In recent years, the crack fault is one of most common faults in rotor system and it still a challenge for position diagnosis hollow shaft system. this paper, method based on Convolutional Neural Network deep metric learning (CNN-C) proposed to effectively identify Center-loss function used enhance performance neural network. Main contributions include: Firstly, dynamic response dual-disks obtained. The analysis results show that will cause super-harmonic resonance, peak value closely...

Abstract The crack fault is one of the most common faults in rotor system, and researchers have paid close attention to its diagnosis. However, studies focus on discussing dynamic response characteristics caused by rather than estimating depth position based obtained vibration signals. In this paper, a novel diagnosis location method for dual-disk hollow shaft system Radial basis function (RBF) network Pattern recognition neural (PRNN) presented. Firstly, model with breathing suitable...

A novel balancing method for rotor based on unsupervised deep learning is proposed in this paper. The architecture of the network described. In network, compared to supervised additional convolution layers are applied not only inverse mapping but also identifying unbalanced force without labeled data. equivalent value and position imbalances two correction planes obtained. case study a with discs supported by sliding bearings conducted. Preset balanced well method. And, using state values at...

In this paper, a simplified dynamic model of dual-rotor system coupled with blade disk is built, and the effects parameters an aircraft engine on characteristics are studied. methodology, as cantilever structure, dynamical equations obtained by means finite element method. The amplitude-frequency response curves orbits shaft centre-vibration shape diagram used to analyze system. results indicate that properties blades have huge impacts critical speed other With increase length blade,...

The issue of aviation safety has always received a great attention and focus, birds are also an important in safety. Nowadays, drones have emerged share the same airspace with at low altitudes. problems associated should be taken into account. For example, small can misused for illegal activities threat from them is on rise. Driven by this situation, we used data provided ICASSP Drone-vs-Bird detection Grand Challenge drone method adding shallow feature pyramid network model SSD [1]...

This paper focuses on the nonlinear response characteristics of an aero-engine dual-rotor-bearing system with flexible coupling misalignment faults in low-pressure rotor (rotor 2). The motion equations are formulated by using finite element method, which excitations induced faults, dual-unbalance two rotors (the high-pressure system-rotor 1 and 2), gravities, force inter-shaft bearing all considered. By numerical calculation solved to obtain responses system. Then, for alignment analyzed,...

This paper deals with the bifurcation and chaotic dynamic characteristic of a single‐machine infinite‐bus (SMIB) power system under two kinds harmonic excitation disturbance, which are induced by external periodic load outer mechanical disturbance. By applying Melnikov’s method, threshold value for occurrence motion is provided. In addition, boundary surface given. The efficiency criteria obtained in this verified diagram, phase portraits, Poincaré section, frequency spectrum. results will...

is paper takes an asymmetric ball bearing supported rotor system with the application of a squeeze film damper (SFD) subjected by unbalanced force and parametric excitation (varying compliance) as research object.e dynamical behaviours in multimode resonance regions SFD are investigated.In each region, amplitude-frequency curves obtained under conditions without SFD, well spectrum analysis solutions.rough results comparisons two cases, damping effects for kind illustrated.Meanwhile,...

Because of the low resolution and limited information small objects, computing resources are in practical applications, object detection is still challenging. In order to improve accuracy detection, we propose a new method. It's included shallow feature pyramid network with an extraction block at features fused multi-scale semantic information. Further, context attention mechanism adopted make focus on significant area. We one top five teams Drone-vs-Bird Detection Grand Challenge. The...

Rotor dynamic balancing is essential in rotor industrial. The conventional methods, including the influence coefficients method and modal method, are effective, but lack economy sufficient usage of data. To overcome disadvantages a using unsupervised deep learning without weight trails had been proposed. proposed network could identify unbalanced forces from data observed just one run labels. validate novel an experimental rig well-designed established. Experimental validation comparison...

A complex discrete dynamical model is built for aero-engine dual-rotor systems supported by six bearings, and two kinds of simplified models were obtained the structure dimensionality reduction method. By using finite element method, accurate was built, to gain four-disc (Model B) three-disc C) based on center mass concentration The critical whirling speed compared indicate that correctness results show that, Model B, 1th–3th relative error correspondingly less than 5%, 22%, 10%, indicated...

The squeeze film damper is usually adopted in the rotor system to suppress vibrating motion of system. In this work, not only are physical parameters analyzed but also parameters, like number dampers used and implementation positions, analyzed. amplitude‐frequency curves obtained by conducting simulation a dual‐rotor, intershaft, oil force concatenated model. Through analysis comparisons results, vibration suppression effects with different parameter configurations summarized. This work...