This project belongs to the field of CNC machine tools and processing. CNC machine tools are the machine tools of the equipment manufacturing industry, and their processing accuracy and intelligence level directly determine a country's manufacturing level. However, there is a huge gap between domestic machine tools and foreign advanced levels. Error compensation is an economical and effective solution, and a necessary function of intelligent machine tools. However, the machine tool configuration and error sources are complex, the thermal characteristics are complex, and the rotating axis errors are coupled with each other, resulting in complex error models, poor robustness, and rotating axis errors measurement. Difficulty, at the same time, the high real-time nature of compensation control (10ms level) and the closed nature of the CNC system seriously restrict the batch application of intelligent geometry-thermal-mechanical comprehensive error compensation for machine tools. With the support of projects such as the 04 Special Project and the National Natural Science Foundation of China, the project team has been conducting common technical research on error measurement, analysis and intelligent compensation of CNC machine tools for nearly 20 years, and has achieved the following innovative results: 1. Established a unified 3-5-axis machine tool Multi-error comprehensive mathematical model: For the first time, multiple error sources such as geometry, heat, and force were considered, and a comprehensive model of multiple errors of machine tools was established. Then a new singular function was used to establish a unified multi-error unified model of machine tools with different configurations of 3-5 axes, realizing the realization of automated and unified modeling of machine errors of different machine tool configurations and different sources in the production line improves the accuracy and speed of modeling. 2. The optimal placement theory of temperature sensors and various highly robust thermal models were created: the mechanism of the action of internal and external heat sources on the machine tool on thermal deformation errors was given, and the optimal placement theory of temperature sensors was proposed to minimize the number of sensors required; Considering multiple internal and external heat sources, a variety of self-learning and self-correcting thermal error intelligent modeling methods based on the heat conduction mechanism and temperature-deformation hysteresis analysis were created. The model is typically used in high-speed and high-precision mold machining centers, realizing adaptive suppression of the influence of thermal errors, and the dynamic accuracy is stable to 5.5% all day long.~6um, exceeding the accuracy index of similar machine tools in Makino, Japan. 3. An error measurement system and identification method for 3-5 axis CNC machine tools were established: a laser interferometer measurement based on vector multi-step and a comprehensive error measurement method based on compound trajectory were proposed to efficiently and accurately identify all 46 errors of 5-axis CNC machine tools., a complete measurement system has been established for the first time at home and abroad. 4. Created a multi-error comprehensive real-time intelligent compensation method and device for CNC machine tools and production lines: A multi-error comprehensive real-time compensator that integrates the compensation functions of all mainstream CNC systems has been developed. The compensation accuracy is higher than that of existing high-end CNC systems, and Real-time compensation of comprehensive errors can be achieved within the interpolation period (about 8ms), increasing the accuracy of single machines by more than 65%, up to 95%. A real-time data interaction and analysis platform for the production line based on compensator architecture and fog calculation has been developed to realize intelligent sensing of the processing status of each CNC equipment in the production line, interactive correlation compensation and intelligent precision self-healing. The results are typically applied to Shenyang Machine Tool High-speed Turning Center Intelligent production line, etc. The project has obtained 15 invention patents, 6 utility models, 3 soft copyright, 1 monograph and 61 SCI papers. The results have been applied in batches in CNC equipment, aerospace, automobile parts, 3C and other industries, and have served the manufacturing of a variety of military products. Users include Shanghai Machine Tool Factory, Rifa Precision Machinery, Shenyang Machine Tool, Ningbo Tianrui Precision Machinery, Xi'an Jiaye Aviation Equipment, etc. More than 40 companies have used more than 1000 units in exclusive batches. The results have won the National Excellent Expo Award and the Shanghai Silver Excellent Expo Award. In the past three years, the new output value has been 755 million yuan, the new profits and taxes have been 185 million yuan, and the funds have been saved by 286 million yuan. Machine tools have been exported to the United Kingdom, Germany and other countries, earning 10.8 million US dollars in foreign exchange, and achieving huge economic and social benefits. This project has significantly improved the accuracy of my country's CNC machine tools and intelligent production lines to a world-class level, leading the development of machine tool error compensation technology at home and abroad.