Abstract:
A method for predicting fault occurrence in a mechanical system and an electrical system. The method comprises: receiving a first dataset of mechanical system condition data, the first dataset being imbalanced by having more data points in a first category than in a second category; generating a plurality of chromosomes from the second category data points in the first dataset; the plurality of chromosomes including information to enable the creation of new datasets; generating a second dataset using the plurality of chromosomes and an evolutionary algorithm, the second dataset being less imbalanced than the first dataset; and predicting fault occurrence in the mechanical system using the second dataset and a machine learning algorithm.
Abstract:
A computational engineering modelling tool having a point generation module arranged to read geometry data representing a domain to be modelled. The point generation module also generates point data for the domain having multiple boundary points located on a boundary of the domain and multiple further points spaced from the boundary within the domain. A point mutation module processes the point generation module output and generates automatically a plurality of alternative point data definitions for the domain in which the location of at least one point differs between each of the alternative point definitions. A blocking module discretizes the domain by creating multiple geometric blocks over the domain using a computational geometric operator wherein each point represents a vertex of at least one block. The blocking module outputs a discretized computational model of the domain and the tool scores the model according to a geometric attribute of the blocks.
Abstract:
A vortex detection method is described. The method comprises storing a plurality of points at locations over a region (32) in which vortex detection is to be performed. A value for each of a plurality of fluid flow parameters, such as velocity, pressure and density, is determined at each point. The points are grouped as being contained in either a vortical flow portion or non-vortical flow portion of the region according to one or more statistical distribution for said fluid flow parameters. A point (p) in a vortex core is identified according to the direction of motion of an array of further points (46) relative to said point in the vortex core. The further points (46) may surround the vortex core point.
Abstract:
A flow feature detection method is described. The method includes storing a plurality of points at locations over a region in which vortex detection is to be performed. A value for each of a plurality of fluid flow parameters, such as velocity, pressure and density, is determined at each point. The points are grouped as being contained in either a flow feature portion or normal flow portion of the region according to one or more statistical distribution for the fluid flow parameters. A point is identified as being indicative of the flow feature by identifying multiple further points at least partially surrounding the point, and determining a plane in which the flow feature is identifiable based upon the relative values of the one or more fluid flow parameter for the further points. The method may be used to detect vortices and to identify a two-dimensional plane representative of a vortex.