Journal Articles

Web guides are mechanisms that are used in roll-to-roll material processing machines to control the lateral position of the material while it is transported over rollers in processing machinery; the flexible, continuous materials are often referred to in the industry as “webs”. The process of controlling the lateral position of webs on rollers is referred to as “web guiding”. With the increasing need to transport and process different types of materials under different operating conditions within the same processing machinery, the existing fixed gain control algorithms do not provide adequate performance under these changing conditions and must be re-tuned often to provide satisfactory guiding performance. A controller that will adapt to parameter changes resulting from changes in the material properties, operating conditions, and size and type of guide mechanisms is desired. This paper discusses the design and implementation of model-based adaptive control strategies for web guides based on the dynamic models which describe the lateral behavior of webs. Extensive experiments are conducted for the designed adaptive control strategies for two types of web guides, namely remotely pivoted guide and offset pivot guide. Practical implementation issues and methods to increase robustness of the adaptive strategies are discussed. Experimental results from the adaptive control strategies will be compared with the existing fixed gain PI controller.

The objective of this work is to investigate on-line optimization-based coordination strategies for robot teams to efficiently accomplish a mission (eg, reach a set of assigned targets) while avoiding collisions. The multi-robot coordination problem is addressed by solving an on-line receding-horizon mixed-integer program to find some suitable inputs for the vehicles. Simulations results verify the feasibility of our approach.

In this paper, we integrate model predictive control (MPC) and mixed integer linear programming (MILP) into a hierarchical framework suitable for solving optimization problems involving robotic networks. A critical issue in MPC/MILP applications is that the underlying optimization problem must be solved on-line. This creates a time constraint which is hard to meet when the number of robots and the number of obstacles increase. To alleviate this difficulty, we develop strategies that significantly improve the efficiency of a hierarchical, decentralized optimization scheme. As an application is considered a case of target assignment problem in urban-like environments. Numerical simulations verify the scalability of the algorithm to the number of robots and complexity of the environment.

Since rotating machinery is used to transport flexible materials (commonly known as webs) on rollers, it is common to observe periodic oscillations in measured signals such as web tension and web transport velocity. These periodic oscillations are more prevalent in the presence of nonideal elements such as eccentric rollers and out-of-round material rolls. One of the critical needs in efficient transport of webs is to maintain web tension at a prescribed value. Tension regulation affects almost all key processes involved during web transport including printing, registration, lamination, winding, etc. Governing equations for web tension and transport velocity that can accurately predict measured behavior in the presence of nonideal rollers are beneficial in understanding web transport behavior under various dynamic conditions and the design of suitable web tension and speed control systems. The focus of this paper is on modeling the effect of eccentric rollers and out-of-round material rolls on web tension and web transport velocity. The new governing equations for web velocity on an eccentric roller and web tension in spans adjacent to the eccentric roller are presented and discussed; a web span is the free web between two consecutive rollers. To solve these governing equations, the location of the entry and exit point of the web on the eccentric roller as it rotates and the length of the web spans adjacent to the eccentric roller are required; a procedure for obtaining this information is described. To corroborate the models and the developed approach, data from experiments on a large experimental web platform are compared with data from model simulations, and a representative sample of the results are presented and discussed.

The design and development of a new fiber-optic sensor for measuring the velocity of a continuous material (also called a web) in material processing systems is described. The development of the proposed sensor is based on the dual beam laser Doppler velocimetry technique and the unique properties of different types of optical fibers. The developed sensor is capable of measuring the true web transport velocity as opposed to the existing methods which infer web transport velocity based on the roller angular speed. Since the sensor design utilizes fibers, signal processing can be performed away from the measurement area, and as a result the sensor can be used in harsh environments within the web processing line. The proposed sensor has been constructed and experiments have been conducted on an experimental web platform. The performance of the sensor is evaluated for a range of web velocities and different web materials. Sensor design, its construction, and a representative sample of the results are presented and discussed.

Analysis and minimization of interaction in multivariable systems employing decentralized controllers with application to roll-to-roll manufacturing systems are considered in this paper. A new interaction metric based on the Perron-Frobenius theory of nonnegative matrices is presented. This new metric may be used to quantify interaction in a large-scale interconnected system, establish constraints on closed-loop system stability, and provide a systematic design procedure for constructing decentralized pre-filters, which minimize interaction. The new interaction metric is applied to a roll-to-roll (R2R) manufacturing system, which utilizes decentralized control systems. R2R manufacturing is a continuous process in which flexible materials are transported on rollers through processing machinery where operations, such as printing, coating, lamination, etc., are performed to obtain finished products. Based on the Perron root interaction metric (PRIM), a comprehensive experimental study to analyze and minimize interaction on a large experimental R2R platform is presented. A representative sample of experimental results which demonstrate the applicability of PRIM is presented and discussed.

In this paper, winding issues in an industrial R2R printing press using a rotating turret winder are investigated by utilizing a new mathematical model and data are collected during production runs. Production data and simulation results from the developed model are analyzed to identify the causes for tension disturbances that affect winding quality. Model simulations are conducted by incorporating production data as inputs to the model to gather insights into the effect of dynamic behavior of the rotating turret winder on winding web tension. The developed model captures the various dynamic events associated with the roll change operation with a rotating turret winder. Measured data are provided to support the results of this work in improving winding tension regulation during the roll change operation. The new model together with the analysis and recommendations provides a good framework for the development of model-based tension control schemes that can further improve winding tension regulation performance, and thereby improves wound roll quality.

A novel sensor to measure web flutter is proposed in this letter. The sensing principle is based on scattering of light and directionally sensitive coupling properties of optical fibers. A linear array of optical fibers, oriented appropriately, is used to collect light scattered from a web. The flutter amplitude is determined by observing the amount of light transmitted by the fibers in the fiber array. Experiments were conducted to demonstrate the ability of the proposed sensing strategy in measuring web flutter with different kinds of web material.

Roll-to-roll (R2R) printing is a continuous process in which thin flexible materials such as paper are passed through a printing press to print the required pattern onto the material. Each printing press may have several printing units depending on the number of colors to be printed and the complexity of the pattern. The flexible material, often referred to as a “web,” is passed successively through each print unit to create a multicolor pattern. Print registration is the process of overlapping successive printed patterns to form a complex multicolor pattern and the registration error is the position misalignment in the overlapped patterns. This paper develops a machine direction print registration model in a printing press with multiple print units whose print cylinders are driven using mechanical line shafts. The registration model considers the effects of interaction between adjacent print units due to variations in material strain and machine dynamics, including various dynamic elements, such as the print cylinder, doctor blade assembly, print unit compensator roller, print unit motor, friction at various locations, etc. Measured data from typical production runs on an industrial printing press are used to corroborate the developed print registration model. Mechanical design and control design recommendations to reduce registration error in print units are also provided. The developed registration model is applicable to many R2R printing technologies, such as offset, flexo, and rotogravure printing.

Existing edge sensors use the concept of blocking/unblocking to determine web lateral position. The most commonly used sensors employ either ultrasonic or optical signals to detect the web edge position by measuring the amount of signal attenuation due to blocking/unblocking of the signal. The main drawback of this sensing method is nonuniform signal attenuation for web material variations and opacity. The research in this paper develops a new sensor which utilizes the phenomena of light scattering from the web edge and the directional sensitivity of optical fibers to determine the web lateral position.