| Abstract [eng] |
Sound systems are used not only in information exchange or entertainment industry but also in control of technological devices. These days audio systems widely use digital signal processors. DSP calculates mathematical signal processing in real time. Adapting the DSP for technological sound analysis makes it possible to project systems for equipment failure identification. Using technological fault identification systems, the faults can be detected in their early stages preventing the stops in production processes. To gather the information about the equipment state various systems, sensors and methods are used. Those can be vibration sensors which detect motor vibration increases, microphones to detect the change of acoustic waves emitted by the equipment, analysis of motor current changes. To use this information for control of processes system first must process the sensor signal. Computers, controllers, or DSPs can be used for this. For sound signal processing and calculation of filter coefficients program MATLAB was selected. Digital signal processor programming was done by using SigmaStudio program. In the research part FIR filters were used to investigate four technological faults: chain conveyor with loose belt between motor and reductor, grain cleaning device with worn out bearings, valve with fan blade touching the protective cover and valve with inaccessible end position faults. For each sound file of the technological fault fast Fourier transform was used to convert the signal from time domain to frequency domain. The effect of foreign technological devices on the recording of the motor emitted sound was also investigated. The study was conducted using four technological devices at distances of 0,5 m, 3 m, and 4 m. It was found that external devices do not affect the sound recording of the investigated technological device. A program which identifies and indicates the technological faults of equipment was created in program SigmaStudio. In this program FIR filters, signal delay, signal comparison and signal generation function blocks were used. A program using digital signal processor ADAU1701 capable of identifying the technological failures of chain conveyor with loose belt between motor and reductor, grain cleaning device with worn out bearings, valve with fan blades touching the protective cover and valve with inaccessible end position has been developed. |
