State Estimation with Coarsely Quantized, High-Data-Rate Measurements

Abstract

This paper treats the problem of estimating a signal corrupted by noise that is sampled and quantized at a high data rate. Local and global processors are proposed to achieve data compression that permits near optimal extraction of information. Two techniquesmaximum likelihood and minimum transform chi square, which are in the class of best asymptotically normal estimators-are investigated for the local processor. Simulation results are presented to demonstrate the feasibility of the approach.

A New Motion Planning Framework based on the Quantized LQR Method for Autonomous Robots

“…Now, our optimization problem can be expressed as follows. We are trying to find the coarsest quantizer that satisfies the control boundaries [u (1) k , u (2) k ] and decreases the performance index in every step( i.e. ∀k∆V (x k ) < 0).…”

“…The discontinuity in the system can be based on the control system transitions like the operation of gearbox shift pattern in vehicles or the steep system dynamics like a change of direction or final stop of a bouncing ball. In the early studies of control theory, this well-known phenomenon in the system dynamics interpreted as a predictable disturbance [1] or the noise on the signal [2]. After the widespread usage of the pulse width modulation signals in electric drives at the beginning of the 90s, sliding mode control techniques [3,4] use this nonlinear switching behavior as a control method for stabilizing nonlinear systems with an on-off(bang-bang) controller.…”