基于多传感器信号融合的真空断路器机械特性状态评价

Mechanical Characteristic State Evaluation of Vacuum Circuit Breaker based on Multi-sensor Signal Fusion

真空断路器机械特性状态评估中存在仅依靠单一传感器信号进行评价的问题。对此,提出了一种利用D-S证据理论对多传感器信号进行信息融合来判断真空断路器机械特性状态的方法。此方法首先对位移传感器测出的运行参数利用模糊综合评判方法计算得出基于位移信号的评价结果。然后,运用小波包-能量熵谱法对加速度传感器监测的振动信号进行信号处理,提取特征向量,结合相似性原则得出基于振动信号的评价结果。最后,运用D-S证据理论中的合成法则对以上两个传感器得出的初步评价结果进行融合,得出最终评价结果。经过实例验证,该方法得出的评估结果具有更高的可信度。

The mechanical properties status of vacuum circuit breakers was usually evaluated by single sensor signal. To solve this problem, an evaluation method was proposed in this paper. This method used D-S evidence theory to combine multi sensor signals. Firstly, we got operating parameters which were measured by the displacement sensor. Then we calculated basic belief assignment of mechanical properties according to fuzzy comprehensive evaluation method. Secondly, we used acceleration sensor to monitor vibration signals. According to the method of wavelet packet-energy spectrum, vibration signals were decomposed and extracted characteristic values. Then based on the similarity principle, the characteristic values were used to determine basic belief assignment of mechanical properties status. Finally, the two basic belief assignments are fused according to D-S evidence theory. And the final fusion evaluation results are obtained. The test results showed the method was correct and effective.