Gearboc signal pre-processing. Time-area characterization: the f-series decelerator calculates peaks, averages, scale differentials, variation factor, deformity factor, steepness factor, waveform indicator, peak indicator, parallel axle deceleration pulse indicator and sufficiency indicator, which are used for signal recognition at the same time, resulting in a decrease in the misdiagnostic rate of the single time-area wave. the parallel axlereducing device signal is reprocessed: the time-area signal is the original signalRight-angle shaft reducer motorThe main basis for artificial recognition of signals is the shape of the time-area signal. AR Modeling Parameters: The above input data are of a relatively large length and the F series decelerator network has a long processing time, so the parallel axle decelerator needs to find a small output-volume method. The parallel axlereducing small-wave neural network was introduced by Zhang Qinghu et al. of the well-known IRLSA information science research institute in 1992. The commonly used seed method is the pre-treatment of signals with small wave analysis, i.e. small wave space as the mode-identified characterization space, and small wave analysis to achieve characterization of the signals, which is then sent into the neurological network; The other is the so-called "F-seriesreducing machine mini-wave neural network " (Wavelet NealNetwork, WNN) or WaveletNetworkWN. The following are common methods of pre-treatment of signals: small wave neural networks are models of neural networks based on small wave transformations, i.e., the Induction motors replace the normal neuron non-linear incentive function (e.g. the Sigmoid function) with a non-linear waverend, which integrates small wave transformations organically with the nervous network and fully continues the advantages of both。GearbocSignal pre-treatment: the purpose of pre-processing of the f seriesreducing machine signal is to improve the signal structure and to maintain certain characteristics of the signal in order to achieve good recognition。

The parallel axlereducing small-wave transformation method: using Daubechis as the primary filter, the 10th-grade small-wave sub-column and sub-details are 512 points (the whole of Daubechis' low-wave-required 2 dots). RollerInduction motorIt is an important equipment that is widely used in processes such as steel rolling. Because the various data units collected do not exist, the data are subject to naturalization. Ar modelling is a 40-step ar model that determines the model parameters of ar models based on input data. In order to ensure that the effects are recognized, the range values should be sub-regulated, as is the case. The parallel axle decelerators work in a harsh and complex working environment, and each time they bite steel and shake it with a larger shock, they are highly vulnerable to malfunctions, such as increased vibrations and shock phenomena caused by the wear of the internal gear system conveyor chain, which can be reflected in the vibration properties of the parallel axle axle. Vibration signal analysis is an important means of monitoring the condition of mechanical equipment and the diagnosis of malfunctions. The f series decelerators are a method between time and frequency characterization. Artificial neural networks provide an excellent solution to topics that traditional model identification methods are difficult to solve successfully, so failure diagnostic techniques are an important area of application for artificial neural networks. Parallel axlereducing machine area signal processing: the power spectrum in the range signal processing is the calculation parameter. parallel axisGeared motorSignal recognition uses the BP Neural Network to learn about pre-processed signals as input parameters to determine the type of signal. The multilayer sensor (MLP) neuro-network theory based on BP algorithms is solid and widely applied in the numerous nervous networks of the F-series slowing machines. http://boserlgearreducer.com/product/list-pingxzoujiansuji-cn.html