Spray quality of Multi-hole nozzles decides the quality of diesels, but the control of spray is a hard work in the field of internal combustion engines. The spray quality can be changed by abrasive suspend jet which is a flexibility manufacturing method and the quality will vary with system pressure, abrasive kind, diameter of abrasive, preparation method, viscosity of working medium and grinding allowance and some other parameters. This paper tries to optimize the spray quality of nozzles by using orthogonal experimental design to select the values of these parameters. To evaluate the quality of spray, a scoring method was offered and input parameters were organized to three parameters before experiment. According to the orthogonal table of factor level (L16(45)), the line charters between the parameters and the spray quality was shown. This paper considers that the optimal combination is A2B2C2. Then the relationship between spray quality and the parameters optimized is deduced. And the significance of the forum was checked, it shows that the forum was significant at level α=0.01.

An embebbed system based on S3C6410 controller and Linux operating system is established to overcome the deficiencies of the traditional control system for high accuracy processing of grinding diesel injector nozzle because of its high speed and real-time performance. In this paper,the general structure of the system is designed firstly. Secondly, the detailed hardware platform is constructed and its peripheral equipment is selected. In addition,the software structure is presented. Finally, the main application program,including of processing control program and algorithm of the flow signal processing which is the key factor for high accuracy maching, is studied. The actual processing data shows that this system has good real-time performance, high processing accuracy and stablity.

Here to introduce a parallel processing video splicing system with multi-processor. The main processor gets the encoded video data from video source and outputs the video data to the coprocessors simultaneously after decoding the data. Those coprocessors capture the video data needed for splicing and display it on the objective monitor. Owing to this method, the sophisticated time synchronization algorithm is no longer needed. The proposed approach also lowers the system resources consuming, and promotes the accuracy of the video synchronization from multiple coprocessors.

In this paper, a control system based on the prediction of processing flow in Abrasive flow machining is designed. In this system,flow is predicted by an improved GM(1,1) model in conformation of background value. Combined with fuzzy control system, it can adapt the pressure to meet the processing requirement automatically. Experiments proved that the improved GM(1,1) model can predict the processing flow accuratly, and the fuzzy control system based on grey prediction can improve the machining accuracy of micro-hole AFM effectively.

In this paper, neural network and grey fuzzy control technology are applied in Abrasive Flow Machining (AFM) to grind the mico-hole in th nozzle of the twin flapper-nozzle valve. An intelligent control system with fine tuning working pressure is established that can predict the process parameters automatically before machining and forcast the flow to adjust the working pressure in machining.The result of experiment indicates that this system has high level of intelligent and can get very high machining accuracy.