中文摘要：

为获得足够和有效的测量信息，提高系统的可诊断性，应采用合适的测量策略并对传感器布置进行优化。为此，提出了量化描述系统可诊断性的三个测度（可检测性／可分离性／可预测性），并将多工位制造系统测量信息处理分为工位间传递和工位内检测。在保障可诊断性的前提下，提出了传感器布置优化方法，降低了测量费用和诊断时间。通过具体加工实例，验证了该方法的正确性和有效性，为传感器布置优化提供了新的理论依据和实践参考。

英文摘要：

To acquire enough and effective sensing data for measurement and improve diagnosability of manufacturing system, a proper measurement strategy should be employed for sensor allocation optimization. To deal with this problem, three indices were defined to quantatively describe the system diagnosability, namely, detectability, isolability, and predictability And then the sensing data utilization for variation control of different stations was divided into variation transmission across-station and variation distinguishing within-station. Thirdly, within the framework of ensuring diagnosability, an optimal sensor allocation strategy was proposed for better sensing cost to reduce measrement time. Finally, an example was used to demonstrate the proposed analytical procedures, which provided an innovative theory basis and practical reference for the sensor allocation optimization and diagnosability analysis of a multi-station manufacturing system.