The paper presents results of simulation investigations of the measuring system, which enables measuring, with very small dynamic error, of rapidly changing signals using inexpensive transducers of not particularly good dynamic performance. Two methods for self-identification of the system sensor parameters are presented. The self-identification is performed using solely the measured signal at the system operating site. The system itself performs the correction of dynamic errors. Therefore the system shows small sensitivity to adverse influence of the operating conditions on the transducers dynamic properties. The investigations have been carried out for the second order, oscillatory and inertial transducers. The presented methods of correction can be applied to measuring of both: periodic and non-periodic signals. The simulation results, presented in a graphical form, illustrate the dependence of the correction effectiveness on the system selected parameters. Created programming tools make possible for testing the system in the neighbourhood of the selected operating point.

Traditional methods and new method for the small non-linearity measuring of electronic components and circuits are described in this paper. There are presented examples of measured intermodulation product levels on passive components.

The Dynamic Railway Diagnostics system (DRD) is a DSP-based measurement system, that analyses and stores the deformation signal of the rail caused by a passing train. The weight of the carriage is calculated by signal processing methods, by digital signal processors. This architecture makes the system capable of performing also other signal processing algorithms on this signal, e.g. faulty wheels can be detected. The DSP-based measurement method is a new approach in the field of in-motion weighing. This method can weigh and perform the diagnostics on trains travelling up to 160 km/h.

Binomial filter may be applied wherever easily predictable smoothing is required. That is possible in lidar signal processing. Digital filtering is a way of extracting noise from signal [1] beside poissonian averaging. Binomial filtering is useful in making data easier to interpret provided that proper perspective relative to measurement errors is maintained. Kaiser algorithm has been used to improve recovered data.

Measurements may contain disturbances like transients or abrupt level changes. Although the duration of these effects is often short, the computed results derived from the measurements may be heavily corrupted. In order to be able to analyse only the non-disturbed region of the signal, the detection of anomalies is needed. In this paper we introduce a method for detecting sudden changes applying windowed moving variance computation technique.