Indoor temperature and humidity control using generalized predictive control-fuzzy cognitive map controller on direct expansion air conditioning system

Nowadays, the application of different controllers on heating, ventilating and airconditioning system (HVACs) are considered as an important issue in order to improve the performance of the system, due to the high demand of these appliances in the buildings and their high energy consumptions in the buildings. Direct expansion air conditioning system (DX A/C) is mostly used in the small to medium size buildings in tropical regions. The DX A/C system is nonlinear, Multiple-Input and Multiple-Output (MIMO) and inherently complex system with strong cross coupling effect between supply air temperature and supply air humidity. The previous researches shows that designing the nonlinear controllers are limited and difficult due to the complexity and uncertainty of the system, and complex mathematical analysis in finding a Lyapunov function. On the other hand, for making the control design easy, the MIMO structure of the system are considered as Single- Input and Single-Output (SISO) system by decoupling the system. In order to consider the coupling effects, MIMO control strategies are required. But, these strategies mostly are applied to the linearized model of the system around operating point and makes the working range of the controller limited to the neighborhood of operating range. For full control of the system, the wider operating range is required. Therefore, the goals for designing the suitable controller on DX A/C system are designing MIMO nonlinear controller by easy mathematic and structure. The simple Fuzzy Cognitive Map (FCM) control algorithm by using generalized predictive control (GPC) for assigning the weights are used to obtain the goals of comfort and energy saving by considering the real characteristics of air conditioning system. The performance analysis of the designed controller was tested by set point tracking test and disturbance rejection test. The results for both tests showed that by changing the compressor and supply fan’s speed, the proposed controller successfully can be implemented to the DX A/C system. Also, the controller work successfully in wider operating range in other set points (22-26 oC). The GPC-FCM controller are compared by LQG controller in different conditions and the results shows the better performance of GCP-FCM controller in comparison with LQG one. The achievements of this research are a new design approach to MIMO nonlinear controller for DX A/C system to stabilize the humidity and temperature of the air conditioned room on desired set points, integration of different control categories in single control scenario by soft computing methodology to response all the requirements of the system, introducing new platform based on the Generalized Predictive Control- Fuzzy Cognitive Map control method for the first time in the literature about HVAC systems, new development in nonlinear control systems with simple mathematics, new solution for approaching to MIMO system with coupling effect without linearization of the model due to a simple structure of FCM, energy saving and energy efficiency by this new control design. In conclusion, by employing the GPC-FCM controller on the DX A/C system a soft, intelligent, hybrid, nonlinear and MIMO control method is obtained. Decreasing the energy usage of the air conditioning system are achieved by using the variable speed supply fan and variable speed compressor and applying the hybrid GPC-FCM Control design for preventing from losing energy by making the controller errors as least as possible.

Preview Text FK 2017 127 IR.pdf Download (1MB) | Preview

Actions (login required)

View Item