**5. TRNSYS Simulation Platform**

#### *5.1. Establishment of the Simulation Platform*

The construction of the test platform based on TRNSYS is shown in Figure 6. In this case, the simulation time is from 15th June to 15th September of the typical meteorological year, for a total of 2231 h. The building is simulated by a single area building module (Type 12) in TRNSYS. A water-water heat pump system is used in this case. Type 668 was selected as the water-water heat pump unit module. The cooling transmission in the system is achieved by the variable speed pump module Type 110. In order to advance the predicted load in advance and consider the delay caused by the flow of the central air conditioning chilled water system, the delay module (Type 93) is added to the test platform. The model is closer to the actual system and lays the foundation for subsequent feedforward control. In the test platform, the switch di fferential controller Type 2 and the Type 155 read the chilled water supply temperature of the heat pump unit and collectively control the heat pump unit operation to make it the same as the actual operation logic.

**Figure 6.** TRNSYS simulation platform.

## *5.2. Verification of Simulation Platform*

## 5.2.1. Indoor Temperature Validation Test

Based on the actual monitoring data, the weather data measured from 6th August to 11th August, 2017 were selected to verify the system simulation model. In order to reduce the influence of the initial conditions on the simulation, the actual indoor temperature on 8th August when the system is on stable operation is compared with the indoor simulated temperature with and without considering the delay effect of the system respectively. The temperature comparison results are shown in Figure 7. It can be seen from Figure 7 that the indoor temperature curve obtained by the simulation platform with delay effect is basically consistent with the actual temperature curve during the period from 9 am to 8 pm. The maximum temperature difference at a single point is 0.8 ◦C, and the average relative temperature difference is 0.4 ◦C. However, there is a large deviation between the actual room temperature and the indoor simulation temperature without considering the system delay effect. The maximum single point temperature difference is 1.9 ◦C, and the average relative temperature difference is 1.1 ◦C. As a result of this, in the simulation of the room temperature, the model considering the delay effect of the system is more in line with the actual situation. In the period from 1 am to 9 am, the reason for the error in the simulation result of the model considering the system delay effect is that the unit was set to be completely closed when the chilled water outlet temperature is lower than 8 ◦C in the simulation model in order to reduce the frequent oscillation of the system parameters during the simulation. This will result in a lower temperature of the chilled water at night, which in turn will result in lower room temperature. The simulated temperature from 4 pm to 7 pm is also low because the simplified room model only considers the effect of outdoor temperature on the indoor load and does not consider the effects of radiation. The indoor load was influenced by solar radiation and the heat storage effect of the wall, which leads to increased room temperature.

**Figure 7.** The temperature in three modes.

5.2.2. Heat Pump Unit and Pump Energy Consumption Validation Test

The energy consumption of the heat pump unit for the actual three hours based on the load rate and power of the heat pump unit during three time periods were monitored and calculated, from 9:22 to 10:22 on 7th August, 15:03 to 16:03 on 8th August, and from 13:17 to 14:17 on 11th August.

As shown in Figure 8, within three test days, the energy consumption of the heat pump unit under the consideration of time delay, without considering time delay, and actual control conditions have been marked, which shows the comparison of energy consumption within 3 h for the model unit considering the system delay effect, without considering the system delay effect, and the actual unit, respectively. It can be seen that the relative errors of the total energy consumption of the actual unit and the simulated total energy consumption of the unit with and without considering the system delay effect are 4.0% and 14.4%. This shows that a system with a delay effect is more energy-efficient than a system without a delay effect because the cooling load is input at an early time. From the above, the simulated energy consumption of the heat pump unit considering the system delay effect is more consistent with the actual energy consumption.

**Figure 8.** Comparison of simulated energy consumption and actual energy consumption of heat pump units.

The two pumps have been operating at full load, with a rated power of each pump at 3 kW. Therefore, the water pump energy consumption per hour is 6 kWh, and the actual total energy consumption of the three-hour pump is 18 kWh. Figure 9 shows the comparison of the system with and without the delay module, and the actual unit pumps energy consumption within three hours. The relative error between the actual total pump energy consumption and the simulated total energy consumption of the pump considering the delay effect of the system is 0.2%, without considering the delay effect of the system is 1.1%. The reason why the difference between both models is small is that the pump operates at a fixed frequency. Since the pump runs at a fixed frequency, the control signal cannot be controlled by the variable frequency like the control unit, and the energy consumption of the pump can only be reduced by changing the start and stop time of the pump. Therefore, the energy consumption of the system pump with or without delay effect is basically the same.

**Figure 9.** Comparison between simulated energy consumption and actual energy consumption of water pumps.
