Publication3

Faults in an air brake system affect the performance of Heavy Commercial Road Vehicles (HCRVs). One of the major faults in the air brake system is the “out-of-adjustment”of pushrod due to excessive brake wear, which may cause a significant yaw angle deviation from the current path and an increase in stopping distance. Hence, this paper aims to design a controller that would maintain the vehicle's directional stability under brake fault scenario. In order to design such a controller, knowledge of vehicle side slip angle is essential, but this is not a measurable quantity. Hence, an Artificial Neural Network (ANN) based estimation scheme for side slip angle prediction is also proposed. As an output regulator problem, Sliding Mode Control (SMC) was used for correction of yaw angle under brake fault scenario through appropriate steering angle input. This controller provided a percentage correction of yaw angle of 93.4 % and 99.8 % respectively for fully laden and fully unladen vehicle on a high friction tire road interface surface, and the same corresponding to a low friction tire road interface was 99.8 % and 98.9 % respectively.

Given the condition of mixed road traffic, few models exist that can predict the motion of a vehicle in it. Mixed road traffic can be defined as being both, lane indisciplined and heterogeneous. This study aims at developing a model that can analyze a given traffic condition, generate a safe trajectory and provide a control input to the vehicle to follow it. The paper explains the flow of the model, starting with traffic interaction, leader detection, and waypoint derivation. Post this, the trajectory is generated, the tracking errors are discussed and a controller is designed to navigate the trajectory by minimizing the discussed errors. Since the assumption of low speed is made, a kinematic model is used when generating feasibility criteria for the trajectory. Once the trajectory is determined to be feasible, a closed-loop Proportional Integral Derivative (PID) controller provides steering input to the vehicle to follow the trajectory. The controller tuning is performed using a dynamic bicycle model considering the error with respect to the trajectory. The trajectory generation model and the controller for trajectory following are implemented in independent simulator environments. The resulting output is a collision-free trajectory as followed by the subject vehicle (SV) to meet the generated waypoints which are based on the traffic scenario.

The paper is concerned with identifying transfer functions of individual input channels in minimal realization form of a Multi-Input Single Output (MISO) from the input-output data corrupted by the error in all the variables. Such a framework is commonly referred to as error-in-variables (EIV). A common approach in the existing methods for identification of MISO systems is to estimate a non-minimal order transfer function under a subset of simplistic assumptions like homoskedastic error variances, known order, and delay. In this work, we deal with the challenging problem of identifying order, delay in each input of minimal realization form separately while estimating the transfer functions. We also estimate the heteroskedastic noise variances in each of the multiple inputs and output variables. An automated approach for the identification of MISO systems of minimal realization form in the EIV framework is proposed. Numerical case studies are presented to illustrate the efficacy of the proposed algorithm in identifying the transfer function along with the order, delay, and noise variances.

Efficient rehabilitation for muscles/nerves using Functional Electrical Stimulation needs control energy to be delivered at exactly a desired point (gait parameter) in a gait cycle. Since both the stimulator and body muscles have a built in lag (to respond), it is important that the gait parameter be predicted ahead of time. Current work focuses mostly on computing gait parameters and triggering the electrical stimulator. In this work, we have used a second-order Box-Jenkins polynomial model in combination with a simple thresholding algorithm to forecast a region around one of the gait parameters - the Heel Strike Region (HSR) - three time steps ahead. Each set of the three forecasts were assigned a certain confidence level to indicate which of the three forecasts were in the HSR. From the results obtained, we can conclude that the HSR can be forecasted 30 milliseconds ahead of time with good accuracy.

Teleoperated robots are helpful in handling distant objects, and is being increasingly relied upon in nuclear centers, space travel, etc. Control of teleoperated systems are plagued by the trade-off between its two objectives, namely stability and transparency. This paper proposes novel controllers for the master and slave sub-systems of the teleoperation system, in order to simultaneously meet both these performance objectives. Passivity analysis is done to guarantee the stability of the system. The uniqueness of the proposed controllers lies in circumventing the problems of gain tuning. Both simulations and experiments are carried out on a 1-degree of freedom teleoperated system, and the performance objectives are observed to be satisfactorily met.