With the continuous development of precise gear processing numerical control technology, the traditional method which continuously improved machine parts designing accuracy, manufacturing accuracy and assembly accuracy to eliminate or reduce the machine error is not able to meet the requirements of the development of high precision gear processing. It has a very big limitation. Even if this method is feasible, it is also very expensive on the economy. So it becomes very important to improve machining precision by CNC machine tool error compensation technology.In this article, CNC gear hobbing machine was treated as the research object. Two kinds of dynamic error mathematics model mainly on the cutting force error and thermal error and their error compensation technology were researched.First of all, the whole error sources of CNC gear hobbing machine were briefly studied and analyzed, which are mainly including the geometric error, cutting force error, thermal error and other errors. About the geometric error, error identification and measuring method was put forward especially in view of the special structure of the rotating shaft of gear hobbing machine. Then, the cutting force error model of CNC gear hobbing machine was established based on the BP neural network and particle swarm optimization. After that, the steady and transient temperature field and displacement field of this machine tool were simulated by the finite element method. Besides, the data of thirteen temperature measuring point and a displacement measuring point were extracted, which were changing over time. Four key points were obtained finally which were used for modeling by the method of hot key point identification. Based on the multiple linear regression method, the mathematical thermal error model of this machine tool was established. Finally, the error compensation strategies were put forward combined with our self-developed gear hobbing CNC system. |