Microprocessor Implementation of Control Algorithms

This chapter is devoted to microcomputer systems suitable for controllers of industrial manipulators. As mentioned, a robotic controller represents a single or multi-processor system aimed at driving a robot arm to move in accordance with a user-written program. Robotic controllers are much more complex in their hardware and software structure than most industrial regulators and programmable automata for numerically controlled (NC) machines. The industrial regulators commonly represent PID single-input single-output regulators designed through the use of digital or analog integrated circuits or microcomputers. Some advanced industrial controllers represent self-tuning regulators [1] and automatically adjust parameters (proportional, differential and integral gains) to the parameters of the controlled process. Such a multichannel PID regulator corresponds to the executive level of a robotic controller (see Chap. 4). However, the executive level of an advanced robotic controller can be much more complex than a set of PID regulators when designed to accomplish trajectory following control (in contrast to point-to-point control). The executive module must compensate for dynamic effects of the mechanical arm. This compensation is usually more complex numerically and imposes the need for fast microprocessors. Robot control systems are usually much more complex than programmable automata, which were used earlier to control simple pick-and-place manipulators.