Review Article

Are Self-Driving Vehicles Ready to Launch? An Insight into Steering Control in Autonomous Self-Driving Vehicles

Table 1

Geometric-based steering control in self-driving vehicles.

ABCDEFG

Chen et al. [24]Pure pursuitProposed the PP-PI controller based on a combination of pure pursuit and proportional-integralSteering angle, radius, look-ahead error, lateral/heading error, PID parametersSimulationsEffectively reduced the dependence on the look-ahead, achieved smaller tracking error and better steering smoothnessSpeed’s effects on the tracking have not been discussed and vehicle dynamics are ignored
Park and Han [25]Presented adaptive pure pursuit algorithm-based steering controller using PI control theoryCornering stiffness, kinematic parameters, lad, steering angle, lateral offset angle, PID parametersAdaptive pure pursuit reduces the path tracking error to 4.0 more than conventional pure pursuitNeglected the effect of throttle brake controller
Andersen et al. [32]Proposed the steering control law for the alternative formulation to the pure pursuit path tracking algorithm.Kinematic parameter, steering angle, root mean square RMS errors, cross-track errorPracticalRMS cross-track error reduced up to 46%, and overcame corner-cutting and overshoot problemThe theoretical derivation of the performance limitations of the controller has not been presented
Shan et al. [36]Presented CF-pursuit-based fuzzy system to measure the steering angle and adjust the look-ahead parameter to control the motion of the AVCurve angle, look-ahead distance, radius, steering angle, vehicle heading, kinematic parametersMaintained stability in steering controller, achieved less cross-track errorVelocity has not been considered that badly influenced the performance of the steering controller
Hoff. et al. [37]StanleyPresented nonlinear steering control law based on the Stanley algorithm for AV to track a trajectory on off-road terrainKinematic parameters, steering angle, steering servo meter, stiffnessAchieved root mean square cross-track 0.1 m less than the standard error = 0.08Overshoot and the steady-state error occurred at high speed
Wit et al. [38]Vector pursuitPresented vector tracking method based on screw theory to deal with a vehicle headingUnit/moment vector, kinematic parameters, translation/rotation/instantaneous screw, look-ahead pointAccurately handle the large error, less reliance on lad, more robust than stanley and pursuit methodThe complex calculation required expertise to tune parameters
Yeul et al. [39]Proposed geometric control approach based on vector pursuit technique to improve the stability in steering controlTranslation/rotation/instantaneous screw, weighting factors, kinematic parameters, look-ahead point, slid slip, radiusSimulationAchieved less changing rate of heading error, the average slip of right and left track has been recorded as 0.0378 and 0.0383, respectivelyHas not been verified in a real or simulated environment

A: authors, B: technique, C: contribution, D: considered parameters, E: testbed, F: strengths, and G: limitations.