摘要
提出了一种可应用于模拟电路故障预测的方法。通过提取被测电路的频域响应信号,计算皮尔逊相关系数,从而表征电路元件的健康度;在获取元件在不同时间点的健康度数据的基础上,推导出电路元件发生故障时的健康度阈值;将经量子粒子群算法优化的相关向量机算法用于故障预测,预测各个时间点的元件健康度变化轨迹并估计模拟电路的剩余有用寿命。该预测方法计算简单、通用性强,适用于实时预测。故障预测仿真实验与实例实验证明了方法的有效性与先进性。
An approach for analog circuit fault prognostics is proposed in this paper.The frequency domain response signals of the circuit under test are extracted,and the component health degree is characterized through calculating the Pearson product-moment correlation coefficient（PPMCC）.The sample data of the health degree at different time are acquired and the health degree failure thresholds of the circuit components are deduced.The relevance vector machine（RVM）algorithm improved by quantum-behaved particle swarm optimiza-tion（QPSO）algorithm is utilized to predict the trend of health degree trajectory with respect to time and estimate the remaining useful life of the circuit.The proposed fault prognostics approach has the merits of simple computation and widespread applicability,and is appro-priate for real time prediction.Simulations and experiment validate the effectiveness and superiority of the proposed approach.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2014年第8期1751-1757,共7页
Chinese Journal of Scientific Instrument
基金
国家自然科学青年基金(61102035)
国家杰出青年科学基金(50925727)
国防科技计划(C1120110004、9140A27020211DZ5102)
教育部科学技术研究重大项目(313018)
中央高校基本科研业务费专项资金(2012HGCX0003)
安徽省科技计划重点项目(1301022036)资助
