Promoting excellence in mobility engineering

  1. FISITA Store
  2. Technical Papers

Subjective Ride Comfort Prediction of Different Drive Modes of Converted Electric Vehicle
FISITA2014/F2014-AHF-004

Authors

Lerspalungsanti, Sarawut - National Metal and Materials Technology Center (MTEC), National Science and Technology Development Agency (NSTDA), Thailand

Abstract

One of main tasks of electric vehicle developers is to design the motor controller, which can sustain passenger demand of ride comfort, performance, and safety. To reduce the number of time-consuming drive tests required for the motor controller program validation, the prediction of the ride comfort using human models is proposed in this study. The objective is to apply the modeling tools developed by the Institute of Product Engineering (IPEK), Karlsruhe Institute of Technology, to predict the subjective human ride comfort of the electric vehicle, which is converted from an internal combustion engine driven vehicle. Two driving modes of motor controller are investigated. The human modeling is carried out by determination of the nonlinear correlation between the subjective comfort evaluation and the objective parameters. The subjective data are obtained from test drives consisting of start-up, acceleration and braking situation on the basis of the 5-digit scale. During the experimental investigation, predefined objective parameters are captured, such as the resulting longitudinal, lateral and vertical acceleration measured at driver seat, the actual vehicle speed and the standardized courses of the acceleration as well as the brake pedal. Correspondingly to the way a driver makes his evaluation, the Artificial Neural Networks (ANNs) are applied to interconnect the subjective evaluation with the significant objective parameters by “trained” weighted network connections. The validity of the elaborated human model is investigated by determination of the prediction accuracy. The results of the investigation have demonstrated that the subjective evaluations performed by different drivers were efficiently correlated with the objective values. By applying further objective data from drive modes, such as normal mode or ECO mode, the prediction accuracy of more than 90% is attained. Due to the achieved prediction performance, the presented approach can be efficiently applied to support the motor controller development. The following modification on the drive train assemblies might be carried out in an early stage of development process.

KEYWORDS – Electric vehicle, Comfort evaluation, Objectification, Human sensation modeling, Artificial Neural Networks

Add to basket