Stochastic approach for controllable measurement uncertainty in Industry 4.0 applications

Abstract

Increasing demands for measurement instrumentation, based on standard sampling with analog-to-digital converters in Industry 4.0 applications, like: high precision, high accuracy, low uncertainty, low power consumption, high reliability and low cost, are resulting in unavoidable tradeoffs. A novel and radically different approach - stochastic digital measurement method, is used to meet all of the former demands in a single solution. Using low-resolution converters made with high-speed voltage comparators, with added stochastic dithering signal for reducing measurement error, enables simple, robust, yet high precision solution, able to operate at high sampling frequencies. The main feature of this solution is rather unique - measurement uncertainty is determined by the length of the time period over which the input is measured, and gives the user full control over the desired level of measurement error by varying the period of measurement. This eliminates the need for many different hardware solutions with different grades of measurement errors, with a single instrument able to perform either at laboratory standard level of precision, or with higher levels of error but with ability to perform a large number multi-channel measurements at high speed, all at low cost and high reliability, thanks to the simple hardware. This solution enables drop-in replacement of the existing analog and digital equipment, and a smart upgrade in Industry 4.0 systems. An example of practical prototype is given, the stochastic smart grid meter with uncertainty level of 0.0080 % for voltage and current measurements. With appropriate sensor and transducer, almost any quantity could be measured with high accuracy and benefits of compatibility with Industrial Internet of Things.