| CPC G01N 15/1434 (2013.01) [G01N 2015/1493 (2013.01); G01N 2015/1497 (2013.01)] | 11 Claims |

|
1. A sensor arrangement for detecting features of particles, the sensor arrangement comprising:
an emitter for emitting electromagnetic radiation;
a detector for receiving the electromagnetic radiation emitted from the emitter and for providing detector signals based on the received electromagnetic radiation;
a measurement volume irradiable by the electromagnetic radiation emitted by the emitter, the measurement volume configured for receiving particles flowing therethrough;
a digitizing unit for digitizing the detector signals; and
an evaluation unit for evaluating the detector signals;
wherein the evaluation unit stores a trained algorithm for machine learning, and wherein the algorithm is configured for determining at least one feature of the particles based on the detector signals, and
wherein the algorithm for machine learning is trained by:
I. measuring and/or simulating test signals from the detector for particles with known features;
II. adapting parameters used by the algorithm for machine learning to reduce a difference between the at least one feature of the particles calculated by the algorithm for machine learning and the known features of the particles; and
III. repeating steps I to II with further particles with known features until the difference between the at least one feature of the further particles and the known features of the further particles falls below a predetermined limit.
|