Neuro-genetic optimization of disc brake performance at elevated temperatures

Finite element analysis on squeal-noise in railway applications

“…This is the assumption underlying the studies carried out by Spurr [71], Earles and Soar [72,73]. In this paper, we show that if two modes are geometrically coupled, for increasing values of friction coefficient, frequencies tend to coincide [74][75]. Once the two modes couple, they become unstable.…”

Management of stresses in the rails on railway bridges

“…vehicle: passenger and/or freight traffic (maximum speed, maximum axle load), and traffic regime (braking/acceleration), track: ballasted (CWR, rail fastening system, concrete sleepers, ballast), and ballastless (CWR, rail fastening system, concrete/asphalt slab, concrete sleepers if necessary, hydraulic stabilized base layer, and bridge: alignment (straight or curve), and structure parameters (static arrangement -system, expansion length, layout and stiffness of supports, deck stiffness). The main contribution of the proposed algorithm is the development of technical solutions of the railway bridges as an integral part of the railway route design by respecting current rail manufacturing technology, climate conditions, traffic regime and avoiding of the rail expansion joints.Mutual adjustment of railway alignment, bridge and track structure increases:(a) ride comfort due to the omittingof the rail expansion joints whenever it is technically possible, and (b) traffic safety through satisfying local climate, topographic and geotechnical requirements [22], as well as considering the rolling stockparameters (i.e.different braking systems [23,24] installed in rail vehicles).…”

Soft computing techniques

“…Static (feedforward) networks have no feedback elements and contain no delays -the output is calculated directly from the input through feedforward connections [10]. Dynamic neural networks are generally more powerful than static networks since they have a sort of memory that can remember the past values and states of the network [12,13]. The output of a dynamic network depends on the current input values as well as on the previous inputs, outputs or states of the network.…”

“…The network function is largely determined by the interconnections between the neurons (the connection weights) [14]. According to [13,[15][16], the output of a dynamic network is affected differently by the weights. Two different effects can be considered: (i) a direct effect (a change in a weight causes an immediate change in the output at the current time step), and (ii) an indirect effect (which implies using dynamic backpropagation to compute the gradients, which is computationally more intensive).…”