**6. Results and Discussion**

Based on the test platform, the indoor temperature and total system energy consumption under the three control strategies (LFF, RTF, and SWT) were simulated. The TRNSYS simulation platform of the LFF control strategy is shown in Figure 10.

**Figure 10.** TRNSYS simulation platform of the LFF control strategy.

#### *6.1. Simulation Results of Indoor Temperature*

The room temperatures of the three control strategies were compared under both full-time simulation and hottest day simulation conditions. The temperature results of LFF, RTF, and SWT control strategies for the full-time simulation correspond to Figure 11a–c, respectively. As can be seen from Figure 11, three control strategies can control the room temperature between 25 ◦C and 27 ◦C when the air conditioning system is in operation. Indoor temperatures above 27 ◦C occur at night when the outdoor temperature is high.

**Figure 11.** (**a**) Room temperature simulation results under LFF control; (**b**) Room temperature simulation results under RTF control; (**c**) Room temperature simulation results under SWT control.

During the entire simulation, the outdoor temperature was the highest on 19th June, so the simulation results of 19th June were selected to analyze the indoor temperature control of the HVAC system under the limit conditions. The room temperature comparison under different control strategies on 19th June is shown in Figure 12.

**Figure 12.** Comparison of room temperature under different control strategies on 19th June.

All three control strategies can control the room temperature within the operating period of the air conditioning system between 25 ◦C and 27 ◦C. Three kinds of control of the temperature range are acceptable. During the 0:00 to 7:00 and 20:00 to 23:00 periods which the air-conditioning chilled water system is not in operation, the indoor room temperature under the three control strategies is almost the same regardless of the trend or value. However, in the period of 7:00 to 20:00 when the air-conditioning chilled water system is in operation, there are obvious differences in the indoor room temperature control effect under the three control strategies. The room temperature under SWT control fluctuates the most. The room temperature reaches a maximum of 27 ◦C at 10:00. At this time, the temperature of the external wall will rise due to the increase of the outdoor temperature, thus making the indoor temperature rise. The room temperature reached a minimum of 25 ◦C at 19:00 because the chilled water temperature of the unit was turned off at 8 ◦C, which in turn caused the indoor temperature to drop. The room temperature under the RTF control fluctuated relatively less with a fluctuation range of 25.4 ◦C and 26.8 ◦C. The RTF control strategy is based on the indoor temperature as the control object, and the fuzzy temperature of the set temperature range is solved to ensure the thermal comfort of the indoor temperature. However, with the LFF control strategy of predictive load feed-forward control, the required load of the system is given from the demand-side response, which can ensure that the system operating temperature just matches the room temperature demand. Therefore, the room temperature fluctuation range under the LFF control strategy is between 25.4 ◦C and 26.4 ◦C, and among the three control strategies, the temperature fluctuation range is the smallest.

#### *6.2. Simulation Results of Energy Consumption*

The total hourly energy consumption of pumps and units under all three control strategies are shown in Figure 13.

The comparison of energy consumption under the three control strategies for the system is shown in Figure 14. It can be intuitively seen from Figure 14 that the LFF control strategy consumes the least amount of energy, whether it is the total energy consumption of the system or the energy consumption of the unit and the pump. The energy consumption of the system under the LFF control strategy is the least because the system is controlled from the demand- side, the system is feed-forward input according to the predicted load advance time delay, and the operation of the pumps and units is fuzzy controlled. Demand-side control can fundamentally solve problems such as the mismatch between system energy consumption and required cooling capacity, and avoid unnecessary startup and overload operation of the unit. Compared with the LFF control strategy, the RTF control strategy, as a feedback control, has a significantly better control effect than the actual control strategy based on the temperature difference between the supply and return water. However, due to problems such as time delay in feedback control, it is impossible to determine the operation of the system from the demand-side feedforward control, which will cause certain unnecessary energy loss in the system. Table 4 shows the energy consumption comparison between LFF and the other two control strategies (SWT, RTF), respectively. Compared to the SWT control strategy, the LFF control strategy has a heat pump unit energy-saving rate of 14.5%, a pump energy-saving rate of 10.2%, and a total energy-saving rate of 13.4%. Compared with the RTF control strategy, the LFF control strategy has a relatively low energy-saving rate of 9.2%, 4.1%, and 7.8%, respectively.

**Figure 13.** The simulation results of energy consumption under three different control strategy.

**Figure 14.** Comparison of energy consumption under three conditions.


**Table 4.** Comparison of energy-saving rates of LFF control strategies compared to the other two control strategies.
