Hongzhong Zhu

Mechanical Deformation Analysis and High-Precision Control for Ball-Screw-Driven Stages

Ball-screw-driven stages are widely used in industry for long-range and high-precision fabrication. One of the main difficulties in dealing with high-precision motion control is the nonlinear friction, which originates at elastic deformation of the mechanical components. In this paper, first, a novel mechanical model of a ball-screw-driven stage is proposed to analyze the elastic deformation dynamics under various motion conditions. It is achieved that the mechanical deformation characteristic of fast motions differs from the case of slow motions in the zero-speed region, and therefore, performing friction compensation should take the motion conditions into account. Then, based on this observation, a sinc-function-based friction compensation method for slow motions including both reverse motions and nonreverse motions and a Sigmoid-function-based friction compensation method for fast reverse motions are proposed to improve the control performance. Thanks to the very few parameters, the proposed methods can save much trouble in design and maintenance and are particularly suitable for control purposes. Finally, the advantages of the proposed methods are evaluated by experiments.

Suppression of Current Quantization Effects for Precise Current Control of SPMSM Using Dithering Techniques and Kalman Filter

High-precision current control is demanded for many mechatronic systems to improve the static and dynamic performances. However, current measurement error inherent in the measurement system degrade the control accuracy, since the motor currents are not guaranteed to follow the desired references. In this paper, quantization error caused by analog-to-digital converters is studied. The dithering techniques combined with Kalman filter are proposed to suppress the quantization effects. First, two dithered systems, including the subtractively dithered and nonsubtractively dithered systems, are designed to whiten the quantization error. In the design of nonsubtractively dithered system, the probability characteristic of the metering noise is utilized to minimize the measurement noise level. Then, Kalman filter is designed to estimate the real current signals from the dithered current measurements. Since the variance of the total measurement error is theoretically calculable according to the proposed dithering techniques, Kalman gain can be determined analytically. Finally, simulations and experiments are performed to verify the effectiveness of the proposed approaches using a high-precision positioning system.

Proposal of nonlinear friction compensation approach for a ball-screw-driven stage in zero-speed region including non-velocity-reversal motion

This paper presents a heuristic friction compensation approach for handling the springlike friction behaviors of a ball-screw-driven stage in zero-speed region. The friction compensation for the motion that the stage is decelerated to stop and then accelerated in the same direction, which is also referred to as non-velocity-reversal motion, is discussed in detail. As the elastic energy stored in mechanical components may not be completely released during non-velocity-reversal motion, the conventional compensation approaches may not work well. Therefore, some other sophisticated friction compensation approaches become necessary for this kind of motion. In this study, a velocity pattern recognition algorithm is presented as the first step to classify the velocity patterns in zero-speed crossing region. Then, sinc function is exploited to model the nonlinear springlike friction of the ball-screw-driven stage. It is proved that the friction can be properly compensated for both the reversal motion and non-reversal motion. Experiments are performed to verify the proposed approach and it is demonstrated that the control performance is significantly improved.

Overcoming current quantization effects for precise current control by combining dithering techniques and kalman filter

Accurate current signals are required for precise current control of AC motor drives. However, current measurement error including the metering error and the quantization error is unavoidable due to the inaccuracy of current sensors and the quantizing feature of analog-to-digital converters. In this paper, the combination of dithering techniques and Kalman filter is presented to suppress the current quantization effects caused by A/D converters. Firstly, the dithering system is designed to whiten the total current measurement error. Then, Kalman filter is exploited to estimate the real current signals from the whitened noise. The effectiveness of the proposed approach is verified via simulations and experiments using a high-precision stage.

Proposal of fast and high-precision control for ball-screw-driven stage by explicitly considering elastic deformation

Ball-screw-driven stages are extensively used in industry for long-range and high-precision fabrication thanks to their high efficiency and high ability to apply/withstand high thrust loads. However, spring-like nonlinear friction in zero-speed region significantly affects the control performance. In this paper, the mechanical deformation characteristics at zero-speed point after low deceleration and after high deceleration are analyzed. It is found that the mechanical deformation characteristic of fast motions differs from the case of slow motions in zero-speed region, and therefore performing friction compensation should take the motion conditions into account. Then, after introducing a variable to evaluate the elastic deformation of the total mechanical components, a novel friction compensation method for fast reverse motions using Sigmoid function is proposed to enhance the control performance by explicitly considering the elastic deformation. Finally, Experiments are performed to verify the effectiveness of the proposed method.

Proposal of Position Reconstruction with Polynomial Fitting Approach for Precise Motion Control

Abstract Optical encoders are extensively used for position measurements in motion control systems. The inherent quantization feature of encoders limits the control performance in many applications. This paper shows a way to reconstruct the smooth position signal from the quantized measurements. The method uses a polynomial fitting approach where the degree of polynomial is automatically selected via a convex optimization method. In addition, the information of the plant is exploited by combining an observer. The effectiveness of the proposed method is verified by simulations and experiments using a high-precision linear stage.

Reconstruction of Plant Output from Quantized Measurement—A Moving Horizon Polynomial Fitting Approach

Abstract This paper presents a new approach to reconstruct the plant output of linear time-invariant systems in the case where the available output measurement is quantized. By fitting the quantized measurement data with polynomials in a moving horizon manner, a smooth approximation of plant output is obtained via solving a convex optimization problem. Applying the signal to an observer, the plant output is reconstructed by taking account of the plant dynamics. It is guaranteed that the error between the true and the reconstructed output is bounded. Experimental validation is given by using a DC motor positioning system. It turns out that the proposed approach achieves small reconstruction error and accurate tracking control. In addition, the approach yields a smooth reconstruction signal so that the plant input is well behaved (or smooth) even if PID controller is employed for the plant subject to quantized output measurement.