Force Sensors for the Active Safety of Road Vehicles

Abstract

Force and moment measurement within road vehicles plays a break-through role in automotive engineering. Both wheel force transducers and instrumented hub carriers are considered in the paper. Both technologies have advantages and disadvantages. Active safety systems (ABS, ESP, up to full automated driving) are expected to be impacted by the measurement of forces and moments at the wheels. Friction potential evaluation and driver model development and monitoring are major field of research. Force and moment measurement technology may also be exploited for lightweight construction purposes. Promising technologies are the ones that don’t need RF data transfer, providing low latency for data transfer and are resiliency against cyber-attacks.

1 Introduction

Most of the algorithms used in Vehicle Dynamics Control are based on on-line measurements of the state of the vehicle. Some measurements are directly taken, such as wheels angular velocities [21]. The measure of the force can be provided by components that measure the forces exchanged at the tyres. A state-of the-art review on the topic is provided in [12]. In [3, 6, 8, 9], the development of an instrumented wheel to measure the forces acting at the wheel hub is presented. The measure is obtained through a patented six-axis load cell [11]. The same technology can be adapted and used by resorting to an instrumented hub carrier [13, 20]. Differently to the RF data transfer implemented on the wheel, the hub carrier permits to use a wired measurement system that leads to a lower latency. The component includes a three-spokes structure. The spokes connect the hub case to the external case. To make the structure statically determined, the spokes have an additional degree of freedom along their own axis. On the spokes surface a system of strain gauges detect the deformation of the material that is converted into force measurement. The moment around the wheel axis cannot be detected since it is the degree of freedom of the wheel respect to the suspension. An instrumented brake caliper can be fitted in order to measure the brake torque. Active safety systems (ABS, ESP, up to full automated driving) are impacted by the measurement of forces and moments at the wheels [1, 8, 12, 13, 21]. New strategies for the control of road vehicles dynamics based on the measurements of the forces acting on the tyres can be implemented

2 Method

Several control algorithm can be applied to ABS system and the main ones are analyzed in [14]. Rule-based ABS are robust to disturbances coming from changes in the road friction coefficient. This is due to the fact that these systems work with the level of acceleration and deceleration of the wheel. As these levels depend only on the difference between the brake and road torque, and not their actual magnitudes, road friction coefficient has little influence on the performance of the system [21]. On the other hand, if the ABS control algorithm is supposed to work based on tyre force measurements (as in this case), the algorithm must be robust enough to guarantee an equivalent performance in the presence of disturbances, model inaccuracies and so on. Under these assumptions, Sliding Mode Control (SMC) seems to meet all the requirements for this particular application. Force based Anti-Lock Braking Systems (ABS) can be developed by using Sliding Mode Control. In literature it is possible to see many example of its application [1, 2, 4, 7, 15,16,17,18]. A complex four-wheels multibody model (14 degrees of freedom) has been used to simulate the force based implementation of the ABS logic in comparison with the standard one. The model represents an high-performance sportscar for motorsport applications. Each wheel has an independent ABS control respect to the other wheels. The control computes the optimal brake torque that is applied to the wheel. To the brake torque is applied a filter that modifies the signal with a time delay, a first order delay, a 100 Hz signal resampling and a random noise addition. Two straight braking events have been selected to compare the two ABSs. The first is a 100 km/h braking with high friction coefficient ($\mu = 1.12$) to simulate a dry road. The second is a mixed surface (high $\mu = 1.4$ and low $\mu = 0.7$) $\mu$-split road. Every 4 m, the high $\mu$ and low $\mu$ sections are inverted giving to the road the shape of a chessboard. The initial speed for this maneuver is 90 km/h.

Fig. 1.

Race motorcycle smart wheel (a). Instrumented wheel carrier featuring a six-axis load cell (b).

2.1 Rule-Based ABS Control Logic

Rule-based ABS control logic is based on thresholds on the wheel rotational speed and wheel slip. The check on the variables is done sequentially: firstly the brake torque is adjusted to maintain the wheel acceleration inside the predefined thresholds. If the acceleration is acceptable, the brake torque is adjusted to maintain the wheel slip inside the predefined thresholds.

2.2 Sliding Mode Control ABS

The SMC is based on the sliding surface defined as:

$$\begin{aligned} s = \kappa - \kappa _{des}+K_i\int {(\kappa - \kappa _{des})dt} \end{aligned}$$

(1)

where $\kappa$ is the slip defined as $\kappa =\frac{V-\varOmega R}{V}$, $\kappa _{des}$ is the desired slip target, $K_i$ is the coefficient of the integral term. The sliding surface was chosen based on the work conducted in [1, 2, 4, 10, 15, 18, 19].

From the sliding surface it is possible to obtain the value of braking torque to be applied to the wheel. The SMC calculates the optimal braking torque to keep the wheel from locking and to maintain the slip close to the desired slip target. The braking torque $T_b$ is defined as:

(2)

To have a complete mathematical derivation of Eq. 2, please refer to [1, 2, 4, 7, 10, 15, 18]. The presented work to exploit the capabilities of the SMC, it integrates an online tyre-road friction coefficient estimation. Based on the work conducted in [3, 20], the measured forces are used to estimate the friction coefficient. The information is then used to calculate the optimal slip target in order to maximize the braking force [5, 21].

Fig. 2.

Wheels angular speed(a) and wheel slips (b). Emergency braking maneuver on high-$\mu$ road ($\mu =1.12$).

3 Results

The results confirm the relevant improvement by using the direct measurement of the wheel forces respect to the actual ABS technology. The simulations have been performed on an emergency brake test in different scenario conditions. The dry braking scenario is representative of conditions where ABS should not be stressed heavily in maintaining the optimal braking force. Results are shown only for the left wheels of the vehicle for symmetry in the results. In both Fig. 2a and Fig. 2b it is possible to see how the rule-based ABS is prone to oscillation. Instead, it is possible to appreciate how the SMC ABS is more precise and less affected by them. figure 2b shows with the black dotted line, the online estimation of the optimal slip target. Once identified the optimal slip target, the SMC ABS is able to almost overlap the real slip to the target one. This lead to the force based ABS to stop the car in 38.9 m respect to the 41.2 m required by the rule-based ABS. While considering a mixed surface, the benefit is particularly relevant. The force based ABS formulation allows to stop the car in 36.7 m, while the rule-based requires 40.8 m. In Fig. 2b it is possible to see even better how the optimal slip target changes whenever the wheel passes from one type of asphalt to the other. On the other hand, the rule-based ABS struggle to keep the wheel from locking when the wheel runs on the low adherence block.

Fig. 3.

Wheels angular speed (a) and wheel slip (b). Emergency braking maneuver on a mixed surface road.