Abstract:
The aim of this project is to explore new directions in diagnostics, control and parameter identification strategies of ac electric drives under critical operating conditions. Main attention will be paid to sensorless drive control and estimation in standstill and low speeds. We propose to explore suitability of methods from Bayesian identification and stochastic control in this area. High quality stochastic model will be developed as well as a new sophisticated simulator for testing of the control methods. The main concern will be with modeling of disturbances from existing deterministic models and their statistical properties. Existing methods will be interpreted as approximations of general theoretical methods; finding a closer approximation will then yield improvements of existing methods. Application of advanced theoretical solutions will reveal closer understanding of the problem and prepare background for further research. The results of this project will be applicable for: 1) improved reliability and safety of modern drives via smart drive diagnostics, and 2) improved efficiency, reliability and cheaper manufacturing cost of drives via robust and adaptive control allowing precise operation of the drive with
reduced number of sensors.