Ultimate Temperature Control of Central Air Conditioning System

Background on air conditioning:

During 1840s, when there was no cool air from air conditioning to expel the extreme heat in summer, tropical disease that survive in hot, humid conditions, such as malaria, flourished and there was no easy way to make patient in hospital to feel more comfortable. Dr. John Gorrie, who was a respectable scientist from Florida, proposed the very first idea of air cooling in hospital. However, his original idea of air cooling required shipping large cubes of ice from other places in the world to Florida. Because of the impracticality of this idea, Dr. Gorrie started working on the idea of producing artificially cooled air. He designed and built the very first machine that produced artificial ice using a compressor powered by horse, water and wind which was granted patent eventually. Although the machine was very rudimentary and not very efficient, Dr. Gorrie laid the foundation for modern air conditioning and cooling.

What does current central air system look like?

In modern days, cities have been swamped with edifices and high-rises. With the development of technology and people’s increasing demands of comfortable ambient living and working temperature, air conditioning is vital to our daily lives and economy and central cooling system was developed and installed in practically all of the buildings. How is a central air conditioning system different from a general household air conditioner? Though both have one overall controlling system, a central air conditioning system directs temperature-adjusted air flow into each room. In other words, each room in the building is connected to the central controlling air unit by a tube in which temperature-adjusted air flow is transferred. Also, each room is connected in series by tubes to allow more transfer of inlet air and more efficient cooling or heating effects. Figure 1 below presents a diagram of a general central air conditioning system when it is used for cooling for one single room.

Figure 1. Diagram of a central air conditioning system when it is used for cooling for one single room.

Instead of introducing the complex infrastructure of an outdoor unit and how a compressor works, the simplest version of the process is that temperature of air flow is set by the user in the room and the outdoor unit will deliver air flow at that temperature to each room in the building after receiving the signal from the user. Air flow into each room can be turned off by the user if the room is overcooled or overheated. The air that is cooler or warmer than the desired temperature will be returned to the air unit and be sent back into the room after the temperature of the air has been altered to the desired temperature.

Problems with current central air conditioning:

Central air conditioning provides us with the convenience of rapid cooling of a whole building, but it only controls the temperature of air flow. The result of this is that although temperatures of inlet air flow to the two room are the same, the resulting temperatures of the two rooms are not same because central air conditioning system does not know the current temperature of the room or any possible heat transfer between the inside and outside due to temperature difference. The result is that two rooms receiving air flows with same temperature end up with different room temperatures. One example is that, on the second floor of Acopian Engineering Center in Lafayette, due to the lack of sunshine, Prof. Soh’s office is usually much cooler than Prof. Woo’s, although, as we can sense, the streams of air flowing to both rooms have the same temperature. Not only at Lafayette college, in most edifices and high-rises around the world, the temperature of the room and the temperature of the hallway are different. Another example which I am sure most people have encountered will explain the problem more clearly. On a freezing winter day, the lobby of a hotel, which is on the first floor, has a temperature that is lower than the temperature of the room on the second floor as more heat loss occurs in the lobby due to the frequent opening of the entrance door although the first floor and second floor are receiving air flow with the same temperature.

Though not critical to the operation of the central air system, the temperature difference between two rooms contributes significantly to the user experience and efficiency of energy consumption. In the first place, walking through adjacent rooms in the same building with different temperatures is an uncomfortable experience, and may even cause sickness. Moreover, the current central air conditioning system may deliver air much cooler than the room temperature at unnecessarily high rates, which means faster heat transfer occurs between the overcooled room and the surroundings, resulting in unnecessary high flow rates and waste of energy and electricity.

What is out new central air system?

Our proposed design and modification of current central air conditioning system will eliminate the problem of different temperatures between two room receiving the air flow with the same temperature due to different heat transfer happening between inside and outside of two rooms. Desired room temperature will be specified by the user and the room should be kept under the desired temperature by our new central air system. The key difference between our proposed design and current central air system is that temperatures of individual room will be measured and monitored and will become our controlled variable. Therefore, we will have one controlled variable in each of our rooms. Also, each room is not connected in series.

From our energy balance of each room:

Variation of air temperature in the room is directly related with the temperature and flow rate of air flowing in and any flow rate of air of the room flowing out under the closed door or through the opened door. Temperature of room can be better monitored and controlled by manipulating temperature and flow rates of air flowing in from the central air unit. Therefore, temperature and flow rates of the air flow into the room will be taken into account and will be treated as our manipulated variables. Our goal is to minimize the temperature difference between the two rooms while maintaining their temperatures at our desired values; the actual room temperatures of the two adjacent rooms should be within 1˚C higher or lower than the temperature specified by the user. In other words, we are trying to achieve automation of temperature control.

What could happen to our system?

Although manipulating the flow rate and temperature of the inlet air flow from central air unit is crucial to controlling the temperature of the room, we could still have deviation from desired room temperature due to potential disturbances. In real life, since the room is usually not insulated and certain temperature difference between the inside and outside of the room always exists, heat transfer to some degree between the inside and outside of the room will inevitably happen. Also, in our proposed design, the system has two adjacent rooms and any heat transfer between the two rooms due to temperature difference can happen and will cause deviation from our desired temperature. The same situation can happen between hallway and each room. If we expand the term of heat transfer to the outside (Qout) in our energy balance above, we can get a clearer correlation among heat transfer, room temperature and outside temperature:

Where h is our heat transfer coefficient between inside and outside, A is the area of the room where heat transfer with outside takes place, h’ is the heat transfer coefficient between the room and the hallway, A’ is the area of the room where heat transfer with hallway takes place and Q1-2 is the heat transferred from one room to the other room. From the equation, we can see that since the room temperature is changing as airflow from air unit is flowing in, heat transfer rate between inside and outside, between hallway and room and between room 1 and room 2 is always changing and heat is not transferred at a constant rate. Also, different outside temperature, different hallway temperature and different room temperatures will result in different heat transfer rate, which will be reflected in changes in room temperature. In other words, heat transfer rate out is a function of the room temperatures, outside temperature, and hallway temperature. Therefore, measuring room temperatures, outside temperature and hallway temperature is a way of assessing heat transfer rate and monitoring disturbance. Moreover, if someone opens a door or a window, air with different temperature will flow in and disturbance to room temperature will happen, resulting in increase of heat transfer rate. In this case, variation and deviation of room temperature will be noticed by the thermometer inside the room and signal will be sent to the air unit to adjust the temperature and air flow rates of inlet airflow to minimize the changes of room temperature and restore the room temperature back to the desired temperature.

Other examples of potential disturbances include heat transfer from sunlight radiation, which is a direct cause of different room temperature in our AEC example, any possible electrical heat generated from laptops or computers inside the room, open of refrigerator, which could result in certain amount of heat loss since the temperature inside a refrigerator is relatively low, etc. All these examples are potential disturbances that can lead to changes in heat transfer rate and will lead to changes in room temperature. Therefore, it is safe to say that monitoring room temperature is a way to monitor disturbances to temperature control.

How is it going to be designed?

Desired room temperature will be specified by the user and will be sent to the temperature transmitter. Adjustments will be made on the temperature and flow rates of inlet air flow based on the signal from temperature transmitter. Different control systems can be applied to control the room temperature. One of the most common system is the feedback loop system. A feedback loop system is a control system where controlled variable is measured and deviation variables are not measured. Since room temperature, which is our controlled variable, will be measured and disturbance variables, such as sunlight radiation and electrical heat generated inside the room, are difficult to measure, a feedback loop system is appropriate and will be utilized in our new central air system. Figure 2 below presents a general process diagram of our new central air conditioning system.

Figure 2. General diagram of new central air conditioning system

In the figure above, thermometer is installed in each room to monitor the temperatures in each room. Desired temperature of each room will be specified by the user. Outside and hallway temperatures will be measured by thermometers outside the room and in the hallway and sent to the junction. Current room temperature will be compared with the set point temperature, outside temperature and hallway temperature at the junction to determine the temperature deviation from the set point temperature and obtain an approximation of heat loss rate to the outside and hallway. Temperature deviation and approximation of heat loss will be collected by the temperature transmitter and a decision signal will be transmitted to the flow control or/and air unit to change the flow rate or/and the temperature of the air stream. Any disturbances to the system will be reflected in the deviation of room temperature and will be sent to the air unit to adjust inlet air flow temperature and rates.



Lester, Paul. “History of Air Conditioning.” U.S. Department of Energy, 20 July, 2015. https://energy.gov/articles/history-air-conditioning




  1. I think this is a very interesting topic to choose. Often times one does not think that there is a way to fix the problem of one room being colder than another unless there is a separate unit. I like the fact that this is a much more technically savvy and applicable idea than the rest that I have read. What I like the most is that it is an improvement to a system that is used by countless people around the world, as central air can be found in the homes of many 1st world countries. That being said, I believe your idea would be most suitable to help with heaters in heated rooms. A problem that I always face is that when it is cold outside I will walk into a heated room and start sweating because the room is too hot to account for the cold air outside. I believe that your proposed changes to the air conditioning unit would be incredibly helpful if applied to central heating.
    I think that your control scheme is very intuitive and well explained. To describe your system you used two equations to show how the tendencies of the temperature in the room change due to other rooms and the environment. This was very helpful and interesting considering many other groups did not include equations. You make it apparent that you are trying to control the temperature of each room as a separate entity by manipulating the output temperature of the air conditioner while accounting for disturbances such as heat gain through the windows, other rooms and from the sun. I believe that this hits all the necessary variables to make a successful change to the room. Due to the multiple temperature readers per room, the manipulated variable should be able to account for all the disturbances and keep the control.
    As much as I think this is a great idea that would work very well, I think there is one major concern for why it hasn’t been produced already: money. In order for this to work according to plan, there needs to be a temperature reader or thermostat in each room of the building/home. Say there is a five room home with three hallways, that means there would need to be around eight temperature sensors. I don’t know exactly how much these sensors would cost, but say each one is one hundred dollars. That would mean that a family will be paying eight hundred dollars to make a household marginally more comfortable. This is not a luxury that most families will feel is necessary enough to have for that price. Along with this, most central air units don’t have air in every single room, it seems that in order for this to work, every room would need to be able to have air inputted separately. This once again would cost more money for the consumer and may not be worth the difference. I think that if you can find a way to make the system more cost effective, it will be a very good and applicable invention.

  2. A reimagination of central air conditioning is not something I had thought of before nor did I ever do the research to understand how central air conditioning worked. I had believed that central air conditioning’s goal was to maintain the entire building at the same temperature. I now see the complexities of using central air conditioning for an apartment building. Your control design for a central air conditioner aims to be more efficient and more effective than the current standard being used. By reducing heat losses, less cold air is needed to maintain the desired ambient temperature, reducing the energy consumed by the system. Saving energy and reducing emissions is also a big plus for any system, and when it can be accomplished on a scale like that of changing the air conditioning systems for skyscrapers and homes, the positive environmental impact that would have would be astounding.
    The control scheme for the air conditioning scheme proposed is more complicated than those of the other groups. The manipulated variable, controlled variable, and potential disturbance variables are all clearly stated and explained. By measuring the temperatures of a room and all of its surrounding areas (neighboring rooms, the hallway, and outside) and deducing the required flow rate to maintain the desired ambient temperature and keeping all the rooms within one degree centigrade of each other would definitely reduce the energy needed for the system. And that was proven through the energy balances performed on the room.
    I can see this control scheme being implemented into a new construction project, specifically for an office building or apartment building. The biggest deterrent for this would be the cost of implementing all of the controllers and thermometers, however, that cost would be masked by the generally high cost for constructing the building. Also, it could be argued that the energy savings over time would eventually pay for the capital cost of the central air conditioning. The only factor that could make that not true would be if the energy consumed by all of the thermometers counter-balanced the savings from reducing heat losses.

  3. First of all, I think you guys did a really good job! Thanks for your guys’ project, now I understand why temperatures in different rooms controlled by the same central air conditioning system might vary significantly. The improvements you guys did on the original central air conditioning system are plausible and appropriate. I believe the proposed changes to the system would be helpful for keeping all rooms in the same building at a relatively consistent temperature.

    The structure of this paper is well organized. It first introduces the history of air conditioner, and then states the current question and potential improvements. The proposed control scheme is well explained and reasonable. I really like those two equations describing the heat transfer and balance in each room. These two equations help me understand the way you guys to determine the controlled variables, manipulated variables and disturbances in this system. This method is straightforward and unique since many other groups just used a bunch of words instead of equations. As you guys said in the paper, the controlled variable is the temperature of each room manipulated by the temperature and flow rates of air flowing in from the central air unit. The potential disturbances for this system are sunlight radiation, temperature differences between inside and outside of the room, hot air coming in due to human’s behaviors and heat produced by the electronic devices inside the room. I think your manipulated and controlled variables are able to hit the desired temperature in each room as thermometers in each room would send the actual temperature to the controller of the central system and the system would make the correct decision based on the temperature difference between the actual number and the set point. For those proposed disturbances, I agree with the majority of your points but I don’t think the heat generated by those electronic devices would make a significant effect on the room temperature as this amount of heat is almost negligible. For example, it is almost impossible to heat the room up by charging cell phones and laptops on a cold winter day.

    Although I do think your modifications are applicable, there is one major disadvantage of this improved system: the high cost. In order to keep all rooms at their desired temperature, you guys choose to install thermometers in every room and every hallway. If a house has six rooms and two hallways, they have to purchase eight thermometers. If this family adopts your plan, they need to pay for the capital cost, labor fee, and the cost of the electronic system updates. It might be a relatively big amount of expense for this family. To decrease the total cost, I think you guys can use suitable software to simulate the detailed heat transfer in this house based on the capacity of the central air conditioner and determined disturbances, and then estimate a possible temperature for each room. Based on the simulated temperature, we can possibly use one thermometer to monitor more than one room’s temperature. In this way, people would spend less money.

  4. Well written & articulated blog post overall. The beginning sections were both interesting and informational. I have to admit, I have never given much thought to why some rooms are hotter than others in a building whose rooms should remain the same temperature. Overall the concept of having an air conditioning system that would maintain the same temperature throughout the house sounds great.

    The control scheme was extremely detailed and very well written. The included process & instrumentation diagram is both detailed and impressive. Based on the topics covered in this class this seems cover all the concepts quite well. The controlled, manipulated, and disturbance variables were all covered in good length. The one thing I wish this blog post addressed more closely would be the practicality of replacing current existing air conditioning system with the “better” system. For example how much would it cost to have thermometers installed in every room in an average suburban house? Personally the cost of the system would have a huge impact if I would use the product or not. Obviously it is beyond the scope of this blog post to get into the costs of the software that would be involved but it would have been nice if it was discussed upon. The only other issue I would see in this product would be the marketing of the device. Getting the consumer to understand the current problems with a central air conditioning system and how your system would fix it would be a major hurdle. After the consumer understands the issue, you would then need to convince said consumer to spend the initial costs of installing the system for the benefit of having a house maintain its temperature throughout it.

    All in all it was a well written blog post regarding a new central air conditioning system, which if cost effective, could theoretically become the “next big thing” in air conditioning. The group addressed all the major points of the project in a good succinct way.

  5. Controlling the temperature of a room is the classic example of process controls and dynamics, given the very clear controlled, manipulated, and disturbance variables. To take such a basic and common idea and attempt to optimize it is very admirable and interesting. Air-conditioning is extremely common in industrialized societies and is definitely not without its obvious flaws. The challenge in optimizing the process of air-cooling is not identifying any problems, as they are quite clear and acknowledged through day to day use. The challenge is found in the techniques used in both system analysis and control within the scope of practical engineering. Many factors must also be taken into account in problem solving, such as cost and consumer need.
    The article is very clear and straight-forward, making it easy to read and understand the problem and solution presented by this project. The focus of the control scheme was laid out explicitly as focusing on centralized air-conditioning systems of entire buildings, the largest scale of air temperature control aside from warehouses and factories. The disturbances of heat transfer are not limited to windows and exterior temperature disturbances. Focus was shined upon the temperature shifts found within the actual building and the singular air-transfer system, leading the temperature differences found between rooms and hallways. The aim of this project is to make people the most comfortable as they can by giving them a more efficient and responsive control scheme.
    The group did a fantastic job of briefly yet effectively explaining the basic process of air cooling and centralized transfer systems to different rooms. I was impressed with the energy balances derived by the group in order to take into account as many disturbance variables as possible. Taking into account the losses in between rooms as well as hallways allows the temperature changes within a single room to be evaluated more accurately for data integration into the proposed feedback loop. The temperature sensors integrated into the system provide a method of ensuring a constant temperature within each room and connecting space within the whole centralized air-conditioning system. This achievement would create an interior environment in which the desired air temperature is constantly being corrected for through measurement of deviation and continuous manipulation of airflow for a more controlled desired outcome.
    While this system has its clear advantages of comfort and control, the implementation of this feedback loop has its obstacles. Centralized air-conditions units manipulate the temperature of an entire building or edifice, which has an endless amount of possible internal disturbances which cannot always be accounted for. In order to effectively maintain a constant temperature, multiple expensive temperature sensors are needed, per room. This amounts to a total technology and hardware need that might not be worth the cost. In order to continuously be analyzing and changing a system, a lot of money is required to add onto an already expensive process scheme. In the end, this process is very useful and definitely desired, but it is up to the consumer to determine the gravity of the problems in temperature fluctuation and deviation. The significance of the heat losses might not meet up to the significance of keeping costs low, especially since comfort cannot be readily quantified. All in all, this is a well presented and extensively thought out process scheme which is backed up by theory and practical use, with the only question being the how realistic the application is.

  6. I had never realized that central air conditioning provided each room with air at the same temperature. I thought the standard systems available now worked more like the one you guys have designed, which speaks to the merits of your idea! I think it’s definitely something that could have a real commercial impact, especially if you extend the idea to central heating as well.

    You guys did an excellent job explaining your control system, and the detailed process diagram presented in figure 2 reflects the time you all put into this. You did well to explain the potential disturbances and deviations to your system, which, when it comes to heat transfer, are numerous, so I feel that your desired steady state value of within 1 oC of your set point is reasonable. Also for this reason, I think the feedback loop you’ve chosen is appropriate. It might be interesting to look into whether some sort of FF/FB loop that also accounts for potential disturbances like outside temperature change could help your controller, but this might introduce additional complexity that you are trying to avoid with your room-specific system. Although, figure 2 shows thermometers inside and outside, so if you’re making these measurements anyway, FF/FB might be reasonable, as a change in outside temp will eventually cause a change in room temp. Obviously you couldn’t use FF control on all of the disturbances, and there are a bunch, but if you notice that outside temp is your major disturbance, FF control on THAT one DV might be helpful. Something else to potentially account for is that air conditioners control both the temperature and humidity of the air in a room, so your system may benefit from inclusion of this parameter somehow.

    As far as practical implication goes, I think this system could get pretty expensive. Your second paragraph suggested that your new central air conditioning system would be a simpler model than those currently in place, but in order to control the temperature of each room individually, I think yours would be more complex. The amount of thermometers and transmitters you would need to control the temp of each room individually (especially in a large building like a high rise or a hotel) would be huge, and if individual control schemes are set up for each room, the system would get really complicated really quickly. However, I think you guys would have a real market in upper-middle class residential settings, where people are willing to shell out some more cash to have a more sophisticated heating/air conditioning set up in their homes.
    Overall, you guys have produced a clearly-written blog and a thoughtfully designed control system. Nice work!

Leave a Reply

Your email address will not be published. Required fields are marked *