Research On Processing Dimension Prediction For Typical Process

The characteristics such as randomness, nonlinearity, uncertainty of machining dimension dynamic system, determine the complexity of machining dimension prediction. Forecast compensation control technology can effectively improve the machining precision and quality in machining, to establish a scientific prediction model is the key to the forecast compensation control technology.In order to improve the dimension accuracy and reliability of the prediction method, the typical machining process was taken as research object in this thesis. Based on grey theory and Markov chain, the dates of honing and turning were analyzed respectively to explore and study the different predictive performance of forecast model adapted to the different dimension characteristics for the corresponding prediction model algorithm. Three kinds of prediction model algorithm were set up to analyze the adaptation of mechanical processing size prediction model by using MTLAB software. The main content would be as follows:1. Give an example of honing procedure: A combined model combined with the equal dimension and new information grey Markov forecasting model was established to forecast the dimension of internal round honing process. The results indicates that this model could describe the processing dimension changing trend, timely follow data randomness, and response the probability of predicted results and the range values. The predicted values could fit the actual values well, the differences between them are small. Thus, this model has higher forecasting precision, and could provide basis for the forecast dimension of the typical machining- honing processing dimension.2. Combining the advantages of both grey model and Markov chain, put forward a weighted Markov SCGM(1,1)c combined model for cylindrical turning dimension prediction. The results show that this model can be used in the machining of cylindrical turning dimension prediction. The simulation analyses prove that this mathematic model not only discovers the trend of the cylindrical turning processing size, but also reduces the influence of stochastic fluctuating, and the precision of the result is satisfied.3. With intensive studied for machining dimension forecast compensation control technology based on two kinds of combined model, different dimension was selected to compare and analyze the degree of predicting precision affected by dimensions. The results show the relationship between the precision of the two models and the selection of dimension is inseparable.4. Based on their characteristics, two groups of date and models were especially analyzed to create a new model named equal dimension and new information grey weighted Markov. And the adaptability problems of forecast compensation control technology were discussed by the three kinds of model’s selection and prediction results.