Highway safety and vehicle performance are two important considerations in the design of a heavy vehicle combination. In this paper a performance measure called the 'rearward amplication ratio' ( RWA) is used as a control criterion in the design of a vehicle-handling controller. This approach is diVerent to conventional control techniques. The RWA is de ned as the ratio of the peak lateral acceleration at the rearmost trailer's centre of gravity (CG ) to that of the lead unit during a lane-change manoeuvre. The vehicle under consideration is a six-axle truck/full trailer, which usually exhibits a high level of RWA leading to roll-over during obstacle avoidance manoeuvres.In this study, several control strategies are examined, namely active yaw control at the truck CG, active yaw control at the dolly CG and active yaw control at the trailer CG. These could be employed individually or in combination. The eVect of an active control torque applied to various vehicle units is examined by using an optimal linear quadratic regulator approach combined with a simpli ed four degrees-of-freedom linear vehicle model. The controller performance index parameters are determined for the vehicle based on acceptable RWA target values. The sensitivity of the controller to tyre cornering stiVness variation is further evaluated.Simulation results indicate that the RWA can be reduced without signi cant change of the uncontrolled vehicle trajectory when active yaw torque is applied to the dolly. The controller can be more eVective in improving the dynamic performance and roll stability of this type of commercial vehicle, if applied to the lead unit (truck) or to the last unit (trailer). However, the path of the vehicle will be strongly in uenced and driving diYculties can be experienced. For active yaw control at the dolly CG, the optimal controller is found to be most sensitive to the dolly's tyres' cornering stiVness variations and least sensitive to steering axle from the RWA point of view. It is also found that the controller is most sensitive to steering axle parameter variations for path following.

Researchers developed a fuzzy-logic model for predicting the risk of accidents that occur on wet pavements. Preventing wet-pavement accidents has been an extremely difficult and elusive task because they are stochastic events whose occurrence is related to a variety of factors, including vehicle, roadway, human, and environmental characteristics. Conventionally, researchers use linear or nonlinear regression models and probabilistic models to predict wet-pavement accidents. However, these models often are limited in their capability to fully explain the process when the underlying physical system possesses a degree of non-linearity. Therefore, the potential of applying fuzzy logic in this area might be promising. Two fuzzy-logic models were developed and evaluated using accident data and the corresponding traffic data collected from 123 sections of highway in Pennsylvania from 1984 to 1986. The models use skid number, posted speed, average daily traffic, percentage of wet time, and driving difficulty as input variables and the number of wet-pavement accidents as the output variable. The first model is based on Mamdani’s fuzzy-inference method, and the second is a Sugeno-type fuzzy-logic model using the fuzzy-clustering method. The two fuzzy-logic models show superiority over the probabilistic model and the nonlinear regression model. Results indicate that, in addition to predicting the risk of wet-pavement accidents, the fuzzy-logic model can be applied conveniently to determine specific corrective actions that should be undertaken to improve safety.

No abstract