Road profile and suspension state estimation boosted with vehicle dynamics conjectures

Vazquez, Angel Gabriel Alatorre; Vaseur, Cyrano; Victorino, Alessandro Corrêa; Charara, Ali

doi:10.1109/ivs.2019.8814242

Cited by 5 publications

(4 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Out of these forces, relative suspension velocity is derived. Vazquez et al [24] provided suspension state, road profile and transfer load estimation methodology using deflection sensors, accelerometers and gyrometers. The observation scheme used is a linear Kalman filter.…”

Section: Introductionmentioning

confidence: 99%

Feasibility of Neural Network Based Virtual Sensor for Vehicle Unsprung Mass Relative Velocity Estimation

Šabanovič¹,

Kojis²,

Šukevičius³

et al. 2021

Preprint

View full text Add to dashboard Cite

With the automotive industry moving towards automated driving, sensing is becoming an increasingly important part of enabling technology. The virtual sensors allow data fusion from various vehicle sensors and provide a prediction for measurement that is hard or too expensive to measure in another way or in the case of demand on continuous detection. In this paper, virtual sensing is discussed for the case of vehicle suspension control, where information about the relative velocity of the unsprung mass for each vehicle corner is required. The corresponding goal can be identified as a regression task with multi-input sequence input. The hypothesis is that the state-of-art method of Bidirectional Long-Short Term Memory (BiLSMT) can solve it. In this paper, a virtual sensor has been proposed and developed by training a neural network model. The simulations have been performed using an experimentally validated full vehicle model in IPG Carmaker. Simulations provided the reference data which was used for Neural Network (NN) training. The extensive dataset covering 26 scenarios has been used to obtain training, validation and testing data. The Bayesian Search was used to select the best neural network structure using root mean square error as a metric. The best network is made of 167 BiLSTM, 256 fully connected hidden units and 4 output units. Error histograms and spectral analysis of the predicted signal compared to the reference signal are presented. The results demonstrate the good applicability of neural network-based virtual sensors for the estimation of vehicle unsprung mass relative velocity.

show abstract

Section: Introductionmentioning

confidence: 99%

Feasibility of Neural Network Based Virtual Sensor for Vehicle Unsprung Mass Relative Velocity Estimation

Šabanovič¹,

Kojis²,

Šukevičius³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The interested reader is referred to Ray [20], Antonov, Fehn [21], Pence, Fathy [22]a n dW e n z e l ,B u r n h a m [ 23] for some examples. Vazquez, Vaseur [24] developed a road profile and suspension state estimator with a Kalman filter technique. The technique was validated with a high fidelity simulator on sinusoidal road profiles at various speeds.…”

Section: Existing Approaches In the Literaturementioning

confidence: 99%

“…The kinematic observer model used for Step 1 of the suspension force prediction algorithm is given in Equations ( 19) to (24). The seven degrees of freedom in the modified model are the four unsprung mass vertical displacements, the sprung mass vertical displacement and roll and pitch motions.…”

Section: Modified Observer and Predictor Modelsmentioning

confidence: 99%

Vehicle suspension force and road profile prediction on undulating roads

Hamersma

Schalk

2020

Vehicle System Dynamics

View full text Add to dashboard Cite

The aim of the proposed estimator and predictor is to predict the effect a change in suspension settings will have on the vehicle's dynamic response before the change in suspension settings is made. The efficacy of advanced driver assist systems (ADAS) such as ABS, traction control, torque vectoring and others deteriorate significantly on undulating roads [1,2]. By changing or modulating the suspension settings, the performance of these systems may be improved. The control system thus needs to make a decision regarding spring and damper settings, and the proposed estimator and predictor is capable of giving information to the controller before it makes its control decision. BackgroundThe use of controllable suspensions is prevalent throughout the vehicle industry. Semiactive suspensions are used on a wide array of vehicles, from heavy vehicle applications, such as military, mining and agricultural vehicles to high performance passenger vehicles and Sports Utility Vehicles (SUVs). The advantage of using a controllable suspension on a

show abstract

“…Out of these forces, relative suspension velocity is derived. Vazquez et al [28] provided a suspension state, road profile and transfer load estimation methodology using deflection sensors, accelerometers and gyrometers. The observation scheme used is a linear Kalman filter.…”

Section: Introductionmentioning

confidence: 99%

Feasibility of a Neural Network-Based Virtual Sensor for Vehicle Unsprung Mass Relative Velocity Estimation

Šabanovič

Kojis

Šukevičius

et al. 2021

Sensors

View full text Add to dashboard Cite

With the automotive industry moving towards automated driving, sensing is increasingly important in enabling technology. The virtual sensors allow data fusion from various vehicle sensors and provide a prediction for measurement that is hard or too expensive to measure in another way or in the case of demand on continuous detection. In this paper, virtual sensing is discussed for the case of vehicle suspension control, where information about the relative velocity of the unsprung mass for each vehicle corner is required. The corresponding goal can be identified as a regression task with multi-input sequence input. The hypothesis is that the state-of-art method of Bidirectional Long–Short Term Memory (BiLSTM) can solve it. In this paper, a virtual sensor has been proposed and developed by training a neural network model. The simulations have been performed using an experimentally validated full vehicle model in IPG Carmaker. Simulations provided the reference data which were used for Neural Network (NN) training. The extensive dataset covering 26 scenarios has been used to obtain training, validation and testing data. The Bayesian Search was used to select the best neural network structure using root mean square error as a metric. The best network is made of 167 BiLSTM, 256 fully connected hidden units and 4 output units. Error histograms and spectral analysis of the predicted signal compared to the reference signal are presented. The results demonstrate the good applicability of neural network-based virtual sensors to estimate vehicle unsprung mass relative velocity.

show abstract

Road profile and suspension state estimation boosted with vehicle dynamics conjectures

Cited by 5 publications

References 15 publications

Feasibility of Neural Network Based Virtual Sensor for Vehicle Unsprung Mass Relative Velocity Estimation

Feasibility of Neural Network Based Virtual Sensor for Vehicle Unsprung Mass Relative Velocity Estimation

Vehicle suspension force and road profile prediction on undulating roads

Feasibility of a Neural Network-Based Virtual Sensor for Vehicle Unsprung Mass Relative Velocity Estimation

Contact Info

Product

Resources

About