安全阀流量试验位移数据优化研究

Research on Displacement Data Optimization of Yield Valve Flow-rate Experiment

位移数据是安全阀压力-流量试验中计算流量的重要依据。由于液体介质的冲击振动及位移传感器数据采集误差等因素的影响,需剔除原始位移数据中的误差数据进行优化处理。依据位移数据分布特征进行数据分类,确定时间间隔,按照时序将每个时间间隔内采集的数据点的集合称为一簇;采用阻尼牛顿型法确立设计变量和优化目标函数,对每簇数据进行优化迭代,搜索其数据中心代替该簇数据集合作为此时间间隔初始时刻的有效位移数据;并采用三次样条插值法对有效位移数据进行曲线拟合得到时间-位移曲线。

In yield valve flow-rate experiment, displacement data are important basis to calculate flow-rate. Due to the influences of liquid impaction and data acquisition error from displacement transducer, the initial displacement data must be optimized. According to the distribution features, the displacement data were divided into different groups. After determining the time interval, following the time sequence, displacement data in the time interval were brought into the same group. When the design variables and target function were determined, each data cluster were iteratively computed to search the data center which was regarded as operative displacement data to the initial time of this time interval, and the data cluster were substituted by the operative displacement data. By the cubic spline interpolation method, the time-displacement curve was obtained through fitting the operative displacement points.