Identification Of Cutter Breakage In Variable Cutting Parameters Milling

With the development of modern production and manufacturing technology, the modernized equipment such as CNC, FMS, CIMS system are applied manufacturing technology widely, in order to guaranteeing these processing system can be produced lastingly and high-efficiently, and avoided causing the precision reducing of processing part because of the wearing and breakage of the cutter, even damaged the machine tool, so we must control the whole processing course effectively. It is indicated that adopting monitoring system can avoid people and technological trouble that factor cause shutting down for 75% of time. So, it is an important subject in an advanced manufacturing technology that the breakage of the cutter is controlled online.There are a lot of methods in the cutter monitors, then we consider the numerous signals, which reflect the information of the cutter, the real processing condition and the complexity sensor installed, we chose the vibration signal as the research object. Imitating the processing course of variable cutting milling during the research of problem we have designed a set of experimental equipment, hardware system includes vibration acceleration apparatus, designing magnifying circuit by myself, data collection card etc. and software adopts C language which can realize the data of collection and memory. This test setting is used to collect the vibration and axis rotate speed of cutter breakage monitoring system.There are often mixed with the noise during the vibration signal gathering because of the test system and locale influence of various kinds of objective factors, so we have carried on the wiping off direct component, lowpass and eliminating bizarre data etc. By time and frequency domain's analysis and disposal of vibration signal we extract the average square root, standard and power spectrum of tooth meshing frequency as characteristic value reflecting cutter breakage, and then get the necessary sample collection in our follow-up process of the test.Finally, utilizing the modern recognition means — the artificial neural network, we carried on BP network, which belongs to the tutor studies of feed forward network. Recurring to the non-linear mapping of BP network, it will carry on non-linear pattern-recognition. According to accuracy, real-time character of cutter breakage monitoring system, we have chosen corresponding BP network structure and designed systematic parameter of network, and then utilized this network system to carry on the network tests of two kinds of operating modes.