A soft-sensing model on hydraulic excavator's backhoe vibratory excavating resistance based on fuzzy support vector machine

In order to measure the backhoe vibratory excavating resistance of a hydraulic excavator fast and precisely, the influences of vibratory excavating depth, angle, vibratory frequency, amplitude, bucket inserting velocity and soil type on the vibratory excavating resistance were analyzed. Simulation analysis was carried out to establish the bucket inserting velocity, amplitude and vibratory frequency considered as secondary variables and excavating resistance as primary variable. A fuzzy membership function was introduced to improve the anti-noise capacity of support vector machine, which is a soft-sensing model on the hydraulic excavator's backhoe vibratory excavating resistance based on fuzzy support vector machine. The simulation result reveals that its maximum relative training and testing error are nearly 0.68% and -0.47%, respectively. It is concluded that the model has quite high modeling precision and generalization capacity, and it can measure the vibratory excavating resistance accurately, reliably and fast in an indirect way.