| 1. |
- Holmberg, Andreas, 1982-
(författare)
-
Experimental Determination of Aeracoustic Sources in Low Mach Number Internal Flows
- 2010
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- In this thesis, the in-duct experimental methods for determining aeroacoustic N-ports of in-duct elements are discussed and improved. The scattering matrix determination methods and the related wave decomposition methods are evaluated from measurements in an empty duct carrying a mean flow. The improvements of a new over-determination method for the source part of the N-port is studied using simulations and measurements; in quiescent air as well as measurements of the flow associated noise of a mixer plate, here a triangular plate inserted at an angle in a duct. The new method is shown to improve suppression of random errors while no improvement is achieved for bias errors. Further, the methods are applied in the study of two different aeroacoustic phenomena; one is the effect on the flow associated noise of the triangular plate achieved by varying the bending stiffness. For the most resilient plate tested, it is observed that when the Strouhal number of the flow noise coalesce with the Helmholtz number of a specific eigen-mode of the plate, the noise is drastically dampened. There is also a weaker broad band effect. The other phenomena studied is the amplification and attenuation obtained for sound waves propagating in a T-junction of rectangular ducts. It is found that by adding only 10% of inflow in the side branch relative to that in the main branch, the amplification is heavily increased. By adding another 10% the amplification is again similar to that of no side branch flow. Adding further flow lessens the effects still.
|
|
| 2. |
- Holmberg, Lars
(författare)
-
Human In Command Machine Learning
- 2021
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Machine Learning (ML) and Artificial Intelligence (AI) impact many aspects of human life, from recommending a significant other to assist the search for extraterrestrial life. The area develops rapidly and exiting unexplored design spaces are constantly laid bare. The focus in this work is one of these areas; ML systems where decisions concerning ML model training, usage and selection of target domain lay in the hands of domain experts. This work is then on ML systems that function as a tool that augments and/or enhance human capabilities. The approach presented is denoted Human In Command ML (HIC-ML) systems. To enquire into this research domain design experiments of varying fidelity were used. Two of these experiments focus on augmenting human capabilities and targets the domains commuting and sorting batteries. One experiment focuses on enhancing human capabilities by identifying similar hand-painted plates. The experiments are used as illustrative examples to explore settings where domain experts potentially can: independently train an ML model and in an iterative fashion, interact with it and interpret and understand its decisions. HIC-ML should be seen as a governance principle that focuses on adding value and meaning to users. In this work, concrete application areas are presented and discussed. To open up for designing ML-based products for the area an abstract model for HIC-ML is constructed and design guidelines are proposed. In addition, terminology and abstractions useful when designing for explicability are presented by imposing structure and rigidity derived from scientific explanations. Together, this opens up for a contextual shift in ML and makes new application areas probable, areas that naturally couples the usage of AI technology to human virtues and potentially, as a consequence, can result in a democratisation of the usage and knowledge concerning this powerful technology.
|
|
