中文摘要：

为提高全自动酶免分析仪等微量移液系统的移液精度和可靠性，提出了一种动态法液位探测和双极值法压力监控相结合提高微量移液精度的方法。首先，通过探针式电容传感器动态法进行液位探测，有效减小电机运动惯性以及浸润现象产生的毛细管作用对液位探测精度的影响；然后，用双极值法对移液过程进行压力监控，避免堵针、样本量不足、气泡等异常移液现象对移液精度的影响。在此基础上，利用基于GM（1，1）模型的灰色动态滤波方法对电容传感器、压力传感器的输出进行滤波，提高传感器输出的稳定性，从而提高对液位判定及移液异常现象判定的准确性。通过微量移液实验对该方法进行了验证，实验结果表明，基于探针式电容传感器的液位探测动态法与传统单向测量法相比，液位探测精度由±0．3mm提高到±0．1mm；在保证液位探测精度的前提下进行基于双极值法压力监控的微量移液操作，能够对移液过程中异常移液现象进行实时判断并保证压力曲线信息的完整性，将移液精度由±5．0％提高到±2．4％。

英文摘要：

In order to improve the pipetting accuracy and reliability of automatic enzyme-linked immunoassay analysis sys- tem,a new method combining liquid level detection of dynamic method and pressure monitoring of double-extremum meth- od is proposed. Firstly,the dynamic method of probe type capacitive sensor is used to perform liquid level detection,the in- fluence of the inertia of motor motion and the micro capillarity caused by infiltration between the probe and liquid on liq- uid level detection accuracy is reduced effectively. Then,the double-extremum method is used to perform pressure monito- ring,and the influence of some abnormal pipetting phenomena such as,air bubble,insufficient sample and tip blocking on pipetting accuracy is avoided. Based on above ,a grey dynamic filtering method based on dynamic GM（ 1,1 ） model is used to filter the outputs of the capacitive sensor and pressure sensor, and improve the stability of the sensors. Consequently, the veracity of liquid level and abnormal pipetting phenomenon determination is improved. Micropipette experiments were car- fled out to verify the feasibility of the proposed method. Experiment results indicate that compared with traditional single direction measurement method,the dynamic method of liquid level detection based on probe type capacitive sensor im- proves the probe type capacitive sensor accuracy from ±0.3 mm to ±0.1 mm. The micropipette operation based on pres- sure monitoring with double-extremum method can judge the abnormal pipetting phenomena in pipetting process in real time,and ensure the integrity of the pressure curve information. The proposed method improves the pipetting accuracy from ±5.0% to ±2.4%.