Research On Shaft-kind Straightening System Based On Support Vector Machine

Shaft lines have become increasingly high demand in industry,and shaft parts straightness are regulated mostly by the alignment link.In the process of straightening shaft,the automatic straightening machine plays a decisive role during shaft alignment.Through the shaft parts three point bend straightening the bending deformation,so as to achieve the elimination of shaft parts of the original warping,and the shaft parts straightness will be improved.Due to the low energy consumption and wide range of application,shaft type parts automatic straightening machine has been widely used in shaft parts processing production,but research on straightening machine automation are still focused on two basic problems--straightness detection problem and straightening stroke determination problem.With the development of automatic control technology,a certain degree of automation of detection of shaft-kind straightening system has been achieved,but due to the existence of random error of the straightening machine driven rotary center which results in workpiece bending inaccurate measurement consequence.The appearance of vector machine and neural network artificial intelligence algorithm is widely used in the compensation of various random errors,and theoretical basis for the straightening machine workpiece bending measurement error compensation has been provided.Aimed at the straightness detection of shaft parts problem,the workpiece bending measurement error model with support vector machine(SSVM)was constructed,and the empirical method of cross validation method(cross validation)was utilized,cyclic voltammetry(CV)was used to optimize the parameters of the model,and straightening machine workpiece bending measurement error were more accurately predicted and compensated by using vector machine(SSVM)· Straightening system electrical schematic was drawn and straightening equipment assembly was debugged,the construction workpiece under pressure volume prediction model of the sample data was obtained.Straightness of shaft parts is a symbol of a country's level of industrialization,and straightening machine is indispensable equipment shaft parts linearity.In the process of the straightening of the shaft parts,the determination of straightened parts(straightening stroke)is the decisive factors after straightening straightness.Hot research spot of straightening stroke focuses on the realization of automatic parameter determination using intelligent methods.In order to illustrate the superiority of the support vector machine,the neural network will belong to the same machine learning language together for the workpiece press amount forecast modeling.The construction of support vector machine(SSVM)model and the BP neural network model of straightening machine workpiece pressure volume were predicted and the prediction results of analysis,and illustrates its feasibility and practicability.