Research On Early Warning System Of Crusher Failure

Crusher is the key equipment for mining work.The crushing production process is an important link in the entire mining production process of mining enterprises.Therefore,the crusher occupies an important position in the mine production process.The working status of the crusher has a direct impact on the mine production.Condition monitoring and fault diagnosis of the crusher can ensure the safe and stable operation of the crushing system.The essence of fault diagnosis is pattern recognition.The key to improving the recognition accuracy lies in the effectiveness of fault feature extraction and the optimization of diagnostic model parameters.Therefore,how to achieve accurate and efficient fault feature extraction and the establishment of optimized diagnostic models have become new research hotspots.Based on the vibration signals of the key parts of the crusher,this paper uses wavelet transform,particle swarm algorithm and support vector machine technology to study the intelligent diagnosis method of cone crusher faults.The main contents of this paper are as follows:(1)First,according to the actual needs of fault diagnosis,the cone crusher vibration data collection system is designed to provide a data source for subsequent fault diagnosis.(2)The wavelet analysis theory is used to extract the fault signal of the crusher.Use wavelet function to de-noise the vibration signal,and compare and analyze with Fourier transform method.Finally,wavelet analysis technology is used for frequency band analysis,and the fault diagnosis query table is established by statistical processing of the wavelet energy feature vectors of typical fault categories.(3)Using a multi-classification strategy,a multi-classification sub-model based on support vector machine technology is established.Using particle swarm optimization algorithm,cross-validation technology is used to optimize the regularization parameter c and the kernel parameter g of each sub-model to ensure that the support vector machine is classified in the optimal state.(4)Finally,the software design of the crusher fault diagnosis system is carried out,SQL Server2008 is used to design the data storage system,and the form application and human-computer interaction screen are written using C# in Visual Studio.Realize the function of online fault diagnosis of crusher and remote release of early warning information.