.

[0228] The advantage of this approach is that it allows the propeller hub to be included in the testing scope, resulting in more comprehensive test results.

[0229] Example 7

[0230] Figure 8 This is a schematic diagram of the structure of the detection device for manufacturing anomalies in a ship propulsion device provided in Embodiment 7 of this application. This solution makes a further improvement to the above embodiment, specifically: the rotor hub comparison result acquisition module is further configured to: perform spatial transformation on the rotor hub model data to obtain a spatial pose matching the rotor hub point cloud data; identify the planar data and surface data of the rotor hub model in the current spatial pose; and compare the planar data and surface data with the planar point cloud data and surface point cloud data of the rotor hub point cloud data to obtain a rotor hub comparison result.

[0231] like Figure 8 As shown, it specifically includes the following:

[0232] The point cloud data acquisition module 810 is used to acquire point cloud data of the propeller blades;

[0233] The hub point cloud data recognition module 820 is used to recognize the hub point cloud data in the point cloud data.

[0234] The propeller hub comparison result acquisition module 830 is used to compare the propeller hub point cloud data with the pre-determined propeller hub model data to obtain the propeller hub comparison result.

[0235] The propeller hub manufacturing standard judgment module 840 is used to determine whether the propeller hub meets the manufacturing standards based on the propeller hub comparison results.

[0236] The report information generation module 850 is used to generate a consistency anomaly report information when the propeller hub does not meet the manufacturing standards.

[0237] The rotor hub comparison result acquisition module 830 is also used to perform spatial transformation on the rotor hub model data to obtain a spatial pose that matches the rotor hub point cloud data.

[0238] Identify the planar and surface data of the propeller hub model in the current spatial pose;

[0239] The planar point cloud data and the curved point cloud data of the rotor hub are compared with the planar point cloud data and the curved point cloud data of the rotor hub to obtain the rotor hub comparison result.

[0240] Spatial orientation can refer to the spatial position and attitude of the propeller hub.

[0241] One method to obtain spatial pose is to perform a spatial transformation on the propeller hub model data, converting it to the space where the point cloud data of the propeller hub under test resides, and then determining the position and attitude information of the propeller hub model data in that space. The position and attitude information of the propeller hub model data in that space can include information such as the point cloud data and shape contour of the propeller hub model.

[0242] Planar data can be point cloud data of a qualified propeller hub without any uneven surfaces. A curved surface can be a collection of continuous trajectories of a straight line in space. Curved surface data can be point cloud data of curved surfaces within a qualified propeller hub.

[0243] The method for identifying planar data and curved surface data can be that the intelligent terminal determines the difference based on the spatial location parameters of the planar and curved surface data. Specifically, if two points of a straight line lie in a plane, then all points on that line within that plane are identified as planar data, and the data information or point cloud data contained within that plane is considered planar data. Conversely, a surface with elevation changes and inflections can be identified as a curved surface, and the data information or point cloud data contained within that surface is considered curved surface data. Using methods similar to these, planar data and curved surface data can be distinguished.

[0244] The method for obtaining the propeller hub comparison results can be to acquire pre-determined propeller hub planar data and the similarity between the planar point cloud data and the surface point cloud data of the measured propeller hub. Specifically, the comparison parameters for the planar data can include information such as the size and spatial position of the propeller hub plane; the comparison for the surface data can include information such as the size, spatial position, and curvature of the surface.

[0245] The advantage of this approach is that by comparing planar and curved surface data separately, different comparison data can be determined for different surfaces, making the comparison results more accurate.

[0246] Example 8

[0247] Figure 9 This is a schematic diagram of the structure of the detection device for manufacturing abnormalities in a ship propulsion device provided in Embodiment 8 of this application. Figure 9 As shown, the specific steps include the following:

[0248] S901, Obtain point cloud data of propeller blades;

[0249] S902, Based on the point cloud data, identify the included angle between the multiple blades of the propeller;

[0250] S903, determine the comparison point cloud data from each blade according to the included angle, and compare them to obtain the comparison result;

[0251] S904: Based on the comparison results between the point cloud data, determine whether the blade parameters are consistent; if not, proceed to S905; if yes, end the process.

[0252] S905 generates a report of consistency anomalies.

[0253] In this embodiment, point cloud data of the propeller blades is acquired; based on the point cloud data, the included angles between multiple propeller blades are identified; comparative point cloud data is determined from each blade based on the included angles, and a comparison result is obtained; based on the comparison result between the comparative point cloud data, it is determined whether the blade parameters are consistent; if not, a report of consistency anomaly is generated. This method for detecting manufacturing anomalies in marine propulsion devices can improve the efficiency and accuracy of marine propeller testing and reduce the requirements for personnel.

[0254] The method for detecting manufacturing abnormalities in ship propulsion devices provided in this application embodiment has the same functional modules and beneficial effects as the device for detecting manufacturing abnormalities in mining ship propulsion devices provided in the above embodiment. To avoid repetition, it will not be described again here.

[0255] Example 9

[0256] like Figure 10As shown, Embodiment Nine of this application also provides an electronic device 1000, including a processor 1001, a memory 1002, and a program or instructions stored in the memory 1002 and executable on the processor 1001. When the program or instructions are executed by the processor 1001, they implement the various processes of the above-mentioned detection device embodiment for manufacturing abnormalities of ship propulsion device and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0257] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0258] Example 10

[0259] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described ship propulsion device manufacturing anomaly detection device embodiment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0260] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0261] Example 11

[0262] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described embodiment of the detection device for manufacturing anomalies of ship propulsion devices, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0263] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0264] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0265] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0266] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0267] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the claims.

Claims

1. A detection device for manufacturing abnormalities in ship propulsion systems, characterized in that, The detection device includes: The point cloud data acquisition module is used to acquire point cloud data of the propeller blades; The blade angle recognition module is used to identify the angle between multiple blades of the propeller based on the point cloud data. The point cloud data comparison module is used to determine the comparison point cloud data from each blade according to the included angle, and to compare them to obtain the comparison result; The consistency determination module is used to determine whether the blade parameters are consistent based on the comparison results between the comparison point cloud data. The report information generation module is used to generate a report information of consistency anomaly when the blade parameters are inconsistent; The point cloud data comparison module is specifically used for: determining comparison point cloud data from each blade according to the included angle, labeling the comparison point cloud data according to the blade to which it belongs, rotating the comparison point cloud data according to the included angle to obtain point cloud information with corresponding relationships, and determining the comparison result based on the distance information between the points with corresponding relationship point cloud information.

2. The detection device for manufacturing abnormalities in ship propulsion devices according to claim 1, characterized in that, The device further includes: The blade parameter identification module is used to identify the blade curvature and blade length of multiple blades of the propeller based on the point cloud data. The extraction boundary determination module is used to determine the extraction boundary of the comparative point cloud data from each blade based on the blade curvature and the blade length.

3. The detection device for manufacturing abnormalities in ship propulsion devices according to claim 2, characterized in that, The extraction boundary determination module is further used for: The width boundary of the comparison point cloud data is determined based on the blade curvature. as well as, The length boundary of the comparison point cloud data is determined based on the blade length.

4. The detection device for manufacturing abnormalities in ship propulsion devices according to claim 1, characterized in that, The point cloud data comparison module is also used for: Determine the comparison cells for the comparison point cloud data on each blade; Based on the included angle and the location information of the comparison cell, the points in the comparison cell are spatially transformed to obtain the distance information between each point, and the comparison result is determined.

5. The detection device for manufacturing abnormalities in ship propulsion devices according to claim 1, characterized in that, The device further includes: A hub point cloud data recognition module is used to recognize hub point cloud data in the point cloud data; The rotor hub comparison result acquisition module is used to compare the rotor hub point cloud data with the pre-determined rotor hub model data to obtain the rotor hub comparison result. The propeller hub manufacturing standard judgment module is used to determine whether the propeller hub meets the manufacturing standards based on the comparison results of the propeller hubs.

6. The detection device for manufacturing abnormalities in ship propulsion devices according to claim 5, characterized in that, The propeller hub comparison result acquisition module is also used for: The rotor hub model data is spatially transformed to obtain a spatial pose that matches the rotor hub point cloud data; Identify the planar and surface data of the propeller hub model in the current spatial pose; The planar point cloud data and the curved point cloud data of the rotor hub are compared with the planar point cloud data and the curved point cloud data of the rotor hub to obtain the rotor hub comparison result.

7. A method for detecting manufacturing abnormalities in a ship propulsion system, characterized in that, The method includes: Obtain point cloud data of the propeller blades; Based on the point cloud data, the included angle between the multiple blades of the propeller is identified; The comparison point cloud data is determined from each blade according to the included angle, and the comparison results are obtained by comparison. The comparison includes: determining the comparison point cloud data from each blade according to the included angle; labeling the comparison point cloud data according to the blade to which it belongs; rotating the comparison point cloud data according to the included angle to obtain point cloud information with corresponding relationships; and determining the comparison results based on the distance information between the points with corresponding relationship point cloud information. Based on the comparison results between the point cloud data, it is determined whether the blade parameters are consistent. If not, a consistency anomaly report will be generated.

8. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method for detecting manufacturing anomalies in a ship propulsion device as described in claim 7.

9. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method for detecting manufacturing anomalies in a ship propulsion device as described in claim 7.

Images