Modeling And Control In Air Conditioning Systems PDF Download

Energy and cost-efficient management of a building's thermal properties requires heating, ventilation and air conditioning (HVAC) systems controllers to be working at optimal settings. However, many HVAC systems employ nonlinear time variances to deal with issues that affect the system's optimal operation. The present work considers an HVAC system at Memorial University's S. J. Carew Building which has been mathematically modeled using a state space multi-input and multi-output system (MIMO) approach for analyses and control system design. An IDA-ICE (Indoor Climate and Energy) simulation program has been applied for modeling the building, note that the four-story Carew Building includes an air-handling unit (AHU) on every floor. Compared with real data for one year's (2016) power consumption, the simulated annual power consumption for the building shows good agreement. Based on that data, two scenarios are applied for building the system models. Scenario 1 considers the HVAC system as a single unit with energy consumption (kWh) as inputs and zonal temperature and CO2 concentrations as outputs. By employing the MATLAB system identification toolbox, a MIMO-based system forms the basis for a state space model. In the model for Scenario 1, there are eight main AHU inputs (hot water power usage and power usage) and eight main outputs (return airflow temperature and CO2 levels). The state feedback controller obtains good results for both responses rise time and stability. In Scenario 2, there are four AHUs in total. Each of this scenario's AHUs features three main inputs (hot water, internal-to-internal air flow, and external-to-internal air flow) and three main outputs (static air pressure, CO2 levels, and temperature). In the first AHU (AHU1), we apply state-of-the-art fuzzy logic controllers (FLCs) to control fan speeds, CO2 concentrations, and temperature in the building in accordance with the flow rates for air and hot water. This strategy represents a novel approach for adapting FLCs by modifying fuzzy rule using the Simulink. The modified system shows improved levels of thermal comfort. The final part of the work presents the design for a supervisor fuzzy logic controller (SFLC) that can be applied to the entire S. J. Carew Building HVAC control. This SFLC features 24 inputs and 12 outputs and employs a state-space model that considers each AHU as an individual system. The SFLC detailed design and system simulation results are presented in this thesis.