Abstract: In a method and arrangement for adapting a sliding mode fuzzy controller to destabilizing changes, control characteristics of the sliding mode fuzzy controller, a closed control loop, are characterized by linguistic variables and are interpreted using linguistic rules in an adaptation block in the control loop which also employs fuzzy control. The rules are configured such that an optimally fast approach of a momentary state condition to the switching line or switching surface of the sliding controller occurs. For adaptation, supporting points are defined in the characteristic of the sliding mode fuzzy controller and these are shifted for the purpose of a desired control behavior, such as by variation of the output fuzzy sets. A separate fuzzy rule set is thus designed for each individual support point.

Abstract: For the purpose of detecting signals by means of fuzzy-logic classification, these signals are sampled in time and the sampled values are compared with fuzzy threshold values with the aid of fuzzy-logic rules. Strengths which evaluate the similarity of the signal to be detected with the relevant master signal given the application of the relevant rule are determined for each master signal of a signal catalog. Furthermore, a membership is calculated for each master signal as the arithmetic mean of all the strengths, averaged over all the rules, and the signal to be detected is assigned to that master signal whose membership is a maximum.

Abstract: A method for adjusting scale factors of fuzzy controllers or conventional non-linear controllers provides that these scale factors are selected in such a way that a standardized cross-correlation function of a scaled input signal is optimized by means of an associated output signal.

Abstract: In a method and apparatus for fuzzy control in contrast to standard methods and arrangements, no sharp quantity is processed as an input quantity of the fuzzy controller; instead, an input quantity is supplied that is affected with a probability density distribution. Thus it is not a sharp value but the probability density distribution of an input value whose membership degree is determined in the inference procedure of the fuzzy controller. This procedure offers advantages, particularly given asymmetrical distribution densities. First, one thus obtains a better response behavior of the controller and, second, the disadvantages that would occur given a pre-filtering of the signal, such as information losses, are avoided.

Abstract: In this controller, a manipulated variable is produced as a sum of a linear compensation term and a product term, the compensation term being a linear combination of a contouring error and its time derivatives, and the product term being formed by multiplying a fuzzy control term by a saturation term. This controller combines the advantages of the sliding-mode controller with boundary layer and those of the fuzzy controller, and thus represents a robust controller for a wide class of nonlinear nth-order systems.