The aim of the study is to investigate whether induction devices can replace fan convectors, which is the climate system that is most common in hotel rooms today. The study compares the importance of the power of the two climate systems, induction devices and fan convectors, for use in hotel rooms. Climate systems controlled by presence require high power if room temperature is to be quickly regulated. The results show that fan convectors with power of 3000 W are often oversized for the hotel room, as the room being studied has a dimensioning cooling power requirement between 600 – 1700 W and heating power needs between 600 – 1000 W. Induction devices with a power of 1500 W are capable of meeting the cooling and heating needs if the room is appropriately designed according to its conditions, including sun protection and high window quality.    When the climate systems have been switched off during the absence period, the room has reached different temperatures depending on the design of the room. The time to cool or heat the room to the desired temperature, which is related to how hot or cold the room has been, differs between fan convectors and induction apparatus. High-power fan convectors have always been able to cool or heat the room in under 1.5 hours, while induction devices often can do it in a shorter time, except in cases where the power requirement is over 900 W.   What distinguishes the rooms studied is the placement of windows in different directions, the quality of the windows such as permeability, g-value and insulation ability, u-value, the rooms also have different thermal masses and sun shielding. The room that requires the highest heat output has a window with high g-value to get the most heat in during the winter months to reduce the power requirement, which turned out to have the opposite effect. When the permeability of light is high, it also contributes to more sunlight and heat being given to the room in the summer, therefore the cooling power requirement is also high for rooms with such a window.   To reduce the power requirement in an existing room, measures such as window replacement, additional insulation or sun protection change can be performed. The measures have different effects and can contribute up to 150 W lower power requirements, which means that both climate systems can be used in hotel rooms without problems.

Syftet med arbetet var att framställa en beräkningsmodell som behärskar att beräkna ett klimatsystems energiförluster samt livscykelanalys, LCC. Samt att använda denna modell vid jämförelse av två befintliga system. Modellen skapades i programmet Excell och använder sig av angiven indata för att beräkna resultat. Den användes för jämförelse av fyra lokaler i VIDEUM ABs byggnader. Två kontor och två föreläsningssalar jämfördes. Den beräknade skillnaden i energiförbrukning kunde i huvudsak härledas till det ena systemets längre drifttid över helgen samt att ett av kontoren drevs med högt konstantflöde hos ventilation även under frånvaro. Huruvida styrning av system sker efter koldioxidkoncentration eller temperatur verkar spela mindre roll för systemets förluster. Dock påverkar onödigt högt ställda flöden samt för långa driftstider energibehovet desto mer. / The thesis focused on compiling a calculation model suitable for calculating an indoor climate system energy demand and life cycle cost, LCC. The model was created in Excell and uses given input data to calculate the results. The model was used to compare four different premises located in buildings owned by VIDEUM AB in Växjö. Two offices and two lecture halls was compared. The calculated differences in energy demand could be derived to longer operating times during weekends for one system. One office had a large constant air flow even during absence which also led to a greater energy demand. Whether the system was regulated by using carbon dioxide concentrations or temperature as indicators on air quality didn’t seem to affect the energy demand significantly. Unnecessary high flow rates and operating times affects the premises energy demand the more.

Recent advancements in processing power, storage capabilities, and availability of data, has led to improvements in many applications through the use of machine learning. Using machine learning in control systems was first suggested in the 1990s, but is more recently being implemented. In this thesis, an embedded climate system, which is a type of control system, will be looked at. The ways in which machine learning can be used to replicate portions of the climate system is looked at. Deep Belief Networks are the machine learning models of choice. Firstly, the functionality of a PID controller is replicated using a Deep Belief Network. Then, the functionality of a more complex control path is replicated. The performance of the Deep Belief Networks are evaluated at how they compare to the original control portions, and the performance in hardware. It is found that the Deep Belief Network can quite accurately replicate the behaviour of a PID controller, whilst the performance is worse for the more complex control path. It was seen that the use of delays in input features gave better results than without. A climate system with a Deep Belief Network was also loaded onto hardware. The minimum requirements of memory usage and CPU usage were met. However, the CPU usage was greatly affected, and if this was to be used in practice, work should be done to decrease it. / Många applikationer har förbättras genom användningen av maskininlärning. Maskininlärning för reglersystem föreslogs redan på 1990-talet och har nu börjat tillämpas, eftersom processorkraft, lagringsmöjligheter och tillgänglighet till rådata ökat. I detta examensarbete användes ett inbäddat klimatsystem, som är en typ av reglersystem. Maskininlärningsmodellen Deep Belief Network användes för att undersöka hur delar av klimatsystemet skulle kunna återskapas. Först återskapades funktionaliteten hos en PID-regulator och sedan funktionaliteten av en mer komplex del av reglersystemet Prestandan hos nätverken utvärderades i jämförelse med prestandan i de ursprungliga kontrolldelarna och hårdvaran. Det visade sig att Deep Belief Network utmärkt kunde replikera PID-regulatorns beteende, medan prestandan var lägre för den komplexa delen av reglersystemet. Användningen av fördröjningar i indata till nätverken gav bättre resultat än utan. Ett klimatsystem med ett Deep Belief Network laddades också över på hårdvaran. Minimikrav för minnesanvändning och CPU- användning var uppfyllda, men CPU- användningen påverkades kraftigt. Detta gör, att om maskininlärning ska kunna användas i verkligheten, bör CPU-användningen minskas.