基于蒙特卡洛方法的冗余传感器信号测量算法

Redundant Sensor Signal Measurement Algorithm Based on Monte Carlo Method

核电厂从安全角度出发针对同一测量信号(温度、压力、流量等)往往设置了多个冗余传感器。为了减少操纵员工作负荷,如何规避测量误差、传输延时和传感器故障等因素影响,正确地将多个信号综合成一个最真实的信号进行显示和处理就显得尤为重要。设计的冗余传感器信号测量算法可根据信号数据质量计算其数据质量权重,然后利用蒙特卡洛方法根据测量信号的测量误差计算每个测量值的计算权重,综合考虑数据质量权重和测量误差权重后,通过对所有信号进行加权平均值计算得出最终的真实信号值。该算法可以有效地减小测量值同真实值之间的误差,并能够极大减小数据传输超时导致数据质量降级、传感器故障导致数据值偏差过大等因素对计算结果的影响,使得计算值尽量接近真实值。通过使用MATLAB对均值法、卡尔曼滤波法和冗余传感器信号测量算法进行比较,得出冗余传感器信号测量算法在有效性和准确性两方面都表现最为优越。

From the perspective of security, nuclear power plants often set up multiple redundant sensors for the same meas-urement signal （ temperature, pressure, flow, etc）. To reduce the operator’s workload, how to avoid the influence of measurement error, transmission delay and sensor failure to integrate multiple signals into one most realistic signal for display and processing correctly is very important. Redundant Sensor Signal Measurement Algorithm is designed to calculate the data quality weight ac-cording to the signal data quality, then uses Monte Carlo method to calculate the weight of each measurement value that based on the measurement error of the measurement signal,and comprehensively considers the data quality weights. After measuring the weight of the error, the final real signal value is calculated by weighted average of all signals. This algorithm can effectively reduce the impact of degraded data quality caused by data transmission timeout, the large deviations in data values due to sensor failures, and making the calculated values as close as possible to the true values. Through using MATLAB to compare the Mean Value Method, Kalman Filter Method and redundant sensor signal measurement algorithm, it is concluded that the redundant sensor signal measurement algorithm is superior in both effectiveness and accuracy.