1. 研究了透析器生产现状、透析器的整体结构以及具体工艺过程、透析器离

1. Study on the status of the dialyzer production, the overall configuration of the dialyzer and the specific process, dialysis, centrifugal machining process, dialysis process centrifuge and causes adverse phenomena, the typical method of processing parameters of the optimization problem. <br>2. Study of the SVM algorithm dialyzer centrifuge process design optimization methods described herein involved. For the original algorithm is slow and prone to over-fitting and other issues, this paper SVM algorithm was improved. Improved contrast before and after the algorithm a set of random normal distribution effect of sample size on the training data set 2000, the improved effectiveness and practical training relative error algorithm are better than the original SVM algorithm. <br>3. dialyzer orthogonal test design process of the centrifuge, and the analysis of the situation after the experiment, the orthogonal Range experimental data and analysis of variance, the main factors determining the dialyzer and the main processing centrifuge time relationship, find the best combination of parameters between the factors. Using orthogonal data, using the optimized SVM algorithm centrifuge machining process modeling to predict the effect of modeling the effect of the current prediction models used several comparison results show that the improved prediction SVM algorithm is better than several other two models. <br>4. Analysis of dialysis centrifuge process parameter optimization requirements, design centrifuge dialysis process parameter based on B / S system optimization software to achieve the main functions of the software. <br>According to the simulation experiments, improved centrifuge process parameters of the dialyzer SVM Optimization This design <br>method can achieve the desired effect prediction model <br><br>Keywords: dialysis, centrifugal processing, orthogonal, SVM

1. The present situation of dialysis machine production, the overall structure of dialysis and the specific process, the process of centrifuge processing of dialysis machine, the undesirable phenomenon and cause of the process of dialysis centrifuge, and the typical method of optimizing the processing process parameters are studied.<br>2. The SVM algorithm involved in the optimization method of the dialysis centrifuge processing process designed in this paper is studied. In view of the problems such as slow calculation speed of the original algorithm and the easy overfitting, this paper makes an improvement to the SVM algorithm. By comparing and improving the training effect of the pre- and post-algorithm on a set of data sets with randomly obeying normal distribution sample size of 2000, the actual training effect and relative error of the improved algorithm are better than the original SVM algorithm.<br>3. The orthogonal experiment of the centrifuge processing process of dialysis machine is designed, and the situation after the experiment is analyzed, the polar difference and anovaizing analysis of the orthogonal experimental data are analyzed, the main factors affecting the process of dialysis centrifuge processing and the primary and secondary relationship are determined, and the best combination parameters between the factors are found. Using orthogonal experimental data, the optimization of SVM algorithm is used to model the centrifuge process, and the modeling effect is compared with the prediction effect of several commonly used models, and the results show that the prediction effect of the improved SVM algorithm is significantly better than that of other models.<br>4. The requirements of the processing process parameter optimization of dialysis centrifuge are analyzed, and the software of the processing process parameter optimization system of dialysis centrifuge based on B/S mode is designed, and the main function of the software is realized.<br>According to the experimental simulation, the processing process parameters of dialysis centrifuge based on improving SVM are designed in this paper.<br>Method can achieve the expected model prediction effect<br><br>Keywords: dialysis, centrifuge processing, orthogonal experiments, SVM

1. The production status of dialyzer, the overall structure of dialyzer, the specific process, the centrifugal processing process of dialyzer, the bad phenomena and causes of the processing process of dialyzer and centrifuge, and the typical methods for the optimization of processing parameters are studied.<br>2. The SVM algorithm involved in the process optimization method of dialyzer centrifuge designed in this paper is studied. In view of the slow calculation speed of the original algorithm and easy to produce over fitting problems, this paper improves the SVM algorithm. By comparing the training effect of the improved algorithm on a group of data sets with random normal distribution sample size of 2000, the actual training effect and relative error of the improved algorithm are better than the original SVM algorithm.<br>3. Designed the orthogonal experiment of the processing technology of dialyzer centrifuge, and analyzed the situation after the experiment, analyzed the range and variance of the orthogonal experimental data, determined the main factors affecting the processing technology of dialyzer centrifuge as well as the primary and secondary relations, and found out the best combination parameters among the factors. Using the orthogonal experimental data, the optimized SVM algorithm is used to predict the centrifuge processing process. The results show that the improved SVM algorithm is better than other models.<br>4. The requirement of processing parameters optimization of dialyzer centrifuge is analyzed, and the software of processing parameters optimization system of dialyzer centrifuge based on B / S mode is designed, which realizes the main functions of the software.<br>According to the experimental simulation, the processing parameters optimization of the dialyzer centrifuge based on the improved SVM designed in this paper<br>The method can achieve the expected prediction effect of the model<br>Key words: dialyzer, centrifuge processing, orthogonal experiment, SVM