- A History of Increasing Complexity
- Mechatronic System Organization
- Amplifiers and Isolation
- Scope: The Unit Machine
- Control
- Real-Time Software
- Nasty Software Properties
- Engineering Design and Computational Performance
- Control System Organization
- Software Portability
- Operator Interface
- Multicomputer Systems: Communication
- The Design and Implementation Process
This chapter is from the book
This chapter is from the book
This chapter is from the book
1.4 Scope: The Unit Machine
Control systems can range from extremely simple to enormously complex. The same design methods and software development techniques will not work across this very wide range. This material is targeted to a middle level in the complexity spectrum: what we call the unit machine or unit system. In general terms, a unit machine encompasses a closely interacting set of physical components. Because of the close interaction there are tight time constraints that apply to the interactions between parts of the machine or system. More specifically, the parts of a unit machine either exchange real physical power, or exchange material with little or no buffering. The complexity can range from a few motors or heaters to a system with tens of such components. A unit machine can stand alone, or be part of a larger system such as a manufacturing cell or plant. However, when it is part of a larger system the interactions with other parts of the system are at a much slower time scale than the interactions within the unit machine.
The first step in any design process is to assess the scope of the problem. Underestimating the complexity is the most dangerous and most common error. If design methods are used that cannot handle the level of complexity needed, the chances of achieving a reliable, high performance control system are severely compromised. With luck, satisfactory performance may be achieved, but only with a loss in predictability of the design and implementation process. At worst, the system will simply never perform satisfactorily and will cost far more than initially estimated as well.
Overestimating complexity is not as dangerous, but it can lead to a loss of an economically competitive position. A system designed using methods suitable for much more complex systems will work well but may use more resources than would be necessary if a simpler method were used.