Research Of Processing Parameters Of Electrical Discharge Machining Titanium Alloy Based On Grey Neural Networks Model

Titanium alloy is widely used in aerospace engines and aircraft manufacturing fields due to its high specific strength, high temperature mechanics performance and strong corrosion resistance characteristics.But titanium alloy parts, which are applied in the engine structure, often have special process and high precision requirements.Therefore,it is difficult to meet the processing requirements by using traditional machining method.EDM removes material with discharge energy, so it is suitable for machining hard cutting material due to its advantage of high machining precision,regardless of the material hardness and non-cutting force.The material of physical mechanism of EDM titanium alloy was researched in this paper, and analyzed the influence on material removal from the polarity effect, area effect, pulse discharge characteristics, the electrode and workpiece material properties, and so on. Peak current, peak voltage, pulse width and pulse interval were chosen as input parameters processing speed and surface roughness were chosen as the target output, then an orthogonal experiment was designed. Gray correlation analysis was applied and gray correlation matrix was established, and got various influencing factors of technology target of EDM titanium alloy.The method of signal to noise ratio was used to analyze orthogonal experiment results, and then the single objective optimal process parameters combination was found out and provided important data support for multiple target process parameters optimization combination.The optimizations with multiple performance objectives were transformed into the signal objective optimization by using gray correlation analysis method, and the optimization combination with multiple performance of EDM machining titanium alloy was gained.Finally, aiming at the characteristics of EDM titanium alloy and its complexity, a new modeling method of gray model combined with neural network was put forward, which had strong nonlinear processing capability, named gray neural networks model(GNNM).With MATLAB software as a platform, the prediction model of EDM titanium alloy processing speed and surface roughness was simulated based on the test data. Processing rule of EDM titanium alloy was successfully mapped out with high fitting precision and ideal predicted precision.At present,the conclusion of paper put up a new way and method.It had guidance significance without clear physical mechanism of EDM...