پیش بینی سرعت فن برای صرفه جویی در انرژی در سیستم HVAC بر اساس سیستم استنتاج فازی مبتنی بر شبکه تطبیقی

In this paper, a HVAC (heating, ventilating and air-conditioning) system has two different zones was designed and fan motor speed to minimize energy consumption of the HVAC system was controlled by a conventional (proportional–integral-derivative) PID controller. The desired temperatures were realized by variable flow-rate by considering the ambient temperature for each zone. The control algorithm was transformed for a programmable logic controller (PLC). The realized system has been controlled by PLC used PID control algorithm. The input–output data set of the HVAC system were first stored and than these data sets were used to predict the fan motor speed based on adaptive network based fuzzy inference system (ANFIS). In simulations, root-mean-square (RMS) and the coefficient of multiple determinations (R2) as two performance measures were obtained to compare the predicted and actual values for model validation. All simulations have shown that the proposed method is more effective and controls the systems quite well.

The comfort of the people in their living environment is partially dependent on the quality and temperature of air in their building. Three interrelated systems are used to provide the desired air temperature and quality. These are the ventilating system, the heating system and the air conditioning system. The purpose of HVAC system of a building is to provide complete thermal comfort for its occupants. On the other hand, energy saving in this system is one of the most important issue because of its cost. Hence, it is necessary to understand the aspects of minimum energy consumption in order to design an effective HVAC system. Teitel, Levi, Zhao, Barak, Bar-lev and Shmuel have employed for variable frequency drives method which is routinely used to vary pump and fan motor speed in heating, ventilating and air conditioning of buildings (Teitel et al., 2008). In these applications, speed control is used to regulate the flow of water or air because speed adjustment is an energy efficient method of flow control. The aim of this study is to present a thermodynamic model for an air-cooled centrifugal chiller which is developed specifically to analyze how the speed control of the condenser fans influences the chiller’s COP under various operating conditions (Yu and Chan, 2006, Yu and Chan, 2007 and Yu and Chan, 2007). Moreover, the other study of the same authors is made to investigate how the use of variable speed condenser fans enables air-cooled chillers to operate more efficiently (Yu and Chan, 2006, Yu and Chan, 2007 and Yu and Chan, 2007). Besides, variable fan speed control is increasingly used for chiller compressors to save power when chillers are operating at part load. The power saving comes from the improved efficiency of the motors when operating at a lower speed under part-load conditions (Aprea et al., 2004 and Tassou and Quereshi, 1998). HVAC systems require control of environmental variables such as pressure, temperature, humidity. Furthermore, HVAC system is necessary to implement a realistic thermal environment in terms of temperature and air flow rate in the space of virtual reality (Shin, Chang, & Kim, 2002; Kaynakli, Pulat, and Kilic, 2005). As in other industrial applications, most of the controllers commissioned in HVAC systems are of the proportional–integral-derivative (PID) type (Bi et al., 2000 and Seem, 1998). This is mainly because PID is simple yet sufficient for most HVAC application specifications. However, tuning a PID controller requires an accurate model of a process and an effective controller design rule. In addition to that, the tuning procedure can be a time-consuming, expensive and difficult task (Bi et al., 2000, Krakow and Lin, 1995, Pinnella et al., 1986 and Riverol and Pilipovik, 2005). The developments in intelligent methods make them possible to use in nonlinear analysis and control. Intelligent methods were first used to increase the robustness of existing models however they have been used to obtain new models in recent years. In addition to PID control of HVAC systems, the various studies using intelligent methods were presented. A neural network (NN) model was developed to predict air pressure coefficients across the openings in a light weight, single sided, naturally ventilated test room (Kalogirou, Eftekhari, & Marjanovic, 2003). The applicability of natural ventilation as a passive cooling system was investigated in modern buildings in Kayseri using model simulations of indoor air velocity by the fluent. Using the simulated data, an ANFIS model was employed to predict the indoor average and maximum air velocities (Ayata, Çam, & Yıldız, 2007). A systematic approach for optimal set point control for in-building section was presented (Lu, Wenjian, Lihua, Shujiang, & Chai, 2005). The major components of in-building section were analyzed to identify the energy conservation potential. In order to save energy for delivery of supply air and chilled water, a variable pressure set point method was analyzed by a simple example and an intelligent neural network model-ANFIS was proposed to compute the variable pressure set points influenced by variation of cooling loads of end users. In addition, fuzzy logic control (FLC) of HVAC systems was studied by many authors (Huang & Nelson, 1994). The obtained results were compared with those of PID control and these studies indicated that FLC had better results. FLC is extensively used in processes where systems dynamics is either very complex or exhibit highly nonlinear characters. Besides, FLC is one of useful control schemes for plants having difficulties in deriving mathematical models or having performance limitations with conventional linear control methods. FLC is designed on the basis of human experience that means a mathematical model is not required for controlling a system. Because of this advantage, fuzzy logic-based control schemes were implemented for many industrial applications (Hung, Lin & Chung; 2007; Tang & Mulholland, 1987). FLCs were successfully applied to many complex industrial processes and domestic appliances in recent years (Tsang, 2001). The first FLC algorithm implemented by Mamdani was designed to synthesize the linguistic control protocol of an experienced operator (Mamdani, 1974). Consequently, different control tuning methodologies have been proposed in the literature such as auto-tuning, self-tuning, artificial intelligence, and optimization methods. In this study, based on considering the above literatures, the required fan motor speed to minimize energy consumption and the required damper gap rates for obtaining the desired temperatures of two different zones for each time step were found by using PID control algorithm. The damper gap rate is also proportional with air flow rate. Besides, in this study, an expert system for fan motor speed and air flow control of HVAC system based on ANFIS is presented. In simulations, root-mean-square (RMS) and the coefficient of multiple determinations (R2) as two performance measures are given to compare the predicted and actual values for model validation. All simulations have shown that the proposed method is more effective and controls the systems quite well. The outline of the paper is as follows. In Section 2, the model of the HVAC system is presented. The design of the considered real-time HVAC system is given Section 3. Section 4 briefly describes the adaptive network based fuzzy inference system (ANFIS). Then, in Section 5, the experimental results are presented. In the experiment, the fan motor speed and the damper gap rate being proportional with air flow rate has been controlled using PID controllers and ANFIS. Finally, conclusions are given in Section 6.

In this study, the cooling process of the system was realized by being cooled the two different zones from the ambient temperature 26.4 °C to the desired temperatures. The required damper gap rates for obtaining the desired temperatures of two different zones and the required fan motor speed to minimize energy consumption for each time step were found by using PID control algorithm. As seen in Fig. 5, Fig. 6, Fig. 7, Fig. 9, Fig. 11 and Fig. 13, the experimental results have been presented in graphical form by using MATLAB Graphical Toolbox. In this work, the fan motor speed and the damper gap rates of a HVAC system with two zones were predicted by using ANFIS method. To assess the effectiveness of our proposal ANFIS, three computer simulation were developed on the MATLAB environment. The ANFIS results were given in the related tables. The simulation results have shown that the ANFIS can be used as an alternative prediction and control method for HVAC systems. In statistical analysis, the RMS value is 3.3475 and the R2 value is 0.9954 for evaporator while the RMS value is 15.6750 and the R2 value is 0.9402 for zone-1 and the RMS value is 17.7019 and the R2 value is 0.9410 for zone-2 for the ANFIS model. This paper shows that the values predicted with the ANFIS can be used to predict fan motor speed and damper gap rate of HVAC system quite accurately. Therefore, faster and simpler solutions can be obtained based on ANFIS.