Fault Diagnosis of the Dynamic Chemical Process Based on the Optimized CNN-LSTM Network.

Deep learning provides new ideas for chemical process fault diagnosis, reducing potential risks and ensuring safe process operation in recent years. To address the problem that existing methods have difficulty extracting the dynamic fault features of a chemical process, a fusion model (CS-IMLSTM) based on a convolutional neural network (CNN), squeeze-and-excitation (SE) attention mechanism, and improved long short-term memory network (IMLSTM) is developed for chemical process fault diagnosis in this paper. First, an extended sliding window is utilized to transform data into augmented dynamic data to enhance the dynamic features. Second, the SE is utilized to optimize the key fault features of augmented dynamic data extracted by CNN. Then, IMLSTM is used to balance fault information and further mine the dynamic features of time series data. Finally, the feasibility of the proposed method is verified in the Tennessee-Eastman process (TEP). The average accuracies of this method in two subdata sets of TEP are 98.29% and 97.74%, respectively. Compared with the traditional CNN-LSTM model, the proposed method improves the average accuracies by 5.18% and 2.10%, respectively. Experimental results confirm that the method developed in this paper is suitable for chemical process fault diagnosis.

Construction of 2D/2D layered g-C3N4/Bi12O17Cl2 hybrid material with matched energy band structure and its improved photocatalytic performance.

A series of visible-light-induced 2D/2D layered g-C3N4/Bi12O17Cl2 composite photocatalysts were successfully synthesized by a one step chemical precipitation method with g-C3N4, BiCl3 and NaOH as the precursors at room temperature and characterized through XRD, FTIR, XPS, TEM, BET and UV-vis DRS measurements. The results of XRD, FTIR and XPS indicated that g-C3N4 has been introduced in the Bi12O17Cl2 system. The TEM image demonstrated that there was strong surface-to-surface contact between 2D g-C3N4 layers and Bi12O17Cl2 nanosheets, which contributed to a fast transfer of the interfacial electrons, leading to a high separation rate of photoinduced charge carriers in the g-C3N4/Bi12O17Cl2 system. Rhodamine B was considered as the model pollutant to investigate the photocatalytic activity of the resultant samples. The g-C3N4/Bi12O17Cl2 composite showed a clearly improved photocatalytic degradation capacity compared to bare g-C3N4 and Bi12O17Cl2, which was ascribed to the interfacial contact between the 2D g-C3N4 layers and Bi12O17Cl2 sheet with a matched energy band structure, promoting the photoinduced charges' efficient separation. Finally, combined with the results of the trapping experiment, ESR measurements and the band energy analysis, a reasonable photocatalytic mechanism over the 2D/2D layered g-C3N4/Bi12O17Cl2 composite was proposed.

Pd-Ag alloy dendritic nanowires: fabrication and application in hydrogen sensor.

Dendritic Pd-Ag alloy nanowires (Pd-Ag DNWs) with different bimetallic composition have been directly self-assembled on the microelectrodes from mixed solutions of palladium and silver ions by applying an alternating current (AC) field. The size and morphology of the final product were controlled via adjusting the electrodeposition parameters and the metal ion concentration ratio. The influence of three factors on the bimetallic composition in alloy nanowires was examined. Structural characterizations suggest that there was a preferential growth along (200) and (111) directions, leading to the formation of Pd-Ag nanodendritic wires with 150-200 nm in stem and branch diameter. The mechanism of forming the DNWs is discussed. The DNWs were studied as sensing materials for the detection of hydrogen, and they were found to exhibit good sensitivity and reproducibility. The Pd-Ag sensing materials with 22.2 wt% Ag content possessed were also found to display a rapid response time of less than 1 min for 4% (V/V) hydrogen concentration.