National Chung Cheng University

Tongtai Machine & Tool Co., Ltd.

Wei-Jen Chen

Chih-Chun Cheng

Wen-Nan Cheng

Ping-Chun Tsai

Wen-Peng Tseng

Yu-Hsin Kuo

Yu-Sheng Chiu

De-Shin Liu

Chen-Wei Chuang

Wen-Chieh Kuo

Yao-Huan Lei

Chih-Ta Wu

Shao-Rong Su

A sensor placement optimization device is provided, which may include a preprocessing circuit and an operational circuit. The preprocessing circuit may perform a pre-process for the sensing signals of a plurality of temperature sensors, installed on a machine tool, to generate a pre-processed data. The operational circuit may execute a normalization for the pre-processed data to generate a normalized data, perform a principal component analysis for the normalized data to generate a dimensionality-reduced data and implement a principal component regression for the dimensionality-reduced data to obtain the contributions of the temperature sensors. Then, the operational circuit may rank the temperature sensors according to the contributions thereof to generate a ranking result and execute a screening process according to the ranking result to select at least one redundant sensor from the temperature sensors; afterward, the operational circuit may remove the redundant sensor from the temperature sensors.

Sensor placement optimization device and method thereof

A machine tool health monitoring method which is to use a predetermined plurality of vibration sensors on a plurality of components of a machine tool and to drive motors of the machine tool to excite the machine tool using an electronic device while the health status of the machine tool is good, and then to perform a diagnostic process to obtain a characteristic cluster consisting of a plurality of modals, and then to define the characteristic cluster as a sample health characteristic cluster. The diagnostic process includes the procedures of vibration transmissibility obtaining, singular value decomposition, curve fitting and modal establishing. In addition, excite the machine tool and proceed the diagnosis process to obtain a current health characteristic cluster. Finally, the current health characteristic cluster is compared with the sample health characteristic cluster to judge whether the machine tool is healthy or not.

Machine tool health monitoring method

Growing community of inventors

Min-Hsiung, Taiwan

Wei-Jen Chen