No abstract

No abstract

Recently, the demand of a high speed and high accuracy increases in the media transport (or feeding) system (MTS) which is one of the most essential parts in copiers, digital printers, ATM’s and so forth. This paper illustrates the development of a high-speed, high-accuracy paper feeding system including a feedback feeding speed control system with a learning controller. The controller is designed to reduce the slippage between the feeding rollers and paper in the presence of the periodic disturbance in the nip force of the driven roller which is mainly caused by the eccentricity of the driven roller. The lack of an accurate model of the plant often makes control system designers to face difficulties at an early stage in designing a robust controller for highly non-linear multi-body dynamics system such as the MTS with flexible media. To mitigate the problem, a so-called co-simulation approach, the combination of a controller development tool and a model developed in MBD, is often adopted and it would allow the controller designer to expedite the design and evaluation of control systems. In the paper, using a co-simulation approah a MTS with a paper is modeled and a repetitive learning controller is develped and evaluated.

In this paper, we made a simple paper feeding system which is one of MTS (media transport system) and controllers. The plant has a flexible paper and two driving rollers and two driven rollers. The control system has two conventional PID controllers. Skew angle and feeding speed of MTS deteriorate the quality of feeding system. In order to control a feeding speed and skew of feeding paper, we control rotational velocity of two driving rollers. Therefore, this controller has two inputs and two outputs as MIMO (multi-input and multi-output) system. The control inputs were the feeding speed and the skew displacement of the paper. The control outputs were the rotational velocity to each driving roller. To find appropriate PID gains of two controllers, we proposed an optimization technique. We assume the system variables and performance of a whole system as follows. PID gains of two controllers for skew and feeding speed are system variables. System performance is both skew and feeding speed. We simulates to making mathematical correlation using global Kriging interpolation. To find appropriate value of system variables, optimization method is simulation in sequence as following method. First, the optimization solver simulates with DOE (design of experiment) tables to find correlation equation of both system variable and performances. Then, the solver guesses the appropriate values and simulates if the system variables are appropriate or not. If the result of validation doesn't satisfy the convergence and iteration tolerance, the solver makes a new Kriging models and iterates this sequence until satisfy the tolerances.