Woodworking tool wear state diagnosis method based on genetic BP neural network

Abstract

The invention discloses a woodworking tool wear state diagnosis method based on a genetic BP neural network. The woodworking tool wear state diagnosis method comprises the following steps of 1, collecting a sample set including a tool front angle gamma, a main shaft rotating speed n, a cutting depth h, an approximation coefficient of machine tool main shaft power Pt and tool wear VB; and 2, establishing a tool wear state diagnosis model by using the genetic BP neural network. According to the technology, a power signal is used as a tool wear state diagnosis signal, the sensor cost is reduced,the sensor installation problem is solved, the signal precision is not disturbed by the machining environment, meanwhile, the tool front angle gamma, the main shaft rotating speed n and the cutting depth h are considered, and the model can be suitable for various cutting parameters; and the genetic BP neural network of a relu training function is used for establishing the tool wear state diagnosismodel, and the calculation speed and the diagnosis precision of the BP neural network are greatly improved.

Description

technical field [0001] The invention relates to the field of tool wear state diagnosis, in particular to a woodworking tool wear state diagnosis method based on a genetic BP neural network. Background technique [0002] Woodworking knives are an important part of the manufacturing process of wood products. During the cutting process, they are affected by factors such as processing parameters and machine tool performance, causing tool wear, resulting in tearing and cutting of the edge surface of wood products, affecting the surface roughness and overall quality of wood products. beautiful. The use of accurate and reliable tool wear state diagnosis technology can increase the utilization rate of CNC machine tools by 50%, reduce production costs by nearly 30%, and can improve product quality and reduce product defect rate. [0003] At present, there are two main types of tool wear state diagnosis technology. One is to collect the characteristic values ​​of various signals rela...

AI Technical Summary

Problems solved by technology

[0003] At present, there are two main types of tool wear state diagnosis technology. One is to collect the characteristic values ​​of various signals related to the tool wear state, analyze the relationship between the signal characteristic value and the tool wear state, establish the tool wear threshold, and analyze the tool wear state. Diagnosis, such as Chen Guohua and others collect the power signal of the machine tool spindle, extract the signal entropy value, and establish the tool wear threshold to diagnose the tool wear state (authorization number: CN108490880B), but this technology ignores the tool wear when the cutting parameters change. The threshold value will also change accordingly; the other is to establish a tool wear state diagnosis model after collecting signals related to the tool wear state to realize the tool wear state diagnosis. For example, Guan Shan et al. extracted image features and combined them with the collected acoustic emission Signal fusion, use the discrete hidden Markov model to establish a tool wear state classifier, and automatically diagnose the tool wear state (authorization number: CN107378641B), but this technology ignores that during signal collection, coolant and chips will affect image quality, and the external environment Interference will also cause too much signal noise, which will affect the accuracy of the tool wear status diagnostic model. In addition, the cost and installation method of the sensor are also the key to restricting the practical application of the technology itself.

Images