Search

This study aims to systematically evaluate the safety of a novel L. plantarum LPJZ-658 explored on whole-genome sequence analysis, safety, and probiotic properties assessment. Whole genome sequencing results demonstrated that L. plantarum LPJZ-658 consists of 3.26 Mbp with a GC content of 44.83%. A total of 3254 putative ORFs were identified. Of note, a putative bile saline hydrolase (BSH) (identity 70.4%) was found in its genome. In addition, the secondary metabolites were analyzed, and one secondary metabolite gene cluster was predicted to consist of 51 genes, which verified its safety and probiotic properties at the genome level. Additionally, L. plantarum LPJZ-658 exhibited non-toxic and non-hemolytic activity and was susceptible to various tested antibiotics, indicating that L. plantarum LPJZ-658 was safe for consumption. Moreover, the probiotic properties tests confirm that L. plantarum LPJZ-658 also exhibits tolerance to acid and bile salts, preferably hydrophobicity and auto-aggregation, and excellent antimicrobial activity against both Gram-positive and Gram-negative gastrointestinal pathogens. In conclusion, this study confirmed the safety and probiotic properties of L. plantarum LPJZ-658, suggesting it can be used as a potential probiotic candidate for human and animal applications.

A Thesis Submitted to the Faculty of Graduate Studies and Research In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Computer Science, University of Regina. xv, 316 p. This thesis proposes a theory of spatial granular computing. A granule is composed of one or more atomic granules, and the rationality of this definition is expounded from four aspects: simplicity, applicability, measurability, and visualization. A projection is introduced, which is a one-to-one mapping from granules to points in an nD unit hypercube. Based on this, some basic questions of spatial granular computing are discussed. When discussing the rough set models from the perspective of micro knowledge space, all the models can be easily applied to incomplete information systems and continuous information systems as well. For any equivalence granule in the macro knowledge space, its spatial rough granule model is introduced and the induced three regions are obtained from each of the three planes. Expansions of the spatial rough granule model are examined in three aspects of binary relation, domain and approximation space. Not only are (fuzzy) tolerance relation, (fuzzy) equivalence relation, and coarse-fine relation but also all operations and models defined by relation matrices, and the properties can be easily proved. By focusing on relation matrices and coarse-fine relation, the reduction of knowledge is strictly related to the general concept of independence, so we can use the bases of micro and macro knowledge spaces for granular analysis, which can successfully solve the bottleneck of computing reducts. The measures are divided into granularity and fineness. All measure functions that satisfy nonnegativity, monotonicity and norm invarant can be generalized to measure granules. Subsethood holds twelve axioms, and it can be generalized and extended to conditional granularity (entropies) and conditional fineness (entropies) which are used to measure the coarse-fine relation between two granules and satisfy twelve corresponding axioms. Different definitions of similarity and difference may have different axioms which can be divided into three aspects: symmetry, boundary conditions and monotonicity. Nonnegativity can be replaced by boundary conditions, and any one of the kinds of triangle inequalities can be used as monotonic condition. Structural problems have always been a major challenge in computational science. This thesis explores the multidimensional structures from the perspective of granular computing. Any nD object can be viewed from different dimensions, and it corresponds to an adjacency relation matrix in each view. By discussing the relation between geometry and topology, granular topology algorithm is proposed and it is the general process of granular analysis that all algorithms should follow. By projecting a hypergraph to a hyperedge project graph, all adjacency relations between vertices of the hypergraph are naturally revealed by the meet of the relation vectors corresponding the vertices of the project graph. In this way, it can not only largely reduce the dimension of tensor but also be used to represent any (directed) hypergraph. Two kinds of graph/digraph information systems are defined, each graph/digraph information system can produce one micro and one macro knowledge spaces in which micro and macro rough graph/digraph models are built on its vertices and its edges respectively. This thesis introduces how to obtain the vertex and edge adjacency relation matrices of a quotient graph/digraph, performs structural analysis for fuzzy granules based on their corresponding fuzzy relation matrices respectively as well as for fuzzy graphs/digraphs based on their corresponding vertex and edge adjacency relation matrices respectively, and generalize all measure methods and operations for (equivalence) granules to (equivalence) graphs/digraphs. Student yes

PurposeTo investigate the quantitative changes in iris and retinal blood flow indices after femtosecond laser-assisted in situ keratomileusis (FS-LASIK) and small-incision lenticule extraction (SMILE).MethodsSeventy-nine patients who underwent FS-LASIK or SMILE were enrolled between July 2020 and September 2020. Participants were followed-up 1 day pre-operatively and 1 week, 1 month, 3 months post-operatively. Optical coherence tomography angiography (OCTA) was used to acquire and quantify the iris and retinal blood flow indices.ResultsThe iris vessel area density (VAD) and vessel skeleton density (VSD) decreased on post-operative day 1 but recovered on day 7. In both cases, the pupil diameter was positively associated with the post-operative iris blood flow indices (p = 0.0013, p = 0.0002). The retinal VAD and VSD in the superficial and deep capillary plexuses decreased after surgery and failed to recover after 90 days. The SMILE group showed significantly lower iris and retinal blood flow indices than the FS-LASIK group. For both procedures, axial length (p = 0.0345, p = 0.0499), spherical equivalence (p = 0.0063, p = 0.0070), and suction duration (p = 0.0025, p = 0.0130) were negatively correlated with the post-operative VAD and VSD.ConclusionsThe SMILE and FS-LASIK procedures induced a short-term decrease in the iris and retinal blood flow indices, although patients finally showed full visual recovery. This phenomenon should be carefully considered, especially in patients prone to anterior segment lesions.

Defective insulator detection is an essential part of transmission line inspections based on unmanned aerial vehicles. It can timely discover insulator defects and repair them to avoid a power transmission accident. The detection speed of defective insulators based on artificial intelligence directly affects inspection efficiency. To improve the detection speed of defective insulators based on YOLOv5s, an improved detection method with faster detection speed and acceptable precision is proposed. First, a new ResNet unit with three branches is designed based on depthwise separable convolution with kernel three and average pooling. To reduce parameters, the new ResNet unit is used to replace the original ResNet unit used in the CSP1_X module in YOLOv5s. Besides, we also introduce channel shuffle in the CSP1_X module to facilitate the flow of feature information from different channels. Second, a new residual CBL module is designed based on depthwise separable and standard convolution. The new residual CBL module is used to replace the two CBL modules used in the CSP2_X module in YOLOv5s to reduce parameters and extract more useful features. Third, we design a separate, coordinated attention module by introducing location information into channel attention. The new attention module is added to the end of the CSP2_X module to improve the ability to extract insulator location information. Besides, we also use convolution to replace the focus model to reduce computation. Compared with defective insulator detection methods, the proposed method has smaller parameters, floating-point operations per second, and higher frames per second. Although it has lower mean precision, it has a faster detection speed. Besides, the increase in detection speed is greater than the decrease in mean precision.

The design of routing protocol method of a wireless sensor network is very important for network performance. It wants to make the sensor network have a long lifetime, a partition-based cluster structure routing protocol is proposed. The proposed protocol uses a cluster structure instead of a chain structure. It selects the wireless sensor node that the current energy has the largest value to be used as the cluster head sensor node in each partition. The minimum spanning tree method is used to form the routing for connecting the clusters and sink sensor node. The simulation results presented in this paper prove that our improved protocol method exhibits better performance than existing algorithms in the aspect of prolonging network lifetime.

The energies of sensor nodes are not unlimited for wireless sensor networks at most condition. In order to make the sensor nodes have better energy efficiency and lengthen sensor nodes lifetime of wireless networks, some protocols are proposed. To further more make the zonal-stable election method have better performance, an improved zonal-stable election method is proposed. It divides the communication zone into three zones. The normal sensor nodes are randomly arranged in intermediate zone, the remaining sensor nodes are advanced nodes which are arranged in the two zones left. In the traditional protocol, the numbers of placed advanced nodes are same in comparing the two zones, but the numbers of placed advanced nodes are different in those two zones in the proposed protocol to improve the energy efficiency. It also introduces the residual energy of node to lengthen the sensor nodes lifetime of wireless network during the cluster head selection. Compared with other methods, the proposed method has longer sensor node lifetime and stability period.

To improve classification accuracy of the multiple power quality disturbances, a method is proposed based on the wavelet transform and the particle swarm optimization-support vector machine. Firstly, wavelet transform is applied to extract the wavelet energy difference of multiple power quality disturbances, and the extracted energy difference is used as feature vector. Secondly, particle swarm method is used to optimize the support vector machine to enhance the accuracy of classification. The simulation results indicate that the proposed method is effective to the classification of the seven single power quality disturbances and four multiple power quality disturbances. Compared with the original method, the new method has higher classification accuracy and strong ability to resist noise.

Artificial Bee Colony (ABC) algorithm is a widely used optimization algorithm. However, it is good in exploration but poor in exploitation. To improve the optimization performance of ABC algorithm, two novel search mechanisms served as the search equations are proposed. Firstly, the improved algorithm integrates the information of previous best solution into the search equation in the onlookers to improve the exploitation. Secondly, the search policy of the scouts is modified, the distance between the optimal and the worst solutions is utilized to replace the worst solution, so as to avoid producing worse solution. The result shows that the proposed algorithm has higher precision and stability compared with other algorithms.