Mohammad Hajizadeh and Mike G. Lipsett
Fault detection, Wavelet transform, Suspension strut, Mining haul truck, Suspension system
Failure of suspension struts on large off-road haul trucks can result in accidents or structural damage. These vehicles typically have a payload measuring system to monitor production in a mine by measuring pressure in the four suspension struts of each truck. In this study we use these pressure signals for fault detection of the suspension struts based on a continuous wavelet transform (CWT). Morlet wavelet functions are employed to estimate the natural frequencies of the system and the frequency components of the signal with relative maximum energy. The fault detection method is evaluated using a half-car dynamic model that has been simulated in MATLAB. The three most common faults in a strut are each modeled in a lumped-parameter model with Gaussian random noise as input to the wheels to simulate road irregularities. Wavelet analyses of the outputs shows that a Morlet SWT can distinguish between different kinds of faulty strut and normal strut behaviour. This approach is now being tested on production and maintenance data from trucks in an operating mine.
Important Links:
Go Back
