Multi-source fusion fault prediction method of hydraulic brake system based on ga-bp network

Abstract

The invention discloses a hydraulic brake system multi-source fusion fault indicating method based on a GA-BP network. The method comprises steps of: sampling oil liquid in a hydraulic brake system, randomly dividing samples into a training set and a test set, analyzing the samples and carrying out multi-source information fusion to obtain analysis data related to the oil liquid; using the genetic algorithm to optimizing an initial value of a neural network; using analysis data of the training set for modeling of the GA-BP network; testing the neural network obtained through training by use of the test set until performance of the neural network can meet requirements; and finally, using the trained GA-BP neural network to analyze states of the hydraulic brake system and indicating possible faults. According to the invention, the states of the monitored hydraulic brake system can be qualitatively and quantitatively evaluated; the fault diagnosis process is quite intelligent; and detection accuracy is quite high.

Description

technical field [0001] The invention belongs to the technical field of hydraulic brake system fault detection, in particular to a GA-BP network-based multi-source fusion fault prediction method for a hydraulic brake system. Background technique [0002] With the advancement of energy saving and emission reduction work and the maturity of hybrid technology, the RTG (tyre crane) energy-saving system based on tire crane hybrid technology has gradually matured. Compared with conventional diesel generator-driven tire cranes, the most prominent advantage of hybrid tire cranes is fuel saving; secondly, because hybrid tire cranes use low-power diesel engines, noise and exhaust emissions are reduced. Hybrid tire cranes eliminate the main disadvantages of conventional tire cranes, and provide technical support for the wide application of tire cranes as a container operating machine. All of these make the hybrid tire crane have great application prospects in practical applications. ...

AI Technical Summary

Problems solved by technology

②Wear particle identification technology is the most distinctive technical core of ferrography analysis in oil analysis, and this analysis method mainly depends on the knowledge and experience of analysts, and the operation is time-consuming and boring; although the research direction of wear particle identification It has attracted many researchers, but due to the low clarity and particle dispersion of ferrographic analysis, the realization of the automatic identification of the spectrum has been restricted, and no breakthrough has been made. The current wear particle identification is still unable to leave the analysts

From the perspective of fault diagnosis of hybrid tire cranes, compared with traditional tire cranes, hybrid tire cranes have a much more complex internal structure and a lot more devices, which also makes the forms of faults more diverse. The modeling of the increased difficulty

Images