Because the process of blade in electrochemical machining(EMC) can be effected by many factors, such as blade shapes, machining electrical field, electrolyte fluid field and anode electrochemical dissolution, different ECM machining parameters maybe result in great affections on blade machining accuracy. Regard some type of aero-engine blade as research object, five main machining parameters, applied voltage, initial machining gap, cathode feed rate, electrolyte temperature and pressure difference between electrolyte inlet and outlet, have been evaluated and optimized based on BP neural network technique. From 3125 possible machining parameter combinations, 657 optimized parameter combinations are discovered. To verify the validity of the optimized ECM parameter combination, a serial of machining experiments have been conducted on an industrial scale ECM machine, and the experiment results demonstrates that the optimized ECM parameter combination not only can satisfy the manufacturing requirements of blade fully but has excellent ECM process stability.
Aero-engine Vibration Signal Blind Separation Based on BP Neural Network Algorithm