With the growing complexity of industrial processes, the scale of production processes tends to be large. The significant amount of measurement data in large‐scale processes poses challenges in data collection, management, and storage. In order to perform effective process monitoring in large‐scale processes, the distributed process monitoring strategy is widely applied. Meanwhile, product quality is an important indicator for industrial production. Therefore, a novel quality‐based distributed process monitoring scheme is proposed. Firstly, the Girvan‐Newman (GN) algorithm in complex network divides process variables into multiple sub‐blocks. Secondly, greedy algorithm‐based high‐dimensional mutual information (HDMI) is used to extract quality‐related variables in each sub‐block, through which the irrelevant and redundant variables are eliminated. Thirdly, the decomposed modified partial least squares (DMPLS) approach is used to detect whether a fault is quality‐related or not in each sub‐block. Finally, the Bayesian inference strategy is adopted to combine the detection results of all sub‐blocks. The effectiveness of the distributed DMPLS approach is illustrated through a numerical simulation and the Tennessee Eastman (TE) process. The results show the superiority of our proposed monitoring scheme.

Traditional process monitoring methods construct a single monitoring model to detect if a fault happened. But the development of industrial technology has made industrial processes increasingly complex and huge. When considering local or quality-related faults, especially tiny faults in complex large-scale industrial processes, it is difficult to accurately detect these faults using a single monitoring model. Therefore, a novel distributed process monitoring framework for quality-related fault detection is proposed in this paper. To availably deal with a large number of process variables in large-scale processes, this paper introduces the idea of community partitioning of complex networks to carry out subblock division. Here, the fast unfolding algorithm is used for multiblock division of process variables. Then, the modified principal component regression (MPCR) model is constructed in each subblock to detect quality-related and unrelated faults. To get an intuitive monitoring result, Bayesian fusion based on probability weighting is applied to combine the detection results of all subblocks. Afterward, the cumulative contribution plot based on multiblock fast-MPCR is used for fault diagnosis. The benefits of distributed MPCR models are illustrated through a numerical experiment and the Tennessee Eastman (TE) process; the results indicate the superiority compared with other monitoring methods.

No abstract

No abstract

The permeability of sand covered with geotextile is affected by the permeability of geotextile, which is related to the tensile state of the geotextile. Considering the weaving mode of geotextile, the effects of warp tension and weft tension on the permeability of sand covered with geotextile were studied by experiment. Four different specifications of geotextiles were selected for warp and weft tension respectively. The changes of permeability parameters of sand covered with geotextile under non tension, warp and weft tension were measured by vertical permeability instrument, and the effects of warp and weft tension on permeability parameters such as seepage velocity, sand loss and gradient ratio were analyzed. The test results show that the water permeability and anti silting performance of the geotextile increase with the increase of tensile strain, and the soil conservation performance decreases with the tensile strain increasing. Meanwhile, the relationship between permeability and warp tensile strain is not monotonic. When the warp tensile strain 3%, the water permeability and anti silting performance of geotextile are the weakest, and the soil conservation performance is the strongest.