The high speed cutting because of its higher efficiency and accuracy fully isapplied in various field. Even so, in order to make it further development, manyrestricting factors is waited for solving. At present, to obtain single one of the highefficiency and accuracy is not enough. How to simultaneously obtain higherefficiency under the parts accuracy guaranteed in order to better and fully play theadvantages of high speed machining. As a result, it has very important realisticmeaning that the key technologies of high speed milling surface roughness arestudied.
The influence of various factors on the high speed milling surface roughness isanalyzed. The method of theory fuzzy is put forward to obtain more accurateprediction model. And the fuzzy system is trained by the algorithm which combinewith least square method and the basic ant colony algorithm. The contrast test showsthat the final convergence error of fuzzy system trained by improved algorithm isthe smallest and the convergence effect of which is optimal.
The model obtained bythe fuzzy system is predicted through the test data and the correctness and accuracyof the model based on the new fuzzy system is proved to be better. In addition,milling parameter optimization mathematical model is established which putmachining accuracy and efficiency as the optimization goal and though the residualstress and the tool life as constraints. And the Dynamic Ant Algorithm GeneticAlgorithm is adopted to optimize the model calculation. The method not only retainsthe advantage of strong robustness of the original single algorithm but also improveits local searching ability and convergence. For different processing requirements,the accurate milling parameter combination can be got more quickly. Finally, withVisual C++6.0 as the software development platform, milling parameteroptimization system is developed to realize the MATLAB software calls, which cancomplete rapid prediction of high-speed milling surface roughness. By the systemthe time of practical production is saved and the processing efficiency is improved.
After the key technologies of surface roughness prediction based on high speedmilling and milling parameter optimization studied, the consumption of theresources in the process of actual production is reduced. The cost is reduced and theefficiency is improved. The advantages of high efficiency, high precision of the highspeed milling are fully played, and which is got more extensive application and lay asolid foundation for the further study of high speed milling.
Omni-directional automated guided vehicle （AGV） is a kind of autonomous guidedvehicle on which assembled omni-directional wheels. By means of structural advantages, ithas three degrees of freedom on the motion plane, including translation along X-axis andY-axis, rotation around Z-axis. The omni-directional AGV has more advantages toconventional wheeled AGV robot when used in any congested and narrow environments.
In this thesis, a new type of omni-directional AGV is designed based on MY3 wheel,composed of mobile platform and body. The body is divided into 3 layers, top, middle andbottom. As functional zone, the top layer reserves cargo; As controlling center, the middlelayer consists of controlling board and driving power of the robot; As driving part , the bottomlayer consists of driving electric engine and driver of engine. The chassis of omni-directionalAGV is composed of 4 groups of MY3 wheels arranged in “+” shape. This omni-directionalAGV could transport cargo weighed in 50kg flexibly in the maximum speed of 0.25m/s. Inthis paper, the kinematics model of four wheeled omni-directional AGV is set up according tothe structure characteristics of the third generation of MY wheel, which lays a goodfoundation for the following control of the robot.
In this thesis, to satisfy the requirements of the stability and real-time performance of thewhole robotic control , a controlling system, which is composed of computer, micro controllerand driver, is developed based on NI sbRIO9631 embedded controlling board. The computermainly completes programming and the human-computer interface. The micro controllerreceives the movement instructions from the PC and then converts them into the target angleof each motor. The system checks external information of the robot by multi-sensors datafusion technology, such as detecting earth orbit by means of infrared sensor, detectingobstacles around by means of ultrasonic sensor, and feeding back practical environmentimages by means of camera. In this system, controlling software is also developed based onLabVIEW, the software is simple to operate and friendly to interface. Finally, someapplication tests are conducted to conform feasibility and validity of the whole controlsystem.
In this paper, research of embroidery machine based on machine vision has been carried out.Based on the principle of machine vision systems,the platform for edge extraction was selected.To remove the effect caused by picture distortion,and to obtain accurate measurements of object,the camera was calibrated. The edge extraction algorithm was proposed,which was based on region morphology and differential gradient operator.In order to changing data between embroidery machine and outside drawing software,a operator was proposed to translate the graphic data into the sewing data.Research on the craft of sewing and the requirement functions of software,building block design for a HMI of this control system was completed.
Firstly,the platform has been built consist of camera,camera lens and light source,according to the requirement of visual analyzing and edge-extracting for small sheets of leather substance.To remove the effect caused by picture distortion,which common exists in machine vision,the camera system has been calibrated by forming a camera model.And through the camera calibrate,the accurate measurements of object was obtained.
Secondly,having studied the basic relationship between pixels,a special method ofedge-extracting of leather was proposed based on region morphology and differential gradientoperator.The result of the extraction was segmented into line, arc and curve with the Rameralgorithm.An edge-extracting operator combines least square fitting method and B-spline curveapproximating method was developed,and the analysis results will be stored in a DXF file.
Then, In order to obtain the edge information of visual analyzing and enable thedata-exchange between embroidery machine and outside drawing software,program in thecontrol system was developed to fetch the DXF file.The operator was proposed to translate thegraphic data into the sewing data.
Finally,having studied the craft of sewing and the requirement functions of software,building block design for a HMI of this control system was completed.
After debugging and processing,The embroidery machine control system achieve theprocessing requirements. The work efficiency of embroidery machine was enhanced by theproposing of the edge extraction algorithm and the support of DXF file.