Aero-engine damage is an important factor affecting flight safety. There are two main problems in the current computer vision-based damage detection of engine borescope image:one is that the complex background of borescope image makes the model detect the damage with low accuracy; the other one is that the data source of borescope image is limited, which leads to fewer detectable classes for the model. In order to solve these two problems, a Mask R-CNN (Mask Region-based Convolutional Neural Network) based progressive detection network from binary classification to multiple classification was proposed for aero-engine damage images. By adding a binary classification detection branch to the Mask R-CNN, firstly, the damage in the image was detected in binary way and regression optimization was performed to the localization coordinates. Secondly, the original detection branch was used to progressively perform multiple classification detection, so as to further optimize the damage detection results by regression and determine the damage class. Finally, instance segmentation was performed to the damage through the Mask branch according to the results of multiple classification detection. In order to increase the detection classes of the model and verify the effectiveness of the method, a dataset of 1 315 borescope images with 8 damage classes was constructed. The training and testing results on this set show that the Average Precision (AP) and AP75 (Average Precision under IoU (Intersection over Union) of 75%) of multiple classification detection are improved by 3.34% and 9.71% respectively, compared with those of Mask R-CNN. It can be seen that the proposed method can effectively improve the multiple classification detection accuracy for damages in borescope images.

The number of defect region samples is far less than that of the normal region samples in aircraft engine borescope inspection image defect detection task, and the defect samples cannot cover the whole sample space, which result in poor generalization of the detection algorithms. In order to solve the problems, a new interactive data augmentation method based on style transfer technology was proposed. Firstly, background image and defect targets were selected according to the interactive interface, and the informations such as size, angle and position of the target needed to be pasted were specified according to the background image. Then, the style of background image was transferred to the target image through style transfer technology, so that the background image and the target to be detected had the same style. Finally, the boundary of the fusion region was modified by Poisson fusion algorithm to achieve the effect of natural transition of the connected region. Two-class classification and defect detection were conducted to verify the effectiveness of the proposed method. The testers achieve 44.0% classification error rate for the two-class classification on the dataset with real images and augmented images averagely. In the detection task based on Mask Region-based Convolutional Neural Network (Mask R-CNN) model, the proposed method has the Average Precision (AP) of classification and segmentation improved by 99.5% and 91.9% respectively compared to those of the traditional methods.

There are two problems to be solved in the traditional System Theoretic Process Analysis (STPA) method. One is the lack of automation means of realization, the other is the ambiguity problem caused by natural language result analysis. To solve these problems, a software safety requirement analysis and verification method based on STPA was proposed. Firstly, the software safety requirements were extracted and converted into formal expressions by the algorithm. Secondly, the state diagram model was built to describe the logic of software safety control behaviors and converted the logic into the readable formal language. Finally, the formal verification was carried out by model checking technology. The effectiveness of the method was verified by the case of a weapon launch control system. The results show that the proposed method can generate the safety requirements automatically and perform formal verification to them, avoid the dependence on manual intervention and solve the natural language description problems in traditional methods.

The physical layer secure transmission in two-way multi-relay system can not obtain the accurate Channel State Information (CSI) of eavesdroppers. In order to solve the problem, a robust joint physical layer secure transmission scheme of multi-relay cooperative beamforming and artificial noise was proposed to maximize the secrecy sum rate in the worst case of channel state under the total power constraint of system. In the proposed scheme, the problem to be solved was a complex non-convex optimization problem. The alternating iteration and Successive Convex Approximation (SCA) methods were used for the alternating optimization iteration of beamforming vector, artificial noise covariance matrix and source node transmit power, and the optimal solution of the above problem was obtained. The simulation results verify the effectiveness of the proposed scheme and show that the proposed scheme has better security performance.

Traditional multi-dimensional text clustering generally extracts features from text contents, but seldom considers the interaction information between users and text data, such as likes, forwards, reviews, concerns, references, etc. Moreover, the traditional multi-dimension text clustering mainly integrates linearly multiple spatial dimensions and fails to consider the relationship between attributes in each dimension. In order to effectively use text-related user behavior information, a Multi-dimensional Text Clustering with User Behavior Characteristics (MTCUBC) was proposed. According to the principle that the similarity between texts should be consistent in different spaces, the similarity was adjusted by using the user behavior information as the constraints of the text content clustering, and then the distance between the texts was improved by the metric learning method, so that the clustering effect was improved. Extensive experiments conduct and verify that the proposed MTCUBC model is effective, and the results present obvious advantages in high-dimensional sparse data compared to linearly combined multi-dimensional clustering.

With the rapid increase of text data sources, topic mining for multi-source text data becomes the research focus of text mining. Since the traditional topic model is mainly oriented to single-source, there are many limitations to directly apply to multi-source. Therefore, a topic model for multi-source based on Dirichlet Multinomial Allocation model (DMA) was proposed considering the difference between sources of topic word-distribution and the nonparametric clustering quality of DMA, namely MSDMA (Multi-Source Dirichlet Multinomial Allocation). The main contributions of the proposed model are as follows:1) it takes into account the characteristics of each source itself when modeling the topic, and can learn the source-specific word distributions of topic k; 2) it can improve the topic discovery performance of high noise and low information through knowledge sharing; 3) it can automatically learn the number of topics within each source without the need for human pre-given. The experimental results in the simulated data set and two real datasets indicate that the proposed model can extract topic information more effectively and efficiently than the state-of-the-art topic models.

With the emergence of various social networks, the social media network data is analyzed from the perspective of variety by more and more researchers. The data fusion of multiple social networks relies on user identification across social networks. Concerning the low utilization problem of heterogeneous relation between social networks of the traditional Friend Relationship-based User Identification (FRUI) algorithm, a new Weighted Hypergraph based User Identification (WHUI) algorithm across social networks was proposed. Firstly, the weighted hypergraph was accurately constructed on the friend relation networks to describe the friend relation and the heterogeneous relation in the same network, which improved the accuracy of presenting topological environment of nodes. Then, on the basis of the constructed weighted hypergraph, the cross network similarity between nodes was defined according to the consistency of nodes' topological environment in different networks. Finally, the user pair with the highest cross network similarity was chosen to match each time by combining with the iterative matching algorithm, while two-way authentication and result pruning were added to ensure the recognition accuracy. The experiments were carried out in the DBLP cooperation networks and real social networks. The experimental results show that, compared with the existing FRUI algorithm, the average precision, recall, F of the proposed algorithm is respectively improved by 5.5 percentage points, 3.4 percentage points, 4.6 percentage points in the real social networks. The WHUI algorithm can effectively improve the precision and recall of user identification in practical applications by utilizing only network topology information.

The existing wireless network performance optimization methods are mainly based on the correlation analysis between indicators, and cannot effectively guide the design of optimization strategies and some other interventions. Thus, a Canonical Causal Inference (CCI) algorithm was proposed and used for wireless network performance optimization. Firstly, concerning that wireless network performance is usually presented by numerous correlated indicators, the Canonical Correlation Analysis (CCA) method was employed to extract atomic events from indicators. Then, typical causal inference method was conducted on the extracted atomic events to find the causality among the atomic events. The above two stages were iterated to determine the causal network of the atomic events and provided a robust and effective basis for wireless network performance optimization. The validity of CCI was indicated by simulation experiments, and some valuable causal relations of wireless network indicators were found on the data of a city's more than 30000 mobile base stations.

Concerning the problem of high time complexity of traditional recommendation algorithms, a new recommendation model based on Latent Dirichlet Allocation (LDA) model was proposed. It was a data mining model applied to Book Recommendation (BR) in library management systems, named Book Recommendation_Latent Dirichlet Allocation (BR_LDA) model. Through the content similarity analysis of historical borrowing data of the target borrowers with other books, other books which had high content similarities with historical borrowing books of the target borrowers were gotten. Through the similarity analyses performed on the target borrowers' historical borrowing data and historical data from other borrowers, historical borrowing data of the nearest neighbors were gotten. Books which the target borrowers were interested in could be finally gotten by calculating the probabilities of the recommended books. In particular, when the number of recommended books is 4000, the precision of BR_LDA model is 6.2% higher than multi-feature method and 4.5% higher than association rule method; when the recommended list has 500 items, the precision of BR_LDA model is 2.1% higher than collaborative filtering based on the nearest neighbors and 0.5% higher than collaborative filtering based on matrix decomposition. The experimental results show that this model can efficiently mine data of books, reasonably recommend new books which belong to historical interested categories and new books in potential interested categories to the target borrowers.