- 上一篇：注塑模具单一浇口改善英文文献和中文翻译
- 下一篇：空调系统英文文献和中文翻译
Control System Design Automation for Mechanical Systems
1. Introduction
Currently, many control system design problems are tackled by skillful control engineers in the industry. One objective of developing the Control System Design Automation (CSDA) is to make control system design methods available to more engineers so that they can carry out control system design and obtain a satisfactory controller in a short time. The users of the CSDA system includes robot designers and machine tool designers, whose training are more in the area of mechanical engineering. The aim is that those engineers would have a userfriendly and comprehensive control system design tool with them when they carry out their design of mechanical systems. The particular targets are industrial robots, X-Y tables, and transfer machines. Nowadays, the Computer-Aided Control System Design (CACSD) tools, such as MATLAB/SIMULINK and MATRIXx, have high performance in analysis, simulation and rapid prototyping. Those tools are very useful for a skilled engineer to design a controller. Further work have been carried out in constructing CACSD tools with which optimal control parameters for a given control structure are automatically calculated [1, 2]. These tools, however, cannot transform the requirements in the application to the specifications items for use in CACSD tools. Furthermore, they do not select an optimal control structure automatically. To overcome these problems, a high level and user-friendly control system design package is constructed for controller design of mechanical systems. The control system design environment is called CSDA. The special features of the CSDA system include the following: (1) The requirements in a particular application may not be expressed in terms of the control design specifications. The CSDA system can transform those application requirements to specifications which can be treated in the control design package.
(2) The CSDA system has a comprehensive set of control design methodologies.
Some of the more recent methods such as the LMI design approach and the
Kessler/Manabe method are included.
(3) The system can select an appropriate control structure automatically in addition to determining a set of control parameters.
(4) A user-friendly interface has been developed using the Graphical User Interface (GUI) in MATLAB.
(5) Analysis of the designed control system can be carried out with different kinds of command input. The aim is to select a suitable command control signal in the final control system.
(6) Verification of the control system design can be carried out with the DSP-CIT system [24].
(7) A knowledge base module in the CSDA system can provide guidance to the user during the design process. Heuristic knowledge from previous design experience can be codified and make available to the user.
(8) A database management module in the system can help to organize the design results of the user. A database on the past design activities can also help the designer to decide on the appropriate control structure and controller parameters. The objective is that users of the CSDA system can design satisfactory controllers easily even if they do not have much experience in control system design. Fine tuning of the controller can be carried out by a parameter optimizer.
In this paper, the configuration of the CSDA system is described in details. Also, detailed description is given to the analysis/design block. The analysis/design block is the most significant block of CSDA, which provides the desired control system design automation.
2. Configuration of the System
The configuration of the CSDA system is shown in Figure 1. The CSDA system consists of five blocks: a requirement interpretation block, a modeling block, an analysis/design block, a database management and knowledge base block, and a verification block. The modeling block, the analysis/design block, and the simulation module of the verification block are constructed using MATLAB/SIMULINK.DSP-CIT (developed by dSPACE GmbH) [21] is used for a rapid prototyping module in the verification block.