Vibration control of vehicle active suspension system using a new robust neural network control system

. All rights reserved. 2053; fax: +90 352 4375784. ). _ I. Eski, S . Yıldırım / Simulation Modelling Practice and Theory 17 (2009) 778–793 779 developed to control a seat suspension system of non-linear full vehicle model [10]. Yıldırım and Uzmay have been investigated the variation of vertical vibrations of vehicles using a neural network [11]. An active horizontal spray-boom suspension, reducing yawing and jolting, has been designed by Anthonis and Ramon [12]. A semi-active control of vehicle suspension system with magnetorheological (MR) damper has been presented by Yao et al. [13]. Spentzas and Kanarachos have been presented a methodology for the design of active/hybrid car suspension systems with the goal to maximize passenger comfort [14]. A methodology for the design of active car suspension systems has been presented [15]. Yagiz and Yüksek have been researched sliding mode control of active suspensions for a full vehicle model [16]. In this paper, a robust neural network based robust control system for whole vehicle’s vibration control is proposed. The paper first describes the full vehicle suspension model under consideration. Second, the proposed control system and standard PID controller are outlined in Section 3. Third, the results of proposed neural based control system and PID control system are given and discussed. Finally, the effectiveness of the proposed control method is concluded in Section 5. 2. Full vehicle model In order to describe the vertical dynamics of a road vehicle which runs at a constant speed along an uneven road, 7 degrees of freedom (DOF) mathematical vehicle model is used and is shown in Fig. 1. The model, including the relative displacement of unsprung masses and the front left suspension mass (z1 z1), the rear left suspension mass (z2 z2), the rear right suspension mass (z3 z3), the front right suspension mass (z4 z4) and the displacement of vertical motion of the vehicle body (z) and including the angular displacement of roll (u) and pitch (h) motion of the vehicle body. The vehicle parameters are given Table 1. According to dynamic analysis, the equations for the full vehicle model are given by: ct1ð _ z1 _ y1Þ 1⁄4 Ft1 ð1Þ ct2ð _ z2 _ y2Þ 1⁄4 Ft2 ð2Þ ct3ð _ z3 _ y3Þ 1⁄4 Ft3 ð3Þ ct4ð _ z4 _ y4Þ 1⁄4 Ft4 ð4Þ In these equations, ct1, ct2, ct3 and ct4 denote the damping coefficients of the left front tyre, the left rear tyre, the right rear tyre and the right front tyre, respectively. z1, z2, z3 and z4 represent the vertical displacement of the left front wheel-axle, the left