A Haptic Gearshift Interface for Cars

Transcription

1 A Haptic Gearshift Interface for Cars Eloísa García-Canseco 1, Alain Ayemlong-Fokem 1, Alex Serrarens 2, and Maarten Steinbuch 1 1 Eindhoven University of Technology, Faculty of Mechanical Engineering, Control Systems Technology Group, PO Box 513, 5600 MB Eindhoven, The Netherlands 2 Drivetrain Innovations BV, Croy 46, 5653 LD Eindhoven, The Netherlands {e.garcia.canseco,m.steinbuch}@tue.nl, a.ayemlong.fokem@student.tue.nl, serrarens@dtinnovations.nl Abstract. This paper presents a two degrees of freedom haptic interface that uses force control to reproduce the behavior of a customary lever and gearshift in automotive applications. The haptic simulation of the gear selector lever has been done by the appropriated design of virtual artificial potential functions. These functions contain parameters that have intuitive physical meaning and that can be easily adjusted to change the force sensations fed to the user. To validate our approach, experiments have been carried out. Keywords: haptic device, gearshift, automobile. 1 Introduction The automotive industry is experiencing not only a strong enhancement of automation in primary driving tasks, e.g., automated gear shifting, lane-keeping systems, adaptive cruise control, park-assists, automatic hill-hold, brake-by-wire, etc, but also, the introduction of more auxiliaries and interior functions, e.g., USB-connectors, mp3 players, navigation, among others. These trends lower the driver s workload significantly, and draw the driver s attention more and more to the interior functions of the car. Adding more functionality to vehicles increases driver satisfaction or pleasure, however, it can also lead to a significant increase of driver distraction. Haptic cues might offer a promising and relatively unexplored alternative to give warnings and other messages to the driver, and also to aid drivers in the execution of their driving tasks. The term haptic comes from the Greek word haptesthai, which refers to the sense of touch. Accordingly, haptic interfaces are devices that provide humans with the means to act on their environment, by generating mechanical signals that stimulate the human touch senses [1]. In the recent years, the automotive industry has been taking a keen interest in haptic [2]. Car makers such as BMW, Audi, Lexus, Nissan and many more have already installed haptic interfaces in their automobiles. Haptic devices in the car can be classified in two categories [3]: a) devices that are constantly in contact with the driver, for instance, the haptic steeringwheel[4,5,6,7,8,9],thenissan haptic gas pedal, and the A.M.L. Kappers et al. (Eds.): EuroHaptics 2010, Part II, LNCS 6192, pp , c Springer-Verlag Berlin Heidelberg 2010

2 316 E. García-Canseco et al. Fig. 1. (left) Haptic gearshift interface concept. (right) 2 dof haptic interface. haptic car seat. Some of these devices use vibration to give the haptic feedback and they can be used to give warnings to the driver; b) devices that the driver must actively touch. Examples are buttons, knobs, levers and tactile displays, such as the BMW idrive [10], the Immersion TouchSense [11]. See also [12,13,14,15] for other applications of haptic in automotive environments. In this paper we investigate a controlled haptic force feedback shift lever that can accurately reproduce the behavior of a customary gearshift during driving, and that might also be used to control interior and comfort functions in the car (Fig. 1). As stated in [12], the primary advantage of haptic levers over customary gear selectors is their ability to reproduce the behavior of manual, automatic or semi automatic transmissions by simply changing the virtual environment, i.e., the force profiles (haptic patterns in Fig. 2). 2 Description of the System The main parts of the system are illustrated in Fig. 2 and can be summarized as follows. 2.1 The Haptic Interface The haptic interface (Fig. 1 (right)) is a two degrees of freedom mechatronic device, whose working principle is based on the self locking property of a worm pair transmission [16]. Without actuation, it is therefore impossible to force the worm to rotate by applying a force to the lever mounted on the worm gear. That is, the operator experiences infinite stiffness of hitting a wall when there is no current fed to the electrical actuator. Vice versa, the actuator is able to move the worm gear and lever around their rotation axis rather easily. The transmission ratio from worm to worm gear can be very large, offering thus a good speed reduction and smaller work load on the actuators. One advantage of this worm pair transmission is the low power consumption, more feedback force requires less power and there is no consumption at all when the lever is kept at its rest position [17].

3 A Haptic Gearshift Interface for Cars dof Haptic Interface Reference trajectories Inner Control Loop Torque Joint positions and velocities Kinematics Transformations Positions and velocities in operational space Virtual Admittance Outer Control Loop Force error + - Operator force Desired force Haptic Patterns Human Operator Fig. 2. Control scheme for the haptic simulation of a gear selector lever The link between the hardware and software is provided by the hardware server and controller board that come along with the dspace c software. The DS1102 controller board provides the inputs and output ports. This controller board is connected via a 62 pin SUB-D connector cable to the hardware server, which is build as a standard PC card. The dspace c software contains the Controldesk testing environment, the RTLib1102 real time interface Simulink c library blockset and the Texas Instrument TMS320C ANSI C compiler. To implement model based control strategies on the system, the dynamical model of the haptic interface has been obtained in the Euler Lagrange framework (see for instance [18]) as M(q) q + C(q, q) q + g(q)+τ f = τ J T (q) f e, where q R 2 is the vector of joint displacements, q R 2 is the vector of joint velocities, M(q) R 2 2 is a positive definite inertia matrix, C(q, q) q R 2 is the vector of centripetal and Coriolis torques, g(q) R 2 is the vector of gravitational torques, τ R 2 is the vector of actuation torques, J R 3 2 is the Jacobian matrix and f e R 3 is the interaction force vector with the environment. The friction torque τ f R 2 is given by [ ( τ f = τ c +(τ s τ c )exp q δ) ] + b q tanh(λ q), (1) q s which corresponds to a modified static friction model with Stribeck velocity [19] (Fig. 3), where λ R and δ R 2 are adjusting parameters, and the constant vectors τ c,τ s, q s,b R 2 represent the coulomb friction coefficients, the static friction coefficients, the Stribeck velocities, and the velocity coefficients, respectively.

4 318 E. García-Canseco et al. **** Data Prediction **** Data Prediction τf1 [Nm] τf2 [Nm] q 1 [rad/s] q 2 [rad/s] Fig. 3. Friction empirical model versus friction predicted model τ f =[τ f1,τ f2 ] T 2.2 Force Patterns We have borrowed some inspiration from [20] to design the haptic patterns that mimic the behavior of a desired gearshift transmission. These force profiles are based on generalized sigmoid functions and can be written as [20] U d (p)= M N i=1 j=1 f sigi (φ j (p)), (2) where p =[x,y,z] T is the position vector of a point in space, φ j (p) is a surface function and 1 f sigi (φ j (p)) = 1 + exp( γφ j (p)) is the generalized sigmoid function. The value of γ depends on the application at hand [20]. In our case, the value of γ is related to the user s preference for force magnitude. drive y [cm] neutral x [cm] reverse Fig. 4. (left) Virtual potential energy function U d (x,y) that mimics an automatic gearshift. x and y represent the position of the lever in the horizontal plane xy. The desired force patterns are generated from the negative gradient of U d as f d = U d p. (right) Experimental results.

5 A Haptic Gearshift Interface for Cars 319 The two main advantages of this technique are on one hand, the smoothness of the force field, due to continuous first and higher order derivatives, and on the other hand, the easy adjustment of the magnitude and effective range of the potential field. To generate the desired attractive forces, the negative gradient of equation (2) is computed as f d (p)= U d p. (3) Fig. 4 (left) illustrates the artificial potential fields which have been designed to replicate the behavior of an automatic transmission. 2.3 Experimental Results As depicted in Fig. 2, we have implemented a hierarchical control scheme consisting of two loops: the inner control loop is a position controller with friction compensation based on inverse dynamics control [18]. This controller computes the input torques as a function of the joint positions and velocities. The outer control loop is a PID force controller with anti windup [21]. This force controller compute the necessary set points as a function of the measured human force and the force generated by the haptic pattern in order to provide a haptic feeling to the operator. Fig. 4 (right) shows the experimental results for the automatic gearshift. The picture has been drawn during 20 trials in which the user changes the gear many times randomly. We notice that the movements are constrained by the force patterns generated by eq. (3), see also Fig. 4 (left). 3 Conclusions and Future Work We have presented the preliminary results of a haptic interface that can be used to test different automobile gearshifts. Current work is ongoing to implement the force patterns of semi automatic and manual gearshift transmissions. In the same spirit of [13], we plan to develop a graphical user interface, were users can easily determine the behavior of the gearshift, and select the desired force profiles. Acknowledgments. This research is supported by the SenterNovem IOP Man Machine Interaction Project MMI07105 Control solutions for human in the loop user interfaces and by DTI Automotive Mechatronics BV. References [1] Hayward, V., Astley, O.R., Cruz-Hernández, M., Grant, D., de-la Torre, G.R.: Haptic interfaces and devices. Sensor Review 24(1), (2004) [2] Bigelow, S.J.: Haptics make it happen. Smart Computing (2004) [3] Hjelm, J.: Haptics in cars. In: Seminar Haptic Communication and Interaction in Mobile Contexts (2008) [4] van Erp, J.B.F., van Veen, H.A.H.C.: Vibro-tactile information presentation in automobiles. In: Proc. Eurohaptics 2001, Birmingham, UK, pp (2001)

6 320 E. García-Canseco et al. [5] Enriquez, M., Afonin, O., Yager, B., Maclean, K.: A pneumatic tactile alerting system for the driving environment. In: Proc. of the 2001 Workshop on Perceptive User Interfaces, pp. 1 7 (2001) [6] Ho, C., Tan, H.Z., Spence, C.: Using spatial vibrotactile cues to direct visual attention in driving scenes. Transportation Research Part F: Traffic Psychology and Behaviour 8(6), (2005) [7] Toffin, D., Reymond, G., Kemeny, A., Droulez, J.: Influence of steering wheel torque feedback in a dynamic driving simulator. In: Driving Simulation Conference North America, Dearborn, MI, USA (2003) [8] Mohellebi, H., Kheddar, A., Espie, S.: Adaptive haptic feedback steering wheel for driving simulators. IEEE Transactions on Vehicular Technology 58(4), (2009) [9] Steele, M., Gillespie, R.B.: Shared control between human and machine: Using a haptic steering wheel to aid in land vehicle guidance. Human Factors and Ergonomics Society Annual Meeting Proceedings 45, (2001) [10] Bernstein, A., Bader, B., Bengler, K., Künzner, H.: Visual-haptic interfaces in car design at BMW. In: Human Haptic Perception: Basics and Applications, pp (2008) [11] Immersion corp., [12] Bengoechea, E., Sánchez, E., Savall, J.: Optimal cost haptic devices for driving simulators. In: Redondo, M., et al. (eds.) Engineering the User Interface, pp Springer, Heidelberg (2009) [13] Gil, J., Díaz, I., Iturritxa, E., Prieto, B.: A haptic interface for automobile gearshift design and benchmark. In: Ferre, M. (ed.) EuroHaptics LNCS, vol. 5024, pp Springer, Heidelberg (2008) [14] Angerilli, M., Frisoli, A., Salsedo, F., Marcheschi, S., Bergamasco, M.: Haptic simulation of an automotive manual gearshift. In: Proceedings of 10th IEEE International Workshop on Robot and Human Interactive Communication, pp (2001) [15] Frisoli, A., Avizzano, C., Bergamasco, M.: Simulation of a manual gearshift with a 2-DOF force-feedback joystick. In: IEEE International Conference on Robotics and Automation, Proceedings 2001 ICRA, vol. 2, pp (2001) [16] Serrarrens, A.: Gear changing device for automotive applications. Patent AF16H5904FI (2005) [17] van Diepen, K.: Dynamic haptic control for a 1 dof shift by wire system. In: Confidential DCT , Eindhoven University of Technology, Eindhoven, The Netherlands (2008) [18] Spong, M., Vidyasagar, M.: Robot dynamics and control. Wiley, Chichester (1989) [19] Olsson, H., Astrom, K., de Wit, C.C., Gafvert, M., Lischinsky, P.: Friction models and friction compensation. European Journal of Control 4 (1998) [20] Ren, J., McIssaac, K.A., Patel, R.V., Peters, T.M.: A potential field model using generalized sigmoid functions. IEEE Transactions on Systems, Man and Cybernetics Part B: Cybernetics 37(2), (2007) [21] Astrom, K., Haaglund, T.: Advanced PID control. ISA The Instrumentation, Systems and Automation Society (2005)