4 - Feedforward Control.pdfVIP

4 Feedforward Control 4.1 Introduction In this chapter we will continue to make the assumptions made in the previous chapter: that the system under control is linear and that the secondary actuator is fully active. A secondary disturbance can thus be generated which destructively interferes with that due to the primary source, as a result of superposition. In contrast to the previous chapter, however, we will now begin to relax the assumption that we have no inform- ation about the original excitation of the mechanical system, due to the primary source. This allows us to use such information as the basis of afeedforward control approach. There are two important examples of where some prior knowledge of the excitation due to the primary source can be obtained. The first of these is where the disturbance is deterministic. In principle the future behaviour of such a disturbance can be perfectly predicted from its previous behaviour. In practice, a reference signal is usually derived from the primary source of the disturbance, and used to maintain the synchronisation of the secondary excitation. For example, in the case of disturbances caused by reciprocating machines such as internal combustion engines, a tachometer signal related to the crankshaft rotation is often used to generate a reference signal. The second example of where prior knowledge of the primary disturbance is available is when the vibrational disturbance is propagating through a mechanical structure, and a sensor can be used to detect this incident disturbance. We should make clear that this detection sensor is not the same as the response sensor discussed in the previous chapter, which was used to generate the excitation signal for