| Summary: | 碩士 === 國立臺灣大學 === 機械工程學研究所 === 106 === During the operation of the machine tool, the machine tool is affected by the temperature of different factors, resulting in non-uniform deformation of the mechanism of the machine tool, thereby changing the original machining path to produce errors and causing the accuracy to decrease. In order to improve and upgrade the machining accuracy, the research on compensation system of thermal deformation error of machine tool has become an important research direction. The main purpose of this research is to analyze thermal deformation of die-sinking electrical discharge machine from the experimental results. Then, selecting the software compensation, which is a relatively inexpensive method to reduce the impact of thermal deformation by Adaptive Network-Based Fuzzy Inference System (ANFIS).
In this study, measurement setups were established using ISO 230-3 that is from International Organization for Standardization (ISO) standard for machine tool temperature deformation to measure the displacement of thermal deformation caused by changing in environmental temperature and temperature rise of dielectric oil. Measurement system that measures thermal temperature rise and thermal displacement was established in the machine tool by using six temperature sensors and five contact displacement sensors.
The experiment was conducted under various conditions, such as changes in the environmental temperature and the temperature of dielectric oil, and simultaneously measured the data of six kinds of temperature and five degrees of freedom displacement. Environmental temperature change experiment was carried out between air conditioners at 20 °C and 30 °C, it was found that the maximum thermal deformation in the Z-axis direction. The temperature rise experiment of the dielectric oil was carried out under constant temperature, it was found that the maximum thermal deformation in the Y-axis direction.
In order to find the relationship between temperature and thermal deformation, using temperature as input and thermal deformation as output, training is performed via ANFIS to establish a model for predicting thermal deformation by temperature. Then, perform the compensation verification by the PC-based controller. Under the experimental conditions of environmental temperature change and the temperature rise experiment of the dielectric, the ANFIS model can compensate the thermal deformation error value of the two axial directions to within ±10 μm.
|
|---|
