Human-In-the-Loop Evaluation of Advanced Semi-Active Primary Suspension Control Systems for HMMWV

Principal Investigators
Steve C. Southward, Mehdi Ahmadian (Virginia Tech)

Government
Alexander Reid (TARDEC)

Student
Daniel Reader (Virginia Tech)

The primary objectives of this effort are:

1. Integrate empirical nonlinear dynamic magneto-rheologic (MR) damper models (Figure 1) with semi-active control laws into the primary suspension of a full vehicle simulation model for a High Mobility Multipurpose Wheeled Vehicle ( HMMWV) or equivalent vehicle.
2. Implement the integrated closed-loop vehicle model on a driving simulator (Figure 2) for human-in-the-loop evaluation of ride, handling, and rollover performance.

The 2007 ARC project on Advanced Semi-Active Control Methods for HMMWV Primary Suspensions established a method for obtaining empirical nonlinear dynamic models of MR dampers that more accurately represent the dynamic response of the devices due to realistic current command inputs as well as relative velocity inputs induced by the dynamic system. The method was shown to be useful for estimating not only a forward model (predict force output due to current and relative velocity input), but also an inverse model (predict what current is required to achieve a desired force output for a given relative velocity input). These empirical models were then exploited to improve the performance of a semi-active primary suspension control system for a quarter-vehicle as indicated in Figure 1. The models can also be used for feedforward linearization of the MR damper.

Figure 1. Prior integration of nonlinear dynamic inverse MR damper model into a semi-active primary suspension control system

The next phase of this development, and the primary focus of this project, is to incorporate these empirically derived nonlinear dynamic MR damper models with semi-active control laws into a dynamic vehicle model of the HMMWV (or equivalent) vehicle for human-in-the-loop evaluation of the closed-loop performance. This will be accomplished using a 6-DOF full motion platform driving simulator which is shown in Figure 2. This platform will be used to enable a human driver to test drive a virtual vehicle on a virtual terrain in order to subjectively evaluate the ride and handling performance of the suspension control system with the virtual MR damper. Semi-active control laws such as the hybrid control algorithm will be evaluated for their ability to selectively optimize performance for ride and handling depending on the road input and/or maneuver. In order to accomplish this human-in-the-loop (HIL) testing, a real-time vehicle dynamics model with the integrated MR damper models and the primary suspension control system must be implemented in the external physics engine of the driving simulator hardware and software .

Figure 2. 6-DOF Full Motion Driving Simulator

Semi-active suspension control solutions represent an obvious choice for improving vehicle performance because they are lighter, less expensive, more reliable, and offer near equivalent performance compared to fully-active solutions. Significant short-term improvements in the suspension performance can be demonstrated using the evaluation platform developed as part of this research. The technology deliverables from this research are broadly applicable to any system using a controllable damper such as an MR device. Potential industrial application areas include semi-active suspension systems on commercial and personal vehicles, seat suspension systems, and so-called “programmable” dampers used in 4-post testing and the automotive and motorsports industries. The integration of the nonlinear dynamic MR damper models into the virtual primary suspension of a real-time vehicle dynamic model can be used for simulation studies of novel future control methods for improving ride, handling, and rollover mitigation. The use of this integrated model in a 6-DOF driving simulator enables human-in-the-loop evaluation of the closed-loop suspension control system. The ability to do human-in-the-loop evaluation of closed-loop suspension performance represents a degree of assessment that is not possible through simulation alone, and has the potential to reduce the cost of evaluating future technologies through virtual prototyping.