| The research on the vibration signal of the washing machine is helpful to understand the running state of the washing machine,so as to perform performance detection,fault diagnosis and optimal control of the washing machine.The GB / T4288-2018 standard requires that a sensor be installed on the outer box of the washing machine to detect general vibrations such as vibration acceleration and displacement physical quantities,but the motor speed can be measured on the outer cylinder.This requires unboxing detection,which is very inefficient.In response to the above problems,this thesis proposes to install a 3D acceleration sensor on the outer box of the washing machine to extract the vibration signal and extract the rotation speed.However,the signal-to-noise ratio of the vibration signal picked up in the outer box is relatively low,and the rotation speed of the dehydration stage changes rapidly,which causes greater difficulties in accurately extracting the rotation speed.Therefore,this thesis conducted an in-depth study,and finally proposed an optimization method for the speed extraction of the washing machine in the dehydration phase based on the short-term average amplitude difference function and adaptive parameter adjustment,so as to achieve the speed extraction without disassembly.The thesis mainly does the following research work:(1)Research on preprocessing of vibration signal of washing machine.Drawing on the method of speech signal processing,the vibration signal of the washing machine is filtered,endpoint detection,framed windowing processing,etc.The method of combining the shortterm average energy and the short-term average zero-crossing rate is used to divide the inlet water,outlet water,washing section and dewatering section.(2)Preliminary research and comparison of instantaneous speed extraction methods based on Hilbert transform and Hilbert-Huang transform(HHT).Both of these methods use the principle of instantaneous frequency to extract the speed.The Hilbert-Huang transform has a slightly higher extraction accuracy than the Hilbert transform,but the overall speed error is still too large to meet the national standard requirements.(3)The research focused on the extraction of the rotation speed of the washing machine in the dehydration phase based on the short-term average magnitude difference function(AMDF)and differential threshold segmentation(DTS).(4)Based on the research in(3),further optimize the algorithm.To solve the problem that the time resolution of the algorithm is too low and the wrong selection of individual extreme points,the vibration signal of the washing machine during the dehydration phase based on APA-AMDF and RBF fitting is proposed.Speed extraction method.A combined algorithm of AMDF and APA is used to extract the speed curve corresponding to the vibration data in the dehydration stage.For the teacher mode,the outer cylinder sample center method is used for adaptive parameter adjustment.For the non-teacher mode,the second-order difference method is used for adaptive parameter adjustment,and then the RBF neural network is used for curve fitting to obtain the final speed curve of the vibration data in the dehydration stage.Three different dehydration time modes of a certain type of pulsator washing machine were used for the experiment.The experiment showed that for the method adopted and proposed in this thesis,the vibration signal speed extraction method based on APA-AMDF and RBF fitting was relatively most effective Excellent,to achieve the relative highest resolution and high precision speed measurement.The speed extraction method in this thesis is not only suitable for the automatic detection of the speed of washing machines in industrial sites,but also can be further generalized to the motor speed measurement in general low signal-to-noise ratio occasions and can be used for remote fault diagnosis services for home appliances for smart homes.Figure[56] Table[10] Reference[48]... |
PDF Full Text Request