A new flow sensor system utilizes AI algorithms to improve accuracy and reduces costs

In the following article you will find information about a new gas flow sensor that uses AI algorithms to save costs. Gas flow measurement algorithms can compensate for environmental factors such as temperature, humidity, and ambient pressure to improve accuracy. For example, a look-up table with over 200,000 support points can be compressed using an artificial neural network into just 31 coefficients, resulting in a 99.98% reduction in memory usage with a maximum error rate of 1%. This cost-effective approach allows for the implementation of low-cost flow sensors. A three-disc system can accurately measure both low and high flows using multiple pressure transducers to calculate the difference across the discs, in conjunction with temperature, humidity, and ambient pressure sensors. For more information contact IMT.