Cooperative Control of VAV Air-Conditioning Systems - Escholarship
Shuai Liu,Baocang Ding,Wenjian Cai
2012-01-01
Abstract:Cooperative control of VAV air-conditioning systems Shuai Liu 1 , Lihua Xie 1 , Wenjian Cai 1 1. EXQUISITUS, Centre for E-City, School of Electrical and Electronics Engineering, Nanyang Technological University, Singapore E-mail: {LIUS0025, elhxie, ewjcai}@ntu.edu.sg Abstract: A dynamic model for a heating ventilating air-conditioning (HVAC) system is developed. Assuming the power of air supply fan is fixed, the air pressure in different rooms and air transmission ducts can be controlled by adjusting variable air volume (VAV) box and its terminals and air diffusers. Since the inlet air pressures in different rooms are interactive, traditional VAV controller, which is to control the dampers individually, will unavoidably generate a large overshoot. In this paper, we consider a novel VAV control strategy which is to control the air diffusers and VAV dampers in a cooperative way. This consensus-based approach can lead to a smoother air pressure change comparing with the traditional approaches. The simulation validates the effectiveness of the new control strategy. Key Words: VAV, HVAC, Cooperative control, Consensus 1 Introduction Energy saving and environmental protection problem has become one of the critical concerns of human being. The world’s growing energy demands require developing a sus- tainable living plan. The largest sector of energy consump- tion in a modern city is buildings. Usually, a great portion of the energy consumption of buildings is attributed to heating and ventilating air-conditioning (HVAC) systems. In fact, a lot of energy is wasted in HVAC systems. Therefore, the controller design of HVAC systems is of great significance. There are two classes of models for HVAC systems: static models for real-time optimization and dynamic models for stabilization control. Most of the existing works are based on static models. [1] is mainly focused on developing a sim- plified model of cooling coils. The model parameters are determined on-line based on commission or catalog informa- tion by linear or nonlinear least squares methods. For over- all HVAC systems, [2] and [3] formulate the minimization of the total power consumption, which is mainly caused by chillers, pumps and fans, as a global optimization problem. A modified genetic algorithm was used to set the optimal operating point of each component. When we consider the dynamic properties of HVAC systems, the control problem becomes more difficult. A new dynamic simulation model for air-handling unit (AHU) is developed by [4]. The param- eters can be easily determined from the measurement of the total fan energy consumption. Since dynamic models can describe the dynamic changing at different time in a day, a dynamic control strategy would often lead to a better per- formance comparing with static control strategy. By using Kalman filtering, [5] presents a temperature prediction al- gorithm based on a simple time varying zone model. The optimal performance is reached by applying a genetic algo- rithm. Other than minimizing energy consumption while keep- ing thermal comfort, there are some works focusing on ther- mal comfort while keeping low energy consumption. The temperature can be affected by adjusting the cool air flow rate into rooms, which is controlled by dampers. There are two temperature control strategies: pressure dependent con- trol and pressure independent control. Pressure dependent control means to control the damper based on the room tem- perature. According to the temperature difference between the room temperature and the desired one, we can adjust the air flow by controlling the damper. The reader can refer to [6], [7], [8], [9] and [10] for detail. Pressure independent control strategy contains two con- trol loops. The inner loop is for temperature control which can provide set point for outer loop based on the temperature difference. The outer loop is for air flow control. Pressure in- dependent control leads to better control performance. This is because a pressure dependent controller would take ac- tion when the air flow changes affect the room temperature. Therefore the response of the system suffers from a time de- lay comparing with systems with pressure independent con- trollers. Keeping the static pressure of the room inlet to be constant is the key step to air flow control. In this paper, we consider the static pressure control problem by cooperatively adjusting the VAV box damper and its terminals. Comparing with the traditional PID controller, the cooperative controller can reduce the pressure differences among each room inlet while tracking the desired pressure, which makes the whole tracking process smoother. A numerical example is given to demonstrate our cooperative control scheme. Preliminaries in Graph Theory A directed graph is denoted by G = {V, E G , A G }, where V = {1, 2, . . . , N } is the set of node with i representing the ith agent; E G is the set of edges which are represented by a pair of node indices (i, j). We consider that (i, j) ∈ E if and only if node i can send its information to node j. In this case, node i is called the parent node and node j is called a child node. The set of neighbors of the ith agent is denoted by N i = {j ∈ V | (j, i) ∈ E}. If (i, i) ∈ E, we say that node i has self-loop. In this paper, we assume that no self-loop exists. A G = [a i,j ] ∈ R N ×N is the the adjacency matrix associated with G. If j ∈ N i , a i,j > 0, otherwise a i,j = 0. If matrix A is symmetric, then the corresponding graph is called undirected graph. A graph is balanced if the in-degree deg in (i) = and the out-degree deg out (i) j∈V\{i} a i,j For example, j∈V\{i} a j,i are equal for all i ∈ V. an undirected graph is a kind of balanced graph.