液体火箭发动机涡轮泵多维度传感器优化布置

A method of multidimensional optimization placement of turbopump sensors is proposed to address the issues of missing structurally important modal information and incomplete extraction of fault-sensitive information during the health monitoring of liquid rocket engine turbopumps. Experimental verification is conducted using a fault simulation test bench. Firstly, a finite element model of the turbopump's optimization area is established to conduct structural constrained modal analysis and transient dynamic analysis of bearing faults. Subsequently, candidate points for sensor placement are indentified based on the discrete artificial bee colony algorithm and transient analysis results. The final sensor placement method is determined through a comprehensive multidimensional evaluation method of sensors. Finally, the comprehensive indicators between the proposed sensor placement method and the traditional method are experimentally compared. The calculation results show that, compared with effective independence, after optimizing the placement of measurement points, sensor monitoring signal modal shape compliance increased by 20. 1%, and the accuracy of bearing fault detection increased by 27. 5%. This verifies that the multidimensional optimization placement method has good comprehensive performance in fault diagnosis.