Abstract: The optimal layout of sensor measuring points has great significance for condition monitoring, fault diagnosis and health management of mechanical equipment. Focusing on the gear transmission system, a method of measuring points optimal layout based on effective independent analysis of box mode and frequency response function is studied, and the fusion signal of optimal points is used to monitor the gear transmission system vibration. The dynamic model of the gearbox is established, and the modal modes of the gearbox are obtained by modal analysis. Based on the modal modes, the initial measuring points layout is obtained by effective independent analysis. At the same time, the main driving bearing seat and direction of the gear fault excitation on the box are analyzed. The frequency response function of the initial measuring points about the direction of the main exciting bearing seat is obtained by harmonic response analysis. The measuring point was further optimized based on principal component analysis and effective independent analysis, and the optimized measuring points sensitive to gear transmission fault excitation is selected by sensitivity analysis. Based on the importance weights of the optimized measuring points, the spectral weighted fusion is carried out. The vibration monitoring of gear transmission is realized based on the fusion spectrum. The measured data analysis of the bevel gear transmission test platform shows that compared with the fusion spectrum of any other measuring point group, the fusion spectrum based on the optimized point group has a larger overall response amplitude, and is more sensitive to the vibration response of the bevel gear transmission, and the monitoring effect is better.