Applied Intelligence最新文献

The lubrication system supplies lubrication and cleans the rotating parts and contacting machinery during the operation of an aero-engine. It is crucial to maintain an adequate amount of lubricant by predicting and analyzing the consumption rate to ensure endurance and maintenance programs are effective. This paper examines the combination of temporal and non-temporal data that impact the characteristic parameters of lubricant consumption rate in aero-engines. Our study focuses on the merging of LSTM (Long Short-Term Memory) + LightGBM (Light Gradient Boosting Machine) + CatBoost, and uses KPCA dimensionality reduction optimization, along with Stacking for the fusion of a multi-feature regression prediction algorithm. On the one hand, this study utilizes integrated learning to fuse feature extractions from LSTM for temporal information and non-temporal information by GDBT (Gradient Boosting Decision Tree). This approach considers the trend and distribution of feature samples to develop a more robust feature extraction method. On the other hand, the integrated learning framework incorporates multi-decision making and feature importance extraction to strengthen the mapping relationship with the predicted output of lubrication oil consumption rate, enabling regression prediction. The algorithm for regression prediction has been executed and the results indicate a final regression prediction MAPE (Mean Absolute Percentage Error) of less than 3%. MSE and RMSE reached 1.28% and 1.33%, the results are in an ideal state. The algorithms used in this paper will be applied in the future to aero-engine lubricant systems and eventually to engines in general.

Unsupervised feature selection (UFS) is a fundamental and indispensable dimension reduction method for large amount of high-dimensional unlabeled data samples. Without label information, the manifold learning technique is leveraged to compensate for the lack of discrimination with the selected features. However, it is still a challenging problem to capture the geometrical structure for practical data, which are often contaminated by noises and outliers. Additionally, the predetermined graph embedded UFS models suffer from the parameter tuning problem and the separated model optimization procedures. To generate more compact and discriminative feature subsets, we propose a Robust UFS model with Adaptive and Flexible (varvec{ell }_textbf{1})-norm Graph (RAFG) embedding. Specifically, the (varvec{ell }_textbf{2,1})-norm is imposed on the flexible regression term to alleviate the adverse effects of both noisy features and outliers, and (varvec{ell }_textbf{2,p})-norm regularization term is incorporated to ensure that the selected transformation matrix is sufficiently sparse. Moreover, the adaptive (varvec{ell }_textbf{1})-norm graph learning characterize the clustering distribution via consistent embeddings, which avoids time-consuming distance computations in a high-dimensional feature space. To solve the challenging problem, we propose an efficient alternative updating algorithm with an iterative reweighted strategy, together with the necessary convergence and complexity analyses. Finally, experimental results on two synthetic data and eight benchmark datasets illustrate the effectiveness and superiority of the proposed RAFG method compared with state-of-the-art methods.

Differential evolution (DE) has been proven as a simple yet powerful meta-heuristic algorithm on tackling continuous optimization problems. Nevertheless most existing DE methods still suffer from certain drawbacks including the use of ineffective mechanisms to adjust mutation strategies and their control parameters that may possibly mislead the search directions, and also the lack of intelligent guidance and reset mechanisms to escape from local optima. Therefore, to enhance the adaptability of DE-based search frameworks and the robustness on optimizing complex problems full of local optima, an adaptive and guided differential evolution (AdaGuiDE) algorithm is proposed. Essentially, the adaptability of the AdaGuiDE search framework is enhanced by three schemes to iteratively refine the search behaviour at two different levels. At the macroscopic level, the AdaGuiDE search framework revises the existing adaptive mechanism for selecting appropriate DE search strategies by counting the actual contributions in terms of solution quality. In addition, the adaption strategy is extended to the microscopic level where a penalty-based guided DE search is employed to guide the search escaping from local optima through temporarily penalizing the local optima and their neighborhood. Furthermore, a systematic boundary revision scheme is introduced to dynamically adjust the search boundary for locating any potential regions of interest during the search. For a rigorous evaluation of the proposed search framework, the AdaGuiDE algorithm is compared against other well-known meta-heuristic approaches on three sets of benchmark functions involving different dimensions in which the AdaGuiDE algorithm attained remarkable results especially on the high-dimensional and complex optimization problems. More importantly, the proposed AdaGuiDE framework shed lights on many possible directions to further enhance the adaptability of the underlying DE-based search strategies in tackling many challenging real-world applications.

Image quality assessment of authentically distorted images constitutes a indispensable part of numerous computer vision tasks. Despite the substantial progress in recent years, accuracy and generalization performance is still unsatisfactory. These challenges are primarily attributed to the scarcity of labeled images. In order to increase the amount of images for training, we use semi-supervised learning to combine labeled images and specifically selected unlabeled images. In our new training paradigm, denominated Selected Data Retrain under Regularization, the selection criteria of unlabeled images is based on the supposition that an image and a certain of its patches ought to have approximate image quality scores. Unlabeled images that meets the aforementioned criteria, named as Highly Credible Unlabeled Images, mitigate the problem of scarcity, thus, improve accuracy. However generalization may be compromised due to selection procedure’s reliance on labeled images and presence of coherent variance existed between labeled images and unlabeled images. Therefore we incorporate a sorting loss function to reduce variation within the new dataset of labeled images and specifically selected unlabeled images, and thus achieve better generalization. The effectiveness of our proposed paradigm is empirically validated using public datasets. Codes are available at https://github.com/dvstter/SDRR_IQA.

Scrap steel is a green resource that can substitute iron ore and is an important raw material in the modern steel industry. To address the many issues such as high risk, low accuracy in grading, and the susceptibility to questioning fairness in the manual inspection process of scrap steel, we propose an efficient intelligent scrap steel classification method based on machine vision, achieving accurate classification and grading of nine types of scrap steel. Firstly, a scrap steel quality inspection system was established at the scrap steel recycling site, where images of various types of scrap steel were collected and various image processing methods were employed for preprocessing, leading to the establishment of scrap steel datasets and carriage segmentation datasets. Secondly, a carriage segmentation model was built based on image segmentation technology to significantly reduce the influence of complex backgrounds of scrap steel images on classification and grading. Subsequently, an intelligent scrap steel classification grading model was established based on the attention mechanism in deep learning, combined with the Spatially Adaptive Heterogeneous Image Slicing (SAHI) image slicing prediction method, achieving accurate classification and grading of scrap steel under complex backgrounds and high-resolution images in scrap steel recycling. Finally, we conducted tests on the proposed method. Experimental results demonstrate the good generalization of our proposed method, accurately detecting various types of scrap steel, meeting the requirements of accuracy, real-time performance, and good generalization in scrap steel recycling classification and grading, achieving initial industrial application, and exhibiting significant advantages compared to traditional manual scrap steel quality inspection.

Temporal knowledge graphs (TKGs) are a form of knowledge representation constructed based on the evolution of events at different time points. It provides an additional perspective by extending the temporal dimension for a range of downstream tasks. Given the evolving nature of events, it is essential for TKGs to reason about non-existent or future events. Most of the existing models divide the graph into multiple time snapshots and predict future events by modeling information within and between snapshots. However, since the knowledge graph inherently suffers from missing data and uneven data distribution, this time-based division leads to a drastic reduction in available data within each snapshot, which makes it difficult to learn high-quality representations of entities and relationships. In addition, the contribution of historical information changes over time, distinguishing its importance to the final results when capturing information that evolves over time. In this paper, we introduce CH-TKG (Contrastive Learning and Historical Information Learning for TKG Reasoning) to addresses issues related to data sparseness and the ambiguity of historical information weights. Firstly, we obtain embedding representations of entities and relationships with evolutionary dependencies by R-GCN and GRU. On this foundation, we introduce a novel contrastive learning method to optimize the representation of entities and relationships within individual snapshots of sparse data. Then we utilize self-attention and copy mechanisms to learn the effects of different historical data on the final inference results. We conduct extensive experiments on four datasets, and the experimental results demonstrate the effectiveness of our proposed model with sparse data.

In the field of food safety, milk has become an indispensable beverage in people’s lives. Therefore, it is of great significance to detect milk freshness. Starting from the milk freshness detection, the perspective can be extended to liquid detection. Liquid detection has attracted much attention and has been applied in many fields. However, current liquid detection methods are either contact detection methods that can lead to sample damage, or require specialized instruments, expertise for operation or cumbersome hardware deployment. This paper introduces a non-contact milk freshness detection method based on image features and tensor construction. Unlike existing liquid detection methods, our method relies on ubiquitous WiFi signals to achieve non-contact and non-invasive milk freshness detection. The design intuition is that the WiFi signals will lead to different multipath propagation when passing through milk with different freshness, which can be used to detect milk freshness. We use existing commercial devices to collect WiFi signal data of milk with different freshness, denoise the collected data and transform the denoised data into multiple types of time-frequency images and spatial-temporal images, and input the images into the deep learning network to extract image features that contain richer and more comprehensive information, and then utilize the extracted image features to perform tensor construction to better retain the original time, frequency and spatial feature information, and apply 3D convolutional layer and fully connected layers for milk freshness detection. The experimental results show that our method achieves a high accuracy of 93.25% in detecting milk freshness.