Nonlinear control with digital signal processors: room for improvement

Realizations and performances of nonlinear control algorithms by digital signal processors (DSP) are discussed in this paper. The author explains some of the important DSP features that make it possible to calculate certain mathematical operations at high speed. Basic principles of a few nonlinear control approaches are summarized, with the objective of pointing out key calculations in adaptive control, feedback linearization and variable structure control. Several references and example implementations are used to illustrate the feasibility of implementing complex nonlinear control algorithms with modern DSPs. The examples presented are: feedback linearizing control of a magnetic bearing, variable structure control of a magnetic bearing, and variable structure control of a resonant link DC-DC converter. Experimental studies are also used to point out some possible shortcomings and pitfalls of using present-day DSPs for nonlinear control. One conclusion is that modern digital signal processing devices are well suited for certain types of control algorithms, but further advances in the technology can be explored