3. Simulation of embedded control implementation
In this section, a theoretical analysis of the proposed methodology reveals the viability of the embedded system to numerically determine indoor comfort temperature for a real-time control system of an air-conditioning unit in practical uses. Indoor air temperature control of a single zone space with an air-conditioning system is subjectively chosen as a conventional problem of interest in this area. Fig. 4 shows a schematic diagram of modeling for a typical air-conditioning system. The indoor air temperature is to be maintained at a reference temperature. In an air-conditioning control system, a temperature controller is implemented to regulate the air-conditioning unit in order to reduce the deviation between the reference temperature and the indoor air temperature.
   In mathematical modeling, the air-conditioned space is regarded as a lumped heat capacitance. According to the principle of energy balance, the rate of change in energy stored in the air-conditioned space is equal to the summation of the following terms: the rate of heat gain through the envelope of space, the rate of cooling from the supply air, the rate of heat generation from appliances, and the rate of heat gain from infiltration and cooling loss from exfiltration, respectively.

4. Implementation of embedded system
The proposed methodology yields systematic implementation of the indoor comfort temperature to be used as the reference temperature all the time. Without this knowledge, as the surveys in Ref. [8], it is reported that the reference temperature is normally set at lower temperature in current usages of the conventional air-conditioning units. On the other hand, the reference temperature might be set at higher temperature as well under dynamic environment. In section 5, survey results from this field study confirm satisfaction of thermal comfort by most occupants. Fig. 6 presents a designed architecture of an embedded system for providing an indoor comfort temperature as a reference temperature of an air-conditioning unit. The embedded system consists of a real-time temperature controller, computing unit for determination of indoor comfort temperature. An analog-to-digital converter is implemented to obtain measurements of the outdoor air temperature and the indoor air temperature. A computing unit then determines a reference temperature according to sequential data of the outdoor temperature. The temperature controller is used to regulate the controlled variable of the conditioning unit, i.e., the temperature of the supply air in such a way that the temperature of an indoor air within the conditioned space follows a reference temperature.

5. Results and discussion
As mentioned in the previous sections, a theoretical analysis of the proposed methodology is presented in order to reveal the principles of the design for the air-conditioning control system via prototyping embedded devices. In this section, numerical simulation of an air-conditioning system within a single-zone air-conditioned space is performed virtually in MATLABTM to understand how to apply the methodology in practical use, according to the flowchart as shown in Fig. 7.
   First, how fast does the response of the indoor air temperature change with respect to time when the air-conditioning unit is switched on? Without loss of generality, the thermal load from internal heat generation or ventilation, and thermal gain from infiltration or exfiltration are excluded for this simulation. The initial temperature of the indoor air is set to be the outdoor temperature, which is fixed at 30 °C. As shown in Fig. 8, the air-conditioning unit is operated with the temperature of the supply air at 18°C. The temperature of the indoor air starts reducing from 30°C to 25°C within 40 minutes and then maintains a steady state condition provided that no disturbance of thermal load acts on the air-conditioned space.
Now, the thermal loads from appliances and infiltration listed in Table 2 take place at the air-conditioned space for a case of “without control action” as shown in Fig. 9. With the same cooling capacity simulated in Fig. 8, some amount of the cooling capacity is used to balance such unwanted thermal loads. In turn, the temperature of the indoor air cannot be brought down to the temperature of 25°C as the previous case in Fig. 8. 嵌入式系统对空调舒适温度控制的英文文献和翻译(3):http://www.youerw.com/fanyi/lunwen_3023.html