Atta, Debabrata (2017) Development of Robust Control Strategies for Autonomous Underwater Vehicles. PhD thesis.
The resources of the energy and chemical balance in the ocean sustain mankind in many ways. Therefore, ocean exploration is an essential task that is accomplished by deploying Underwater Vehicles. An Underwater Vehicle with autonomy feature for its navigation and control is called
Autonomous Underwater Vehicle (AUV). Among the task handled by an AUV, accurately positioning itself at a desired position with respect to the reference objects is called set-point control. Similarly, tracking of the reference trajectory is also another important task. Battery recharging of AUV, positioning with respect to underwater structure, cable, seabed, tracking of reference trajectory with desired accuracy and speed to avoid collision with the guiding vehicle in the last phase of docking are some significant applications where an AUV needs to perform
the above tasks. Parametric uncertainties in AUV dynamics and actuator torque limitation necessitate to design robust control algorithms to achieve motion control objectives in the face of uncertainties. Sliding Mode Controller (SMC), H / μ synthesis, model based PID group controllers are some of the robust controllers which have been applied to AUV. But SMC suffers from less efficient tuning of its switching gains due to model parameters and noisy estimated acceleration states appearing in its control law. In addition, demand of high control effort due to high frequency chattering is another drawback of SMC. Furthermore, real-time implementation of H / μ synthesis controller based on its stability study is restricted due to use of linearly approximated dynamic model of an AUV, which hinders achieving robustness. Moreover, model based PID group controllers suffer from implementation complexities and exhibit poor transient and steady-state performances under parametric uncertainties. On the other hand model free
Linear PID (LPID) has inherent problem of narrow convergence region, i.e.it can not ensure convergence of large initial error to zero. Additionally, it suffers from integrator-wind-up and subsequent saturation of actuator during the occurrence of large initial error. But LPID controller
has inherent capability to cope up with the uncertainties. In view of addressing the above said problem, this work proposes wind-up free Nonlinear PID with Bounded Integral (BI) and Bounded Derivative (BD) for set-point control and combination of continuous SMC with Nonlinear PID with BI and BD namely SM-N-PID with BI and BD for trajectory tracking.
Nonlinear functions are used for all P,I and D controllers (for both of set-point and tracking control) in addition to use of nonlinear tan hyperbolic function in SMC(for tracking only) such that torque demand from the controller can be kept within a limit. A direct Lyapunov analysis is pursued to prove stable motion of AUV. The efficacies of the proposed controllers are compared with other two controllers namely PD and N-PID without BI and BD for set-point control and PD plus Feedforward Compensation (FC) and SM-NPID without BI and BD for tracking control.
Multiple AUVs cooperatively performing a mission offers several advantages over a single AUV in a non-cooperative manner; such as reliability and increased work efficiency, etc. Bandwidth limitation in acoustic medium possess challenges in designing cooperative motion control
algorithm for multiple AUVs owing to the necessity of communication of sensors and actuator signals among AUVs. In literature, undirected graph based approach is used for control design under communication constraints and thus it is not suitable for large number of AUVs participating in a cooperative motion plan. Formation control is a popular cooperative motion control paradigm. This thesis models the formation as a minimally persistent directed graph and
proposes control schemes for maintaining the distance constraints during the course of motion of entire formation. For formation control each AUV uses Sliding Mode Nonlinear PID controller with Bounded Integrator and Bounded Derivative. Direct Lyapunov stability analysis in the
framework of input-to-state stability ensures the stable motion of formation while maintaining the desired distance constraints among the AUVs.
|Item Type:||Thesis (PhD)|
|Uncontrolled Keywords:||AUV; Parametric uncertainties; Actuator saturation; Nonlinear PID; Integrator wind-up; Sliding Mode Control; Cycle Free Minimally Persistent Formation; Cascaded system|
|Subjects:||Engineering and Technology > Electrical Engineering > Power Electronics|
|Divisions:||Engineering and Technology > Department of Electrical Engineering|
|Deposited By:||Mr. Kshirod Das|
|Deposited On:||05 Oct 2017 12:21|
|Last Modified:||04 Dec 2019 16:57|
Repository Staff Only: item control page