Design and implementation of LQG\LTR controller for a magnetic telemanipulation system-performance evaluation and energy saving

Abstract

This paper deals with designing a telemanipulation system (TMS) for microrobotics applications. The TMS uses magnetic levitation technology for the threedimensional (3-D) manipulation of a microrobot. The TMS is made up of two separate components: a magnetic drive unit and a microrobot. The magnetic drive unit is developed to generate the magnetic field for propelling the microrobot in an enclosed environment. The drive unit consists of electromagnets, a disc pole-piece for connecting the magnetic poles, and a yoke. To handle the 3-D high precision motion control of the microrobot, experimental magnetic field measurements coupled with numerical analysis were done to identify the dynamic model of levitation. This approach leads to the design of a linear quadratic gaussian (LQG) control system, based on the derived state-space model. Based on the PID controller performance, the LQG controller provides considerable improvement in transient response and cross coupling errors. The 3-D motion control capability of the LQG control method is verified experimentally, and it is demonstrated that the microrobot can be operated in the TMS workspace, vertical range of 30 mm and the horizontal range of 32 Â 32 mm 2 , with RMS error on the order of 10 lm in the vertical and 2:2 lm in the horizontal direction. In the vertical motion, the cross coupling error of the LQG controller is nine times smaller than that of the PID controller. A pre-magnetized pole-piece is proposed to compensate for gravity effect and reduces the system's energy consumption. This pole-piece provides 66% energy saving for the system's workspace operations.

Adaptive Controller and Observer for a Magnetic Microrobot

“…Since by definition F |K = F , system (15)- (16) or (15)- (18) coupled with (21) gives: e = LHue + FL(X,Ẑ,θ, u) − FL(X, Z,θ, u), (A. 36) with Hu given by (22), and mapping FL = (0, 1 L 2 F ). Since Hu is Hurwitz, there exists a matrix P symmetric positive definite such that: H t u P + P Hu = −I4 (A.37)…”

“…To deal with the nonlinearities of the magnetic field, the authors in [21] have used decoupling and linear parametrization to synthesize an optimal controller that minimizes a quadratic cost. To face the same kind of problem, a linear quadratic gaussian controller based on the linearized model has been addressed in [22]. Recently, a new approach referred to as magnetic resonance navigation has been proposed to steer and track in real-time endovascular magnetic devices in deep tissues to target areas of interest [23], [24].…”

Medical Technologies and Challenges of Robot-Assisted Minimally Invasive Intervention and Diagnostics

“…In the field of control of magnets more generally, telemanipulation of magnets using external electromagnets has been achieved with a linear quadratic gaussian controller, haptic feedback with virtual fixtures, and force sensing using hall effect sensors on a capsule [111,112,113].…”

Micro-Domain Force Estimation Using Hall-Effect Sensors for a Magnetic Microrobotic Station

“…More details of controlling the magnetic forces and torques have been reviewed in Ref. (5), (16), (17). The magnetic untethered microrobotic system (MUMS), as the heart of micromanipulation platform, is made up of two separated subsystems: a magnetic drive unit (MDU) and a microrobot.…”

“…The drive unit consists of six-pair of electromagnets, a disc pole-piece for connecting the magnetic poles, and a yoke. The effectiveness of this structure for the 3D position controlling of the microrobots was reported in (5), (16), (17) . The maximum magnetic field in z-direction produced by the MDU is in the range of 0.1-0.2 Tesla inside the workspace.…”