I am measuring the open circuit voltage (Voc) of triboelectric nanogenerator that I have designed and the instrument I am testing with is a Keithley 2400. Can you tell me why the bottom or top baselines are not flat in both test images? I know how to handle these by using origin, but I do not know that what triggers these problems, such as environmental factors or incorrect operation during measurements?

In the first two images, the baseline has some degree of tilt; normally the baseline should be smooth. In the third image, the top baseline undergoes a periodic deviation-reversion process.