Research On Fault Diagnosis Of Roller Hydraulic System

As a kind of important engineering compaction machine,vibratory roller is widely used in engineering fields such as road engineering,airport port and municipal construction.Vibratory rollers are usually cooperated with engineering machinery group in large construction site.When the rollers fail,the machine equipment with which they work together will be forced to shut down,which will affect the progress of the project,delay the duration and even cause great economic loss.Because of the worse working environment,relatively complex condition,the hydraulic system which is the main system of vibratory roller,has a higher probability of failure.The research work on the fault diagnosis of the hydraulic system of vibratory roller is beneficial to timely eliminating the malfunction of the hydraulic system and exerting the maximum efficiency of vibratory roller,which can bring great economic and practical significance to ensure the engineering quality,accelerate the project progress and improve the economic benefit.By analyzing the working principle of a hydraulic vibratory roller and the basic faults and eliminating methods of hydraulic system,this paper proposes the fault feature extraction based on principal component analysis method.Through pattern recognition of faults by two algorithms: fuzzy inference and fuzzy neural network,the fuzzy neural network is better robust and stable for fault recognition by simulation and comparison.In this paper,the traditional principal component analysis method is improved.By taking the axial plunger pump as an example,make twice dimensionality reduction of monitoring data to reduce the redundancy,and to ensure that the data after the dimensionality reduction still carries enough data of the original sample.Then,using the data as sample,fuzzy inference and fuzzy neural network are used to identify faults.Through simulation analysis,this paper establishes a fault diagnosis model of the hydraulic system of the vibratory roller on the basis of principal component analysis and fuzzy neural network.The simulation results show that this model has good fault tolerance and robustness to the fault diagnosis of the hydraulic system of the vibratory roller,and avoids the disadvantage of the standard neural network that easily trapped in local convergence,which can be widely used in the fault diagnosis of the hydraulic system of the roller.