This monograph opens up new horizons for engineers and researchers in academia and in industry dealing with or interested in new developments in the field of system identification and control. It emphasizes guidelines for working solutions and practical advice for their implementation rather than the theoretical background of Gaussian process (GP) models. The book demonstrates the potential of this recent development in probabilistic machine-learning methods and gives the reader an intuitive understanding of the topic. The current state of the art is treated along with possible future directions for research.
COBISS.SI-ID: 29101607
A number of supporting subsystems is required for optimal exploitation of PEM fuel cells. In this manner, an effective online condition monitoring system provides the means for achieving requirements regarding reliability and durability. Contemporary commercial PEM fuel cell power systems are equipped with DC-DC converters and voltage monitors, which provide only basic information (i.e., voltage measurements at low sampling rates and resolution). The designed embedded system extends the basic functionalities of a DC-DC converter and voltage monitor with advanced condition monitoring capabilities. The originally devised concept of the designed system is the interconnection between the DC-DC converter and the voltage monitor. The designed system incorporates: • a DC-DC converter module capable of injecting excitation signals, which are required for the condition monitoring purposes, • a high resolution voltage monitor capable of acquiring voltage and current measurements on all of the cells within the fuel cell stack, and • a condition monitoring algorithm capable of detecting the presence of PEM fuel cell faulty operation. The designed embedded condition monitoring system aims at bridging the gap between the laboratory-grade measurement equipment (highly accurate and prohibitively expensive for the industrial applications) and contemporary commercial diagnostic systems (inexpensive, but capable of only basic monitoring). The system is jointly commercially viable and precise enough to perform sophisticated diagnostic measurements.
COBISS.SI-ID: 28568359
A novel fast-filtering method for rotational speed measurements, applied to the brushless direct current motor drives, is proposed. The algorithm is tested on a valve actuator. The experimental results show that the speed jitter is substantially reduced, whereas fast reaction of the speed control loop is preserved in contrast with the classical filtering methods. The benefit of using the proposed algorithm is expected to become evident after a long-term operation of valve actuator due to reduced stress on mechanical components, and, consequently, reduced wear. A lower consumption of electrical energy, a lower probability of failure during the operation, and longer life span of the device are also evident.
COBISS.SI-ID: 29087783
The contribution concerns the data-driven design of a virtual sensor for maximum and minimum stack temperatures, which is (i) of high precision, (ii) simple to implement on conventional platforms like programmable logic controllers, and (iii) to maintain reliability in spite of degradation processes. By careful application of subspace identification, supported by physical arguments, we derive a simple estimator structure capable of producing estimates with 3% error irrespective of the evolving stack degradation. The degradation drift is handled without any explicit modelling. The approach is experimentally validated on a 10 kW SOFC system.
COBISS.SI-ID: 29917735
For the autothermal reformer, the heart of the fuel processor, a complex controller has been developed, to decouple the multivariate process, including a look-up-table treated feedforward action. Such control system enables complete control over the reaction parameters and fast, accurate response to a set-point change as well as efficient suppression of unmeasured disturbances. The control concept has been implemented in the form of embedded control system, and successfully tested with the operational unit. It thus expands the library of our control tools and allows simple transfer to other applications. The achievement was published in one of the most prestigious journals in the fields of energy, fuels and chemical engineering, Journal of Power Sources.
COBISS.SI-ID: 30059047