1,721,039 research outputs found
Multipurpose model for pantograph dynamic interaction with flexible or rigid catenary
No full textRailway transportation powered by electrical supply systems is the most employed technology to provide passengers safety and vehicles reliability during the running operations. Both the flexible and rigid catenaries are employed to power up trains through interaction with pantographs. Within this context, principal issues are wear estimation and pantograph-catenary contact loss prediction. Therefore, various modelling approaches are adopted to build simulation models for these phenomena prediction, particularly related to the contact force exerted between the pantograph and the catenary. In this paper a new and versatile analytical model of the pantograph dynamic interaction with the catenary is presented. The model is characterized by a low computational load and the pantograph model can be interfaced with both rigid and flexible catenary models. Therefore, the proposed model is a versatile solution to reproduce the pantograph-catenary interaction behavior. Simulations have been carried out in both the contact sceneries and the results are reliable and in accordance with the system dynamics
Output-only estimation of lateral wheel-rail contact forces and track irregularities
No full textThe contact between the wheel and the rail heavily affects the dynamic behaviour of railway vehicles and their running safety. The real-time knowledge of wheel-rail contact forces is tricky due to the several running conditions of the train and the combined tribological and wear characteristics of wheels and rails. Furthermore, their measurement and that of track irregularities, sensitive to the harshness of the wheel-rail contact, require dedicated sensors. Therefore, the development of strategies for estimating wheel-rail contact features is essential to perform the condition monitoring of rails to improve the safety related to the entire railway system. A nonlinear model-based estimation procedure, based on a Central Difference Kalman Filter, is proposed in this work to estimate the lateral wheel-rail contact forces and moments, including the identification of lateral track irregularities in the estimation process related to rails. Furthermore, the developed estimation technique can..
Railway pantograph contact strip monitoring through image processing techniques
No full textDuring the train operation and especially at high speed, the pantograph-catenary contact is influenced by electromechanical wear phenomena. Consequently, the overall reliability of the railway system is reduced. In particular, the pantograph contact strips are subjected to severe abrasions during the running operations. Their maintenance and inspection are usually made manually with negative consequences in terms of costs and safety. Therefore, in the last years, no-contact monitoring and measurements techniques for the pantograph-catenary elements have been developed. A practicable methodology to measure the pantograph contact strip thickness is proposed in this paper. The employed approach is based on image processing techniques useful to detect geometrical variations of the pantograph contact strips. The no-contact measurement system is made through a cost-effective single camera. Several images of the pantograph equipped with unworn and worn contact strips are employed to verify the reliability of the proposed technique
A novel adaptive-gain technique for high-order sliding-mode observers with application to electro-hydraulic systems
Full text linkIn this paper, a novel technique for adaptive gain selection in high-order sliding mode observers to estimate the state and the disturbance acting on a hydraulic actuators is designed and tested. The proposed gain adaptation technique is based on the evaluation of the absolute value of the errors computed between the real measurements available on the system and their estimation provided by the observer. Experimental results are reported to validate the effectiveness of the proposed technique in a real context. The experiment are conducted in two different conditions, i.e. without external load and with an unknown visco-elastic seismic isolator attached to the hydraulic actuator. The proposed gain-adaptation method is also compared with the case of manually-tuned gains and a classic gain adaptation technique reported in literature, in which the gain adaptation is based on fixed variation rates. The proposed method provides performance comparable or superior to both manual and classic adaptive gain selection. Moreover, with respect to classic techniques, the proposed one has the advantage of improving the convergence rate in case of large estimation errors and limiting the growing of observer gains in case of noisy measurements
A supervised machine learning framework for smart tires
No full textArtificial intelligence techniques are today among the most important challenges for the future due to their ability to analyze big data and identify key characteristics. In transportation, the growing amount of data measured on vehicles and infrastructures constitute a wealth of information useful for energy, safety and environmental sustainability. In this paper, a supervised machine learning methodology applied to smart tires is presented. Starting from the definition of sensors inside the tire, it is shown how a predictive model can be used for the learning phase through the generation of virtual datasets. Following the learning phase, the architecture and the functional scheme of the machine learning algorithm is presented. The methodology is capable to estimate key variables of the tire operation starting from common measurements on the vehicle
A Nonlinear Estimation Approach for Vehicle and Tire-Road Monitoring with No Interaction Modelling
No full textThe reliability of control systems implemented onboard vehicles is strictly dependent on the correct estimation of the sideslip angle coupled with tire-road features. The prediction of tire-road forces and friction is strongly dependent on tire models typically characterizable through expensive experimental tests. In this paper, a nonlinear estimation approach, based on an Extended Kalman Filter, of lateral tire-road forces and friction coupled with the vehicle sideslip angle is proposed. The estimator is based on a single-track vehicle model, including a parametric estimation approach to avoid a specific tire model employment. The results obtained through the proposed technique are compared with a detailed vehicle model in both non-noisy and noisy measurements
AN ESTIMATOR BASED ON A SIGMA-POINTS KALMAN FILTER FOR VEHICLE AND TIRE-ROAD CONDITION MONITORING
No full textIn the automotive field, the estimation of key variables related to vehicle dynamics, such as the sideslip angle coupled with characteristic tire-road parameters, is fundamental for improving the safety performance of vehicle control systems. Typically, expensive tests are made to characterize tire parameters included in models functional for estimating tire-road forces. A nonlinear model-based estimator, based on the Central Difference Kalman Filter, is proposed in this work for making the vehicle condition monitoring through the coupled estimation of the sideslip angle, the tire-road friction coefficient and lateral forces. An estimator design model is designed around the single-track vehicle model coupled with a parametric estimation strategy functional for estimating tire-road features without specific modelling of tires. The proposed technique is assessed by comparing the estimated data with the ones obtained from a Multibody vehicle model. Two different tests are made considering both non-noisy and noisy measurements
Synthesis and comparative analysis of three model-based observers for normal load and friction estimation in intelligent tyre concepts
No full textRecently, the growing trends towards autonomous driving and full automation of terrestrial vehicles have intensified the need for intelligent and interconnected systems to collect and communicate reliable data in real-time. Since the tyres represent the primary sensing system between the vehicle and the ground, it comes natural to designate them to acquire information about their interaction with the road. Intelligent or smart tyre technologies can, indeed, be used to estimate both vehicle’s performance and environmental conditions, leading to handling, NVH and comfort improvements. Inspired by some encouraging results found in previous works, in this investigation we present three model-based estimators to detect the forces acting in the contact patch for an intelligent tyre. The underlying mathematical foundation is the Flexible Ring Tyre Model (FRM), which is able to describe the in-plane dynamics of intelligent tyre systems. More specifically, it represents the tyre treadband by means of a flexible ring restrained at its mean radius by a viscoelastic foundation. Due to its relatively simplicity, the FRTM allows to obtain a closed-form solution for the treadband displacements, accelerations and circumferential strain, whilst being accurate enough to capture all the relevant phenomena concerning the tyre dynamics. The governing Partial Differential Equations (PDEs) of the system are reduced to a set of coupled Ordinary Differential Equations (ODEs) by performing the Fourier series expansion. An optimal observer is then designed based on the Unscented Kalman Filter to account for the nonlinearities which arise when the vertical load acting on the tyre and the friction coefficient are included in the augmented state space representation. The three different technologies considered in this study are: strain, displacement, and acceleration-based intelligent tyres. All of them are shown to be capable of estimating the quantities of interest – the actual state of the system, the normal force and the adhesion coefficient – with high accuracy in very short times. Furthermore, the robustness of the proposed approach is validated by considering a step variation in both the vertical load and the available friction. The performances of the observers are finally compared by using the RSS and MSE indices
Nonlinear estimation of the Bouc-Wen model with parameter boundaries: Application to seismic isolators
No full textThe Bouc-Wen model is one of the most widely used parametric models of hysteresis. The exactness of the Bouc-Wen model depends on the correctness of the model parameters that are subjected to bound constraints in accordance with their physical meaning. Constraints and nonlinearities of the model introduce a complexity for the parameter estimation. In this paper, a constrained unscented Kalman filter (CUKF) is proposed in order to identify the hysteresis model parameters in a robust and reliable way. The adopted parameter estimator has been compared with the well-known extended Kalman filter (EKF) that is often used for nonlinear system identification. The effectiveness of the proposed approach has been verified by numerical simulations and experimental tests. In particular, for the experimental verification of the CUKF, the Bouc-Wen model has been adopted to describe the hysteretic shear behaviour of a seismic isolator. The results show that the CUKF provides better parameter identification than the EKF also taking into account parameter boundaries
Design and experimental validation of an adaptive fast-finite-time observer on uncertain electro-hydraulic systems
Full text linkThis paper presents the design of an adaptive fast-finite-time extended state observer for electro-hydraulic
actuator systems. First, the system model is divided into three parts, and fast-finite-time state observers are
designed independently for each part. This guarantees the fast-finite-time uniform ultimate boundedness of the estimation errors. Then, based on the designed state observers and without neither any knowledge about the upper bounds of the uncertainties nor their derivative, supplementary observers are presented to estimate the unknown terms. Rigorous analyses of the proposed strategy are provided through the Lyapunov approach.
The suggested adaptive framework can improve the convergence rate for zones both far from the equilibrium points and around them. The adaptive gains are computed based on the straightforward evaluation of the absolute value of the observation errors, thus their values are valid in real life applications, achieving finite-time estimate of both the full state variables as well as uncertainties. Comparative simulations are presented to analyze the effectiveness of the proposed observers with and without unknown measurement noise. Finally, the effectiveness of the proposed approach in real-life conditions is demonstrated through experimental studies
- …
