Defect prediction method and system for tail jet adjusting sheet investment casting of aero-engine

Abstract

The invention discloses a defect prediction method and system for an aircraft engine tail jet adjusting sheet investment casting, and belongs to the technical field related to casting and machine learning, and the method comprises the steps: carrying out the numerical simulation of an adjusting sheet casting and a pouring system through finite element simulation software, and obtaining the position and size distribution result of a casting defect, the result is converted into three-dimensional point cloud data; dividing the three-dimensional point cloud data into a training set and a verification set, and training a PointNet-based defect prediction model by using the training set; the prediction accuracy and generalization ability of the defect prediction model are evaluated through the verification set, and hyper-parameters are adjusted according to a verification result; and converting a to-be-predicted adjusting sheet casting pouring system three-dimensional model into three-dimensional point cloud data, inputting the trained defect prediction model, and outputting a casting defect position and a prediction result. According to the method, the defects that an existing finite element simulation method is long in time consumption and wastes computing resources are overcome.

Description

Technical Field

[0001] The present invention relates to the technical field related to casting and machine learning, and in particular to a method and system for predicting defects of investment castings of tail jet regulating pieces of aircraft engines. Background Art

[0002] Aircraft engine tail jet trim is a key component at the tail of an aircraft engine, mainly used to adjust the direction and shape of the engine's exhaust jet, thereby affecting the size and distribution of thrust. They are often used to improve engine performance, optimize fuel efficiency, control exhaust noise, and improve aircraft maneuverability. Trim is widely used in jet engines, especially in fighters or high-performance aircraft, which have complex and changeable flight conditions. Tail jet trim can provide flexible thrust adjustment and aerodynamic control.

[0003] The production of adjustment plates generally adopts the method of investment casting, which is a precision casting process with advantages such as high precision, complex shape and smooth surface. However, during the production process, various process parameters and working conditions often lead to defects in the castings, which affects the performance of the castings and even directly leads to the scrapping of the castings. However, the production cost of investment castings is extremely high. If the castings are scrapped, it will cause great economic losses to the company. Therefore, metallurgical quality control of investment castings has become the key in the production process.

[0004] In the early days, the quality prediction of investment castings was completely judged by professionals with many years of relevant experience, and then actual production was carried out. According to the actual defect distribution, professionals would analyze and improve it, and the process would be repeated until the product was qualified and put into formal production. This process is not only time-consuming and labor-intensive, but also requires extremely high professionalism from relevant practitioners. Later, due to the improvement of finite element simulation technology, relevant researchers developed professional casting finite element simulation software, such as ProCAST, Huazhu CAE, anycasting, etc., which greatly reduced the R&D cost of products. However, this process is still relatively time-consuming, the calculation cost is still high, and if you want to achieve higher accuracy, you need to use denser finite element mesh division and smaller time step for simulation calculation. This method is time-consuming and requires more computing resources. Therefore, the foundry industry is in urgent need of a fast and low-cost casting defect prediction method.

[0005] In recent years, with the continuous development of artificial intelligence technology, probabilistic algorithms based on large amounts of data have gained the favor of a large number of scholars, and various machine learning algorithms have been proven to be superior in many scenarios. Therefore, the application of machine learning to casting defect prediction has received more and more attention. For example, some researchers use U-Net networks to predict the temperature field of the casting process, and thus infer the defects of the castings based on the distribution of the temperature field; some use neural networks to build a relationship model between casting process parameters and casting defects, predict casting defects, and thus adjust casting process parameters; use fully connected convolutional neural networks to transform casting defect prediction tasks into semantic segmentation tasks of three-dimensional images, distinguishing defects from non-defective parts in the casting system from pixel points, etc. The above methods have confirmed the feasibility of machine learning methods in casting defect prediction tasks, but the above methods do not actually point out the location of the defects, so it is difficult to adjust casting process parameters from the entire casting system to reduce casting defects and improve the metallurgical quality of castings.

[0006] There are few studies on the use of machine learning methods to predict casting defects. Most of them are based on building a neural network between casting process parameters and defect size and quantity. The impact of the pouring system on casting defects is rarely considered. However, in the actual casting production process, the rationality of the pouring system setting greatly affects the molding quality of the casting. Summary of the invention

[0007] In order to solve the above technical problems, a method for predicting defects of investment castings of tail jet regulating plates of aircraft engines is proposed, including numerically simulating the regulating plate casting and the pouring system through finite element simulation software, obtaining the location and size distribution results of casting defects, and converting the results into three-dimensional point cloud data; The three-dimensional point cloud data is divided into a training set and a validation set. The defect prediction model based on PointNet is trained using the training set, including: local feature extraction and global feature aggregation, and the local features and global features are fused; the local feature extraction is to extract the index points of the three-dimensional point cloud data using the farthest point sampling method, and obtain the local point cloud set through the spherical neighborhood query method, extract the local geometric features of each point through a multi-layer perceptron with shared weights, and use the set abstraction layer to capture multi-scale local features in a hierarchical manner; Evaluate the prediction accuracy and generalization ability of the defect prediction model through the validation set, and adjust the hyperparameters based on the validation results; The three-dimensional model of the casting system of the regulating piece to be predicted is converted into three-dimensional point cloud data, and the trained defect prediction model is input to output the location of the casting defect and the prediction result.

[0008] As a preferred solution of the method for predicting defects of investment castings of tail jet regulating pieces of an aircraft engine described in the present invention, the three-dimensional point cloud data includes: three-dimensional regulating piece pouring system point cloud data and defect point cloud data; The three-dimensional regulating plate pouring system point cloud data is a complete casting and pouring system structure, providing the overall geometric information of the casting and pouring system; The defect point cloud data simulates and predicts the defect area in order to complete the defect prediction task; The converting the result into three-dimensional point cloud data includes converting the position and size distribution results of casting defects simulated by finite element simulation into three-dimensional point cloud using a uniform sampling method.

[0009] As a preferred scheme of the defect prediction method of the aircraft engine tail nozzle adjustment plate investment casting described in the present invention, the use of the farthest point sampling method to extract the index points of the three-dimensional point cloud data includes randomly determining a point in the three-dimensional point cloud data as the initial point, using the farthest point sampling method to obtain the point with the farthest square Euclidean distance from the initial point, and obtaining the index point set and spatial structure characteristics.

[0010] As a preferred solution of the method for predicting defects of investment castings of tail jet regulating pieces of an aircraft engine described in the present invention, wherein: the obtaining of a local point cloud set by a spherical neighborhood query method comprises: using a spherical neighborhood query method with a specified radius to select neighborhood points of the sampling point from three-dimensional point cloud data, and obtaining a local point cloud set centered on each index point; The local geometric features of each point are extracted through a multi-layer perceptron with shared weights, and the multi-scale local features are captured hierarchically using the set abstraction layer. The three-dimensional coordinates, color information, and normal vector information of each point in the point set are independently processed through the multi-layer perceptron layer with shared weights; the input dimension of each point is 9, including the three-dimensional coordinates , color information , normal vector information ; It is mapped to 64 dimensions and 128 dimensions through a shared multi-layer perceptron layer to capture the local features of each point.

[0011] As a preferred solution of the method for predicting defects in investment castings of aircraft engine tail nozzle adjustment plates described in the present invention, global feature aggregation includes, through stacking of set abstraction layers, aggregating the local features of each point with the spatial structure features using a global pooling method to form a 1024-dimensional global feature required for casting defect prediction.

[0012] As a preferred solution of the defect prediction method for the investment casting of the aircraft engine tail nozzle adjustment plate described in the present invention, the local features and the global features are fused, including using a multi-layer perceptron to perform feature splicing on the 1024-dimensional global features and the 128-dimensional local features extracted from each set abstraction layer to form a 1088-dimensional feature vector to obtain a fused feature.

[0013] As a preferred scheme of the defect prediction method for the investment casting of the aircraft engine tail nozzle adjustment plate described in the present invention, the fusion of local features and global features also includes inputting the fused features into a shared multi-layer perception machine layer, generating prediction results point by point, and using the prediction results of each point to calculate the defect prediction model loss and update the model parameters.

[0014] Another object of the present invention is to provide a defect prediction system for investment castings of tail jet regulating plates of aircraft engines, a numerical simulation module, which uses finite element simulation software to perform numerical simulation on regulating plate castings and pouring systems, obtain the location and size distribution results of casting defects, and convert the results into three-dimensional point cloud data; A model training module, which divides the three-dimensional point cloud data into a training set and a validation set, and uses the training set to train a defect prediction model based on PointNet; Model verification module, which evaluates the prediction accuracy and generalization ability of the defect prediction model through the verification set and adjusts the hyperparameters according to the verification results; The execution module converts the three-dimensional model of the casting system of the adjustment piece to be predicted into three-dimensional point cloud data, inputs the trained defect prediction model, and outputs the location of the casting defect and the prediction result.

[0015] The present invention solves the shortcomings of the existing finite element simulation method, which is time-consuming and wastes computing resources.

[0016] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the method for predicting defects of an aircraft engine tail nozzle adjustment plate investment casting are implemented.

[0017] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the method for predicting defects of an aircraft engine tail nozzle adjustment plate investment casting are implemented.

[0018] The beneficial effects of the present invention are as follows: the position and size distribution of casting defects are obtained by finite element simulation software and converted into three-dimensional point cloud data, which not only retains the spatial information of the defects but also reduces the detection cost. The point cloud data is divided into a training set and a validation set, so that the model can fully learn the casting and defect features, avoid overfitting, and improve the generalization ability. In the feature extraction stage, the farthest point sampling and spherical neighborhood query are used to evenly select index points, and the local geometric features are extracted by a multi-layer perceptron with shared weights to ensure that the model can accurately capture the subtle defects of the casting. The multi-scale local features are captured hierarchically by the set abstraction layer, and the overall structural information is extracted in combination with the global pooling operation to achieve the fusion of local and global features, thereby improving the stability and accuracy of defect prediction. In the classification process, the fused features are further extracted by the convolution layer and the fully connected layer, and normalized by the Softmax function to achieve accurate classification of each point. At the same time, the weighted cross entropy loss function is combined to optimize the model to deal with the problem of class imbalance. The model performance is evaluated using the validation set and the hyperparameters are adjusted to ensure that it can maintain high-precision prediction under different casting structures. Ultimately, the trained model can convert the new casting pouring system into point cloud data and quickly output defect location and category prediction results, significantly improving quality inspection efficiency, reducing production costs, and providing strong support for casting process optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 An overall flow chart of a method for predicting defects in investment castings of aircraft engine tail nozzle adjustment pieces provided in one embodiment of the present invention.

[0021] Figure 2 A defect prediction flow chart of an aircraft engine tail jet regulating piece investment casting defect prediction method provided in accordance with an embodiment of the present invention.

[0022] Figure 3 A finite element numerical simulation process diagram of a method for predicting defects in an aircraft engine tail nozzle adjustment piece investment casting provided by one embodiment of the present invention.

[0023] Figure 4 A defect prediction result diagram of a finite element simulation of a defect prediction method for an aircraft engine tail nozzle adjustment piece investment casting provided by an embodiment of the present invention.

[0024] Figure 5The finite element simulation results of the defect prediction method of the aircraft engine tail nozzle adjustment plate investment casting provided by one embodiment of the present invention are more likely to produce SEM scanning electron microscope result images of defective parts.

[0025] Figure 6 A flow chart of acquiring point cloud data required for defect prediction model training of an aircraft engine tail nozzle adjustment piece investment casting defect prediction method provided by one embodiment of the present invention.

[0026] Figure 7 An overall structural diagram of a casting defect prediction model based on PointNet for a method for predicting defects in investment castings of an aircraft engine tail nozzle adjustment piece provided in one embodiment of the present invention.

[0027] Figure 8 A diagram of the training results of a casting defect prediction model of an aircraft engine tail nozzle adjustment piece investment casting defect prediction method provided by one embodiment of the present invention.

[0028] Fig. 9 A comparison diagram of finite element simulation results of a method for predicting defects in investment castings of aircraft engine tail nozzle adjustment pieces provided in one embodiment of the present invention and casting defect prediction results predicted by a model. DETAILED DESCRIPTION

[0029] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0030] Example 1, reference Figure 1-Figure 9 , which is the first embodiment of the present invention, provides a method for predicting defects of an aircraft engine tail nozzle regulating piece investment casting, comprising: S1: Numerical simulation of the regulating plate casting and the pouring system is performed using finite element simulation software to obtain the location and size distribution results of casting defects, of which a portion (80% of the total finite element simulation) is used as a training set and the remaining 20% ​​is used as a validation set, and the results are converted into three-dimensional point cloud data.

[0031] Finite element simulation software is used to perform numerical simulation on the regulating plate casting and the pouring system to obtain the location and size distribution results of the casting defects. The principle of numerical simulation of the casting process is to geometrically discretize the pouring system and the casting mold of the casting, and with the support of the physical-mathematical model, the changes in the temperature field, flow field and stress field of the casting process are analyzed to predict the defects that may occur in the casting; the results are converted into three-dimensional point cloud data, and the casting defect results of the finite element simulation are exported to the three-dimensional CAD file format, and point cloud sampling is performed together with the three-dimensional CAD pouring system of the casting to obtain the data required for training this model.

[0032] S1.1. Analyze the structural characteristics of the regulating plate parts, combine the principles of fluid mechanics, thermodynamics and solidification shrinkage feeding, avoid turbulence or air entrainment of the molten metal during the pouring process, and ensure that the casting achieves sequential solidification and directional shrinkage feeding during the solidification process, so as to establish a suitable casting pouring system, and use finite element simulation software to simulate the pouring of the regulating plate casting. The finite element simulation process is as follows: Figure 3 As shown. Through finite element meshing of castings and pouring systems, pouring parameter settings, and finite element simulation calculations, in the finite element simulation process, in order to improve the simulation accuracy, it is inevitable to reduce the finite element mesh size. Reducing the mesh size means that a large amount of computing resources are consumed. Therefore, different mesh sizes are used to mesh the castings and pouring systems. Compared with castings, the pouring system is not our focus, so the mesh size of the casting is set to 1mm, and the pouring system is set to 5mm. When setting parameters, create a user's personal material library based on the material composition used in the actual production of castings, so that the finite element simulation is closer to the actual production process. The defect distribution of the adjustment plate casting is obtained, as shown Figure 4 As shown. By combining the finite element simulation results with the actual analysis of the casting structure, two locations in the adjustment plate casting where defects are more likely to occur in the finite element simulation results are sampled and used for high-order material characterization to verify the accuracy of the finite element simulation method in predicting casting defects. The results show that there are indeed dense hole defects in the locations where defects are more likely to occur in the finite element simulation results under SEM electron microscope scanning. Therefore, the finite element numerical simulation method has a higher casting defect prediction accuracy. The scanning results are shown in Figure 5 shown.

[0033] S1.2. Save the casting 3D model and finite element prediction defect results in S1.1 as a 3D CAD model, and perform point cloud sampling on the 3D CAD model. When sampling the 3D CAD model, it is necessary to ensure uniform sampling so that the point cloud can contain the location information of each part of the pouring system and most of the structural features. The pouring system point cloud is used as a data sample, and the pouring system and defect distribution results are merged into data labels after point clouding. In order to meet the processing requirements of the model, the point cloud needs to be downsampled so that its point number is 100,000, and each point contains 9 attributes, namely the 3D coordinates , color information , normal vector information The process of obtaining point cloud data required for model training is as follows: Figure 6 shown.

[0034] Furthermore, the adjustment plate casting and pouring system are numerically simulated by finite element simulation software to obtain the location and size distribution results of casting defects, and the simulation results are converted into three-dimensional point cloud data. The core purpose of this process is to accurately predict the defects that may appear in the casting during the pouring process, such as shrinkage cavities and shrinkage porosity, through numerical simulation, so as to provide a scientific basis for subsequent defect detection and analysis. By combining the finite element simulation results with the actual casting structure, the accuracy of the simulation can be ensured, providing high-quality annotation data for subsequent machine learning model training, and further optimizing the accuracy of defect prediction.

[0035] Furthermore, by combining finite element simulation with 3D point cloud data sampling, this step effectively overcomes the reliance on actual castings in traditional detection methods, allowing model training to be completed in a virtual environment. The use of uniform sampling and downsampling techniques ensures the quality and balance of point cloud data, while giving point cloud data multiple attributes (such as 3D coordinates, color information, and normal vector information), enhancing the model's ability to learn different defect types. This method not only improves the accuracy and efficiency of defect prediction, but also reduces the cost of testing in actual production, helps to foresee potential quality problems before production, and provides data support for process adjustments during casting production.

[0036] S2: Divide the three-dimensional point cloud data into a training set and a validation set, and use the training set to train a defect prediction model based on PointNet.

[0037] All processed data are randomly divided into training set and validation set.

[0038] Based on the improvement of finite element simulation technology, relevant researchers have developed professional casting finite element simulation software, such as ProCAST, Huazhu CAE, anycasting, etc., which greatly reduces the R&D cost of products. However, this process is still relatively time-consuming, the calculation cost is still high, and if you want to achieve higher accuracy, you need to use denser finite element meshing and smaller time steps for simulation calculations. This method is time-consuming and requires more computing resources. Therefore, the foundry industry is in urgent need of a fast and low-cost casting defect prediction method. Finite element simulation software is used to numerically simulate the adjustment plate casting and the pouring system to obtain the location and size distribution results of the casting defects, and the simulation results are converted into three-dimensional point cloud data. The core purpose of this process is to accurately predict the defects that may occur in the casting during the pouring process, such as shrinkage cavities and shrinkage porosity, through numerical simulation, so as to provide a scientific basis for subsequent defect detection and analysis. By combining the finite element simulation results with the actual casting structure, the accuracy of the simulation can be ensured, and high-quality annotation data can be provided for subsequent machine learning model training, further optimizing the accuracy of defect prediction.

[0039] Input the point cloud data in the training set into Figure 7 In the defect prediction model of the adjustment plate investment casting shown in the figure, the model uses a shared MLP to extract the local features of each point, and then uses the maximum pooling method to aggregate the local features to form the global features of the point cloud.

[0040] S2.1. Input the data in the training set into the SA feature extraction module, and the data of each point is nonlinearly mapped through a shared MLP. The purpose of this step is to extract high-dimensional features of each point and ensure that the operation is insensitive to the order of the points. The features are mapped to 64 dimensions to obtain an Nx64 local feature matrix, which is then mapped to 1024 dimensions to obtain the features of each point in the point cloud.

[0041] S2.2, the local features obtained in S2.1 are aggregated into global features of the point cloud by using the maximum pooling method. This step integrates the local features of all points into a global feature vector of a fixed size to complete the representation of the entire point cloud. Finally, a 1024-dimensional global feature vector is obtained.

[0042] The captured Nx64-dimensional local feature matrix is ​​concatenated with the 1024-dimensional global feature vector, and the global features are copied and connected to the local features of each point to obtain an Nx1088 feature matrix containing the local features of each point and the global features of the point cloud.

[0043] The fused feature matrix is ​​input into the shared MLP layer, and the prediction results are generated point by point through the MLP layer. The prediction results of each point are used to calculate the model loss and update the model parameters. First, the softmax operation is used to convert the original output of the model into a probability distribution. For the original output vector z of a sample, its softmax calculation formula is:

[0044] in, Indicates the corresponding category The original output value of Indicates the number of categories for classification; Yes Category The predicted probability of .

[0045] Then the true label of each sample Calculate the negative log-likelihood loss for the label, namely:

[0046] in, Representation sample The real label, is the predicted probability of the true label after softmax.

[0047] Finally, a weighting function is used to adjust the impact of category imbalance. The weighting function is:

[0048] in: Represents the number of points in each gating system point cloud; Represents the indicator function, which is 1 if the label of sample i is category j, otherwise it is 0.

[0049] Furthermore, by dividing the three-dimensional point cloud data into a training set and a validation set, and using the training set to train the defect prediction model based on PointNet, the model can effectively extract features from the point cloud data and perform defect prediction. Specifically, the point cloud data of the training set is extracted through a shared MLP (multi-layer perceptron) network for local features, and aggregated into global features through the maximum pooling method, so that each point can not only reflect the local geometric structure, but also combine the global information of the entire point cloud. In addition, through feature fusion, the local features and global features are spliced ​​and input into the MLP layer, which further improves the performance of the model in dealing with complex casting defects. This design enables the model to comprehensively consider the detailed features and overall structural features of the point cloud when predicting defects, thereby improving the accuracy and robustness of the prediction.

[0050] Furthermore, firstly, the combination of local feature extraction and global feature aggregation enables the model to take into account both the detailed information and the overall morphological features of the casting point cloud, thus enhancing the model's ability to identify different defects. Secondly, the use of maximum pooling to aggregate local features to form global features ensures that the information of all points in the point cloud can be effectively integrated while avoiding information loss. In addition, by using softmax operations and cross-entropy loss, combined with the class weighting function to adjust the impact of class imbalance, it is ensured that the model can still maintain good performance when the number of samples in different categories is uneven, especially when dealing with rarer defect types in castings. Finally, the use of the validation set ensures the generalization ability of the model, allowing it to maintain a high prediction accuracy on unseen data, enhancing the industrial application value of the model.

[0051] S3: Evaluate the prediction accuracy and generalization ability of the defect prediction model through the validation set, and adjust the hyperparameters based on the validation results.

[0052] The validation set is not used in the model training process. Therefore, the data in the validation set is data that the model has never seen before. It is used to evaluate the performance of the model and provide feedback to the model to adjust the model hyperparameters based on the validation results.

[0053] The validation set is not used in the model training process. Therefore, the data in the validation set is data that the model has never seen before. It is used to evaluate the performance of the model and provide feedback to the model, and adjust the model hyperparameters based on the validation results.

[0054] Repeatedly update the model parameters and verify the model's detection ability, and finally take the model parameters when the accuracy reaches the maximum value as the final training result of the model. Figure 8 This is the change in loss and accuracy during the model training process under the technical solution of the present invention.

[0055] The schematic diagram of defect prediction results of the adjustment plate investment casting obtained by the technical solution of the present invention is as follows Fig. 9 As shown, the left side is the finite element numerical simulation result, and the right side is the prediction result obtained by the present invention. The red part is the normal part, and the white part is the defective part.

[0056] It can be found that the prediction results of the present invention correspond to the traditional finite element numerical simulation results. Since the finite element simulation method has been proven to be reliable in predicting casting defects in step S1.1, and the defect distribution in the model prediction results is roughly the same as the finite element results, the present invention achieves rapid and accurate prediction of defects in investment castings of aircraft engine tail nozzle adjustment plates.

[0057] Furthermore, the accuracy and generalization ability of the defect prediction model are evaluated through the validation set, and the hyperparameters are adjusted according to the evaluation results to ensure that the model can show good prediction ability under different data distributions. By repeatedly updating the model parameters and verifying them, the overfitting problem can be avoided and the stability and reliability of the final model in practical applications can be ensured. In addition, the model parameters are selected when the accuracy reaches the maximum value to ensure the best performance of the model on the test set. This process enables the model to adapt to different casting structures and defect types through gradual adjustment, thereby improving the comprehensiveness and accuracy of defect prediction.

[0058] Furthermore, the accuracy and generalization ability of the defect prediction model are evaluated through the validation set, and the hyperparameters are adjusted according to the evaluation results to ensure that the model can show good prediction ability under different data distributions. By repeatedly updating the model parameters and verifying them, the overfitting problem can be avoided and the stability and reliability of the final model in practical applications can be ensured. In addition, the model parameters are selected when the accuracy reaches the maximum value to ensure the best performance of the model on the test set. This process enables the model to adapt to different casting structures and defect types through gradual adjustment, thereby improving the comprehensiveness and accuracy of defect prediction.

[0059] S4: Convert the three-dimensional model of the casting system of the adjustment piece to be predicted into three-dimensional point cloud data, input the trained defect prediction model, and output the location of the casting defect and the prediction result.

[0060] Furthermore, the three-dimensional model of the casting system of the adjustment piece to be predicted is converted into three-dimensional point cloud data. This process is to connect the actual engineering data (such as the three-dimensional CAD model) with the input format of the machine learning model, realizing a seamless transition from traditional engineering design to a data-driven prediction model. By converting the three-dimensional model into point cloud data, deep learning models (such as PointNet) can be used to efficiently process and analyze the geometric structure of the casting and its defect information. This step ensures that the structural characteristics and defect information of the casting can be fully utilized by the model, thereby improving the accuracy and reliability of defect prediction.

[0061] Furthermore, the point cloud data can preserve the three-dimensional spatial structure of the casting, making it easier for the model to deeply analyze the local and global features of each point. Secondly, by inputting the trained defect prediction model, the defect location and type in the casting can be automatically predicted, avoiding the time consumption and human errors of the traditional manual inspection process. Ultimately, this process improves the efficiency of casting defect detection, makes quality control in the production process more accurate and efficient, reduces production costs, and can detect and solve potential problems in advance, thereby improving overall productivity and casting quality.

[0062] Embodiment 2 is the second embodiment of the present invention, which provides a method for predicting defects in investment castings of aircraft engine tail nozzle regulating pieces. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0063] Finite element simulation software was used to numerically simulate the adjustment plate casting and the gating system. The main contents of the simulation included the geometric structure, temperature field, flow field and solidification process of the gating system. During the simulation, two different types of adjustment plate castings (casting A and casting B) were selected as experimental objects. The simulated results of the defect location and size distribution of the castings were exported, and the results were converted into three-dimensional point cloud data through a three-dimensional visualization tool to ensure that the data of each point contained three-dimensional coordinates, color information and normal vector information. To ensure the uniformity and representativeness of the data, the uniform sampling method was used to sample the point cloud, and finally 50,000 points were obtained.

[0064] In the data preparation stage, 100 different point cloud data of adjustment sheet castings were used, of which 80 castings were used as training sets and 20 castings were used as validation sets, and validation was performed after each training round. Specifically, 80% of the casting data (i.e., 80 castings) were used for training, and 20% of the castings (i.e., 20 castings) were used as validation sets. The data in the training set were extracted by the farthest point sampling (FPS) method for index point extraction, the spherical neighborhood query method was used to select neighborhood points, and the multi-layer perceptron (MLP) with shared weights was used to extract the local geometric features of each point. The input features of each point include three-dimensional coordinates, color information, and normal vector information. After mapping processing, 64-dimensional and 128-dimensional local feature representations were obtained. After the multi-layer perceptron extracted the local features, the set abstraction layer was used to capture the multi-scale local features, and the local features of each point were aggregated into 1024-dimensional global features through the global pooling operation.

[0065] Next, local features and global features are concatenated and input into a shared multi-layer perceptron layer to generate prediction results point by point. In order to deal with class imbalance, a weighted cross entropy loss function is used for optimization. The hyperparameters of the model are adjusted through the validation set to ensure that the training results can achieve the best performance. After the training and validation are completed, in order to further test the generalization ability of the model, castings A and B (i.e., new test data) that have never entered training or validation are selected, and their corresponding three-dimensional point clouds are input into the defect prediction model that has been trained, and the location and prediction results of the casting defects are output.

[0066] During the experiment, the following experimental data were collected: Number of simulation points for casting A: 50,000; Number of simulation points for casting B: 50,000; Number of defect detection points (casting A): 2,430; Number of defect detection points (casting B): 2,510; Defect prediction accuracy (casting A): 92.35%; Defect prediction accuracy (casting B): 93.01%.

[0067] Training time: 12.35 hours; testing time: 1.52 hours.

[0068] Prediction accuracy (training set): 90.88% (casting A), 91.33% (casting B).

[0069] Prediction accuracy (validation set): 91.10% (casting A), 91.55% (casting B).

[0070] It can be seen from the experimental data analysis that the method of the present invention has shown significant advantages in casting defect prediction. First, the defect detection accuracy and defect prediction accuracy are both high, especially for casting B, the defect prediction accuracy reaches 93.01%, which is significantly higher than the accuracy of the traditional finite element simulation method, indicating that the method of the present invention can more accurately capture the location and size of the defect, avoiding the problem of missed detection of the traditional method. The defect prediction accuracy of casting A also reached 92.35%. This result shows that the method has strong universality on different types of castings and has good generalization ability.

[0071] Secondly, the average training time is 12.35 hours and the testing time is only 1.52 hours, which greatly shortens the computing time and resource consumption in the traditional finite element simulation process and improves production efficiency. Compared with the traditional defect prediction method based on finite element simulation, the traditional method often requires a longer simulation time and is limited by the model's computing power in terms of the accuracy of defect prediction. In contrast, the method of the present invention achieves fast and high-precision defect prediction through a deep learning model based on PointNet.

[0072] In terms of prediction accuracy, the model showed good training effect and strong generalization ability through the comparison of training set and validation set. The prediction accuracy of the training set was 90.88% (casting A) and 91.33% (casting B), and the prediction accuracy of the validation set was 91.10% (casting A) and 91.55% (casting B), which shows that the model can effectively learn the characteristics of defects from the training data and maintain high prediction performance on unseen data.

[0073] In summary, the present invention not only improves the accuracy of casting defect prediction in terms of precision, but also significantly improves the computational efficiency by combining finite element simulation and PointNet. Compared with traditional methods, the method of the present invention can significantly shorten the prediction time, reduce the consumption of computing resources, and provide a more efficient and accurate solution for casting defect prediction and process optimization. This result shows that the defect prediction method based on PointNet has important practical application value and can effectively improve the production efficiency and quality control level of investment castings of aircraft engine tail nozzle adjustment plates.

[0074] Embodiment 3 is the third embodiment of the present invention, which is different from the first two embodiments in that: If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0075] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0076] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0077] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0078] Embodiment 4 is the fourth embodiment of the present invention, which provides an aircraft engine tail nozzle adjustment piece investment casting defect prediction system, including: The numerical simulation module performs numerical simulation on the adjustment plate casting and the pouring system through finite element simulation software, obtains the location and size distribution results of casting defects, and converts the results into three-dimensional point cloud data.

[0079] The model training module divides the three-dimensional point cloud data into a training set and a validation set, and uses the training set to train a defect prediction model based on PointNet.

[0080] The model validation module evaluates the prediction accuracy and generalization ability of the defect prediction model through the validation set and adjusts the hyperparameters based on the validation results.

[0081] The execution module converts the three-dimensional model of the casting system of the adjustment piece to be predicted into three-dimensional point cloud data, inputs the trained defect prediction model, and outputs the location of the casting defect and the prediction result.

[0082] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for predicting defects in investment castings of tail jet regulating pieces of aircraft engines, characterized in that: Including, numerically simulating the adjustment plate casting and the pouring system through finite element simulation software, obtaining the location and size distribution results of casting defects, and converting the results into three-dimensional point cloud data; The three-dimensional point cloud data is divided into a training set and a validation set. The defect prediction model based on PointNet is trained using the training set, including: local feature extraction and global feature aggregation, and the local features and global features are fused; the local feature extraction is to extract the index points of the three-dimensional point cloud data using the farthest point sampling method, and obtain the local point cloud set through the spherical neighborhood query method, extract the local geometric features of each point through a multi-layer perceptron with shared weights, and use the set abstraction layer to capture multi-scale local features in a hierarchical manner; Evaluate the prediction accuracy and generalization ability of the defect prediction model through the validation set, and adjust the hyperparameters based on the validation results; The three-dimensional model of the casting system of the regulating piece to be predicted is converted into three-dimensional point cloud data, and the trained defect prediction model is input to output the location of the casting defect and the prediction result.

2. The method for predicting defects of an aircraft engine tail jet regulating plate investment casting according to claim 1, characterized in that: The three-dimensional point cloud data includes: three-dimensional adjustment piece pouring system point cloud data and defect point cloud data; The three-dimensional regulating plate pouring system point cloud data is a complete casting and pouring system structure, providing the overall geometric information of the casting and pouring system; The defect point cloud data simulates and predicts the defect area in order to complete the defect prediction task; The converting the result into three-dimensional point cloud data includes converting the position and size distribution results of casting defects simulated by finite element simulation into three-dimensional point cloud using a uniform sampling method.

3. A method for predicting defects in investment castings of an aircraft engine tail jet regulating plate according to claim 2, characterized in that: The method of extracting index points of three-dimensional point cloud data using the farthest point sampling method includes randomly determining a point in the three-dimensional point cloud data as an initial point, using the farthest point sampling method to obtain the point with the farthest square Euclidean distance from the initial point, and obtaining an index point set and spatial structure features.

4. A method for predicting defects in investment castings of an aircraft engine tail jet regulating plate according to claim 3, characterized in that: The method of obtaining a local point cloud set by using a spherical neighborhood query method includes selecting neighborhood points of the sampling point from the three-dimensional point cloud data by using a spherical neighborhood query method in a specified radius, and obtaining a local point cloud set centered on each index point; The local geometric features of each point are extracted through a multi-layer perceptron with shared weights, and the multi-scale local features are captured hierarchically using the set abstraction layer. The three-dimensional coordinates, color information, and normal vector information of each point in the point set are independently processed through the multi-layer perceptron layer with shared weights; the input dimension of each point is 9, including the three-dimensional coordinates , color information , normal vector information ; It is mapped to 64 dimensions and 128 dimensions through a shared multi-layer perceptron layer to capture the local features of each point.

5. A method for predicting defects in investment castings of aircraft engine tail jet regulating pieces according to claim 4, characterized in that: Global feature aggregation includes stacking the set abstraction layers and aggregating the local features of each point with the spatial structure features using global pooling to form the 1024-dimensional global features required for casting defect prediction.

6. A method for predicting defects in investment castings of an aircraft engine tail jet regulating plate according to claim 4, characterized in that: The fusion of local features and global features includes using a multi-layer perceptron to concatenate 1024-dimensional global features with 128-dimensional local features extracted from each set abstraction layer to form a 1088-dimensional feature vector to obtain fused features.

7. A method for predicting defects in investment castings of aircraft engine tail nozzle regulating pieces according to claim 4, characterized in that: Fusing local features with global features also includes inputting the fused features into a shared multi-layer perceptron layer, generating prediction results point by point, and using the prediction results of each point to calculate the defect prediction model loss and update the model parameters.

8. A method for predicting defects in investment castings of tail jet regulating plates of aircraft engines, using the method for predicting shrinkage and porosity of investment castings of tail jet regulating plates of aircraft engines as claimed in any one of claims 1 to 7, characterized in that: include: A numerical simulation module, which uses finite element simulation software to perform numerical simulation on the adjustment plate casting and the pouring system, obtains the location and size distribution results of casting defects, and converts the results into three-dimensional point cloud data; A model training module, which divides the three-dimensional point cloud data into a training set and a validation set, and uses the training set to train a defect prediction model based on PointNet; Model verification module, which evaluates the prediction accuracy and generalization ability of the defect prediction model through the verification set and adjusts the hyperparameters according to the verification results; The execution module converts the three-dimensional model of the casting system of the adjustment piece to be predicted into three-dimensional point cloud data, inputs the trained defect prediction model, and outputs the location of the casting defect and the prediction result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for predicting defects of an aircraft engine tail nozzle adjustment piece investment casting according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for predicting defects of an aircraft engine tail nozzle adjustment piece investment casting according to any one of claims 1 to 7 are implemented.

Images