In real industrial processes, factors, such as the change in manufacturing strategy and production technology lead to creation of multimode processes continuous emergence new modes. Although SCADA system has accumulated a large amount historical data, which can be used for modeling monitoring certain extent, it is difficult model learned from data adapt emerging modes, resulting mismatch. On other hand, updating with modes allows continuously match but may cause lose ability represent...

Smart manufacturing has become mainstream in the development of industry, where Industrial Internet Things plays a critical role. In this article, systematic intelligent technique for procurement supply chain (PSC) optimization is proposed. technique, an integrated approach based on variational mode decomposition and long short-term memory network used to predict market price. Considering factors, such as production plan fluctuation, multiperiod dynamic purchasing model built. A stacked...

With the digital transformation of process manufacturing, identifying system model from data and then applying to predictive control has become most dominant approach in control. However, controlled plant often operates under changing operating conditions. What is more, there are unknown conditions such as first appearance conditions, which make traditional methods based on identified difficult adapt Moreover, accuracy low during condition switching. To solve these problems, this article...

Commodity prices are important factors for investment management and policy-making, price forecasting can help in making better business decisions. Due to the complex volatile nature of market, commodity tend change frequently fluctuate violently, often influenced by many potential with strong nonstationary nonlinear characteristics. Thus, it is difficult obtain satisfactory prediction effects only using historical data individually. To address this problem, a novel dynamic method based on...

The evolution of a cooperative strategy on multilayer networks is arousing increasing concern. Most the previous studies assumed that agents can only choose cooperation or defection when interacting with their partners, whereas actual provisions in real world scenarios might not be discrete, but rather continuous. Furthermore, evolutionary game, often make use memory which keeps most successful past, as well best current gained by directed neighbors, to find available strategies. Inspired...

Complex dynamic network is a representative model for the interactions of complex system, such as Internet network, smart grid, and biological network. Many studies have investigated dynamics in networks control networks. Among these works, an accurate topology essential prerequisite. Therefore, reconstruction from measured node data important yet challenging. By analyzing extracting underlying feature unweighted undirected networks, we propose structured compressive sensing method that...

At present, network model is a general framework for the representation of complex system, and its structure fundamental prerequisite control other applications networked system. Due to advent Big Data era, scale expanding sharply. Obviously, traditional centralized reconstruction methods require high-performance computing resources can hardly be suitable in practice. Therefore, it challenge reconstruct large-scale networks with limited resources. To resolve problem, distributed local method...

Non-ferrous metals, as important basic raw materials, are the strategic supports for national economic development. For non-ferrous metal smelting enterprises, material procurement is focal and most session. Due to fluctuation of production volumes future changes in raw-material prices, cost materials high with a risk shortage. In this paper, we propose multi-period rolling robust model considering price demand uncertainties. particular, design data-driven method construct budget-based...

Deep neural networks (DNNs) as one of the key enabling technologies have been widely used in industrial artificial intelligence (IAI). However, recent research has revealed that they are quite vulnerable to adversarial attacks, arousing serious concerns about DNNs' robustness many IAI-driven applications such video analysis tasks. Considering attack efficiency and effectiveness, it is essential study sparse examples. Nevertheless, current methods' performance limited by insufficient sparsity...

Zinc is an indispensable base material for the development of national economy and construction defense industry, price forecasting great significance investors, policy makers researchers. Considering complexity, dynamic strong nonlinearity zinc changes, it usually affected by a variety external factors, difficult to obtain satisfactory effect only analyzing underlying pattern historical data changing. To solve aforementioned problem, novel method based on factor selection long short-term...

Industrial processes usually have strong nonlinearity and dynamics, which lead to extreme difficulties in the traditional control method for effective control. In this study, a nonlinear adaptive predictive based on wavelet transform bi-directional long short-term memory(WT-BiLSTM) is proposed. Wavelet firstly conducted decompose original data of system output into multiple sub-sequences different frequency bands, can reduce non-stationarity time series. Secondly, prediction input matrix...

Nonferrous metal industry is the foundation of China’s substantial economy and plays a key role in national defense construction. Industrial software crucial for high-quality development nonferrous associated with in-depth implementation strategies. Currently, industrial significantly restricted by lack knowledge models. Hence, we propose method constructing an intelligent model library industry. Considering meta-model-driven engineering, define metallurgical meta-model its attributes,...