A method and system for controlling indoor temperature of a building based on model predictive control

Abstract

The invention discloses a method and system for controlling indoor temperature of a building based on model predictive control, and relates to the field of indoor environment control. First, an indoor temperature prediction model is established to predict the room temperature. On this basis, a cost function is established, which optimizes the cooling capacity of the air-conditioning system to reduce the system energy consumption while ensuring the indoor temperature. The particle swarm algorithm is used to roll and optimize the function to minimize the value in the prediction time domain, and a control sequence of the cooling capacity is generated. At the current moment, each device is controlled according to the first value of the sequence to minimize the overshoot and ensure the indoor temperature in the control time domain. At the next moment, the indoor temperature is fed back to correct the deviation of the prediction model, the prediction accuracy is ensured, and the rolling optimization is repeated to enhance the stability of the system. The invention is different from the current feedback control indoor temperature method based on system temperature difference or pressure difference, reduces the energy consumption of the air-conditioning system under the premise of improving indoor thermal comfort, and has good control quality for both indoor temperature and cooling capacity.

Description

Technical Field

[0001] The present invention relates to the field of indoor environment control, and more specifically, to a method and system for controlling indoor temperature of a building based on model predictive control. Background Art

[0002] According to statistics from energy agencies, building energy consumption accounts for 40% of total energy consumption, and the energy consumption of air conditioning systems is the main part of building energy consumption, accounting for about 50% of total building energy consumption. As people's requirements for building comfort and energy saving are increasing, how to control the air conditioning system to achieve the maximum energy saving while achieving thermal comfort is becoming more and more important.

[0003] Traditional PID feedback control strategies have disadvantages such as large overshoot, slow adjustment speed and poor stability, and cannot balance comfort and energy consumption. Under the premise of ensuring human comfort in the future, the required cooling / heating and control strategies are also changing. By rolling optimization of control variables, the goal of reducing energy consumption while meeting comfort can be achieved. Summary of the invention

[0004] The present invention proposes a method and system for controlling indoor temperature of a building based on model predictive control, which is theoretically different from the currently commonly used PID feedback control method. It realizes dynamic matching of air-conditioning supply / cooling heat and building load while meeting indoor thermal comfort conditions, achieves maximum system energy-saving effect, and realizes optimal control of air-conditioning system equipment.

[0005] The specific implementation steps of a building indoor temperature control method and system based on model predictive control proposed by the present invention are as follows:

[0006] Step 1: Establish a building indoor temperature prediction model based on the historical data of the air-conditioning system, and calculate the predicted indoor temperature value at future times under different cooling capacity and weather forecast data conditions.

[0007] y*(k+1)=a×u(k)+b×Q(k)+c×y(k)

[0008] Among them, y * (k+1) is the predicted value of indoor temperature at time k+1, a, b, c are the parameters to be identified, u(k) is the cooling capacity at time k, Q(k) is the building load under the meteorological data at time k, and y(k) is the indoor temperature at time k.

[0009] Step 2: Based on the indoor temperature prediction model, the cost function J(k) is established by collecting the performance parameters of the air-conditioning system equipment. This function optimizes the cooling capacity of the air-conditioning system as much as possible to reduce the system energy consumption while ensuring the indoor temperature.

[0010]

[0011] Among them, N is the prediction time domain which can be set by yourself, q is the temperature error weight coefficient, y set Set the temperature for the indoor temperature.

[0012] The first term on the right side represents the cost of temperature output error, which forces the air conditioning system output to be as close to the indoor temperature set point as possible; the second term on the right side represents the cost of the change in the cooling capacity control variable, which smoothes the control variable as much as possible to reduce system energy consumption.

[0013] Step 3: In each control cycle, the particle swarm algorithm is used to predict the indoor temperature and optimize the cost function in a rolling manner, so that the objective function J(k) is minimized within the prediction time domain N. Under the constraints of the equipment performance parameters, a control sequence of the cooling capacity setting value is generated, and the first u value is output for execution.

[0014] Step 4: The field controller optimizes and controls each device of the air-conditioning system according to the set value, and adjusts the cooling capacity of the air-conditioning system to the set value, thereby minimizing the overshoot and adjusting the indoor temperature within the control time domain.

[0015] Step 5: Collect the indoor temperature of the building in real time for feedback correction to correct the deviation of the prediction model, ensure the accuracy of the prediction results and enhance the stability of the control system.

[0016] In general, compared with the prior art, the above technical scheme conceived by the present invention performs model predictive control on the indoor temperature of the building. Compared with the fixed water supply temperature and fixed pressure difference control, the indoor temperature can be directly controlled. Compared with the traditional PID feedback control, the indoor temperature model predictive control has a feedforward control characteristic with small overshoot and no steady-state error. It is suitable for large-delay complex air-conditioning systems, has good temperature and energy consumption control quality, and has obvious advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The invention discloses a method for controlling indoor temperature of a building based on model predictive control and a system flow chart.

[0018] Figure 2 It is a system diagram of an embodiment of the present invention.

[0019] Figure 3 This is a comparison diagram of the indoor temperature control effect of the embodiment of the present invention and PID simulation

[0020] Figure 4 This is a comparison chart of the control effect of the embodiment of the present invention and the PID indoor temperature experiment

[0021] Figure 5This is the embodiment of the present invention and the PID energy consumption simulation effect diagram

[0022] Figure 6 This is the effect diagram of the embodiment of the present invention and the PID energy consumption experiment DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0024] See attached Figure 1 The present invention provides a method and system for controlling indoor temperature of a building based on model predictive control, which mainly comprises the following steps:

[0025] (1) Collect historical data of the air conditioning system, including supply and return water temperature, flow, indoor temperature, building structure parameters, historical outdoor dry-bulb temperature and relative humidity data. Use the supply and return water temperature and flow to calculate the cooling capacity at different times, and use the outdoor dry-bulb temperature, relative humidity, solar radiation illumination and building structure parameters to calculate the building cooling load at different times. Finally, form training sample data with the indoor temperature at the corresponding time, and use the genetic algorithm to train the following formula.

[0026] y*(k+1)=a×u(k)+b×Q(k)+c×y(k)

[0027] Among them, y * (k+1) is the predicted value of indoor temperature at time k+1, a, b, c are the parameters to be identified, u(k) is the cooling capacity at time k, Q(k) is the building load under the meteorological data at time k, and y(k) is the indoor temperature at time k.

[0028] Finally, the prediction model after training in this embodiment is as follows.

[0029]

[0030] Among them, C p , V, ρ represents the specific heat capacity, volume and density of the building, Q represents the cooling load, the ΔT control step is 1 hour, y* represents the predicted indoor temperature, and y represents the real-time indoor temperature.

[0031] (2) Based on the indoor temperature prediction model, the performance parameters of the air-conditioning system equipment are collected, mainly the cooling capacity of each chiller. A constrained cost function J(k) is established, which optimizes the cooling capacity of the air-conditioning system as much as possible to reduce the system energy consumption while ensuring the indoor temperature.

[0032] The cost function is:

[0033] The air conditioner operation time is from 8:00 to 19:00, the prediction time domain N = 12, and the indoor temperature y set =25°C q=[10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 2].

[0034] The constraints are: min ≤u(k+i)≤u max , i = 0, 1, ..., N-1

[0035] where u min and u max are the minimum and maximum cooling capacities of the air-conditioning system respectively, which in this case are 0kW and 40kW.

[0036] (3) In each control cycle, the particle swarm algorithm is used to predict the indoor temperature and optimize the cost function in a rolling manner, so that the cost function J(k) is minimized in the prediction time domain N, and a control sequence U of the cooling capacity setting value is generated, and the first value u(k) is output for execution. The objective function of the particle swarm algorithm rolling optimization in this embodiment is:

[0037]

[0038] (4) The air conditioning system optimizes and controls each unit according to the set value. First, the number of chillers to be turned on is determined according to u(k). Second, the particle swarm algorithm is used to calculate the optimal supply and return water temperature and flow rate under the cooling capacity with energy consumption as the target. Finally, the field controller starts and stops the unit, sets the supply and return water temperature of chilled water, and controls the electric regulating valve and water pump to adjust the system flow rate, so as to minimize the overshoot and meet the indoor temperature within the control time domain.

[0039] (5) At the beginning of the next moment, the indoor temperature of the building is collected for feedback correction to correct the deviation of the prediction model, and the optimization is repeated again. This cycle ensures the accuracy of the prediction results and enhances the stability of the control system.

[0040] e(k+1)=y(k+1)-y * (k+1)

[0041] Among them, y(k+1) is the real value of the indoor temperature at time k+1, and y* is the predicted value at time k+1. This deviation e is used to correct the prediction for future moments.

[0042] Y*(k+1)=y*(k+1)+e(k+1)

[0043] Among them, Y*(k+1) is the corrected value of the predicted indoor temperature at time k+1.

[0044] The simulation results are compared with Figure 3 and Figure 5 As shown in the figure, the indoor temperature of MPC reaches the set point of 25℃ 1 hour earlier than PID; MPC has almost no overshoot phenomenon, while PID has a large overshoot and its stability is worse than MPC. The average indoor temperatures of MPC and PID are 25.1℃ and 25.3℃ respectively. MPC saves 15% energy in cooling tower, 16% energy in water pump, 13.9% energy in chiller and 14.8% energy in total energy consumption. Figure 4 and Figure 6 As shown in the figure, the indoor temperature of MPC reaches the set point of 25℃ about 1.5 hours earlier than PID. There is almost no overshoot in MPC, while there is a 24.2℃ overcooling overshoot in PID. Compared with the PID control strategy, MPC saves 8.9% of energy in cooling tower, 7.3% of energy in water pump, 8.6% of energy in chiller, and 8.1% of total energy consumption. Therefore, simulation and experiments show that the indoor temperature and energy consumption control effects of MPC are better than those of traditional PID strategy.

[0045] To make it easier for those skilled in the art to understand, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method and system for controlling indoor temperature of a building based on model predictive control, characterized in that The steps include: Step (1), based on the historical data of the air conditioning system, a building indoor temperature prediction model is established to calculate the indoor temperature prediction value at a future time under different cooling capacity and weather forecast data conditions; the indoor temperature prediction model is established using a mechanism method, and the formula is as follows: Among them, y * (k+1) is the predicted value of indoor temperature at time k+1, a, b, c are the parameters to be identified, u(k) is the cooling capacity at time k, Q(k) is the building load under the meteorological data at time k, and y(k) is the indoor temperature at time k; Step (2), based on the indoor temperature prediction model and the performance parameters of the air conditioning system equipment, a cost function J(k) is established, which optimizes the cooling capacity of the air conditioning system as much as possible to reduce the system energy consumption while ensuring the indoor temperature; Step (3), in each control cycle, the indoor temperature is predicted and the cost function is optimized by using an intelligent algorithm to minimize the objective function J(k) in the prediction time domain. Under the constraints of the equipment performance parameters, a control sequence of the cooling capacity setting value is generated and the first value is output for execution; Step (4), the field controller optimizes and controls each device of the air-conditioning system according to the set value, adjusts the cooling capacity of the air-conditioning system to the set value, and ensures that the overshoot is minimized and the indoor temperature is adjusted within the control time domain; Step (5) collects the indoor temperature of the building in real time for feedback correction to correct the deviation of the prediction model, ensure the accuracy of the prediction results and enhance the stability of the control system.

2. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: The air conditioning system historical data includes supply and return water temperature, flow, indoor temperature, building structure parameters, outdoor dry bulb temperature, relative humidity data and solar radiation illumination.

3. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: The weather forecast data includes outdoor dry-bulb temperature, relative humidity and solar radiation illumination.

4. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: The feedback correction includes real-time building indoor temperature, outdoor dry-bulb temperature, relative humidity and total solar radiation illumination.

5. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: Genetic algorithm and historical data of air conditioning system are used to identify model parameters a, b, and c, and the accuracy of the prediction model is within ±10%.

6. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: The equipment parameters include the cooling / heating capacity, fixed / variable frequency and input power of the cold and hot source units, and the head, flow rate and input power of each water pump.

7. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: The cost function J(k) is established as follows: J(k)=\sum ^{N}_{i=1} {q(k+i)}[y*(k+i)-{y}_{set}(k+i){]}^{2}+\sum ^{N-1}_{i=1} {[u(k+i)}-u(k+i-1){]}^{2} Among them, N is the prediction time domain which can be set by yourself, q is the temperature error weight coefficient, y set Set the temperature for the indoor temperature; The first term on the right side represents the cost of temperature output error, which forces the air conditioning system output to be as close to the indoor temperature set point as possible; the second term on the right side represents the cost of the change in the cooling capacity control variable, which smoothes the control variable as much as possible to reduce system energy consumption.

8. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: The specific implementation of step (3) is as follows: At time k, the particle swarm algorithm is used to globally optimize the cost function J(k) for the next N time periods, and the minimum value of J(k) is calculated. The optimal variable in J(k) is the cooling capacity u, and the first u(k) in the u sequence is output and executed at time k.

9. A method and system for controlling indoor temperature of a building based on model predictive control as claimed in claim 1, characterized in that: The specific implementation of step (4) is as follows: The operating parameters of the current air-conditioning system are optimized according to the cooling capacity set value u(k). The number of chillers to be turned on under this cooling capacity set value is determined mainly based on energy consumption. The particle swarm algorithm is used to optimize the supply and return water temperature set values ​​and the chilled water flow rate to ensure the lowest operating energy consumption. Based on the optimized parameters, the field actuators mainly start and stop the units and set the supply or return water temperature of the units. The DDC controller adjusts the cooling capacity by controlling the bypass flow of the electric regulating valve and the water pump frequency.

Images