Computational Studies for Smart Building Applications: Localization, Car Parking and Healthcare

智能建築應用的計算研究:定位,泊車和醫療

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date22 Jan 2021

Abstract

Smart building is realized as building that supports various business functions by means of remotely or locally collecting, preprocessing and generating data. Among smart buildings, sensors, microprocessors and actuators are utilized to aggregate and manipulate data pertinent to the nature of services or business, thereby bringing social benefits. In recent years, smart building technology has developed rapidly, from personal applications to industrial applications. A tremendous number of sensors have been deployed for various purposes, such as crowd control and air quality monitoring. These sensors require well-designed algorithms, schemes, and configurations to fulfill the requirements of different applications and to deliver the envisaged services. The development of smart building systems encounters various deficiencies, such as lack of systematic design tools and diverse system requirements. Evolutionary techniques are thus needed to enhance the performance of conventional system designs and to customize the systems to fulfill the requirements of smart building applications. With these concerns in mind, this thesis addresses three smart building use cases to demonstrate the feasibility of exploiting evolutionary techniques in smart building development. Crowd management and healthcare are two essential smart building services. They can be applied in most residential and commercial buildings, such as object tracking, navigation, and remote health monitoring. Indoor localization, car parking, and healthcare are three crucial applications, and they can be generally found in most modern buildings. After analyzing the deficiencies of these applications, this thesis has developed three analytical works to overcome the deficiencies and make the applications smart, thereby enhancing citizens’ living quality.

Indoor localization techniques can be used in various contexts, such as surveillance and asset tracking. In this thesis, a new indoor localization scheme is developed with primary objectives of improving localization accuracy and reducing computational time. This scheme comprises a newly developed algorithm, namely precise-fast convergence particle swarm optimization (P-FCPSO), to analyze the data of link quality indicator (LQI) and received signal strength (RSS) and produce localization results. Its performance is better than that of existing schemes. In addition, a stochastic area clustering (SAC) method is designed to increase the accuracy of indoor localization. Stochastic cluster factoring (SCF) is also developed to shorten the processing time of the PSO-based indoor localization.

Cruising for parking spaces is a severe problem for modern buildings since it may cause traffic congestion around buildings. An array of sensors is installed on buildings to measure the traffic conditions nearby. Advanced applications can then be developed by exploiting these measured data. A new parking cruising-prevention (PCP) scheme is developed to help users locate the optimal parking lots with the minimized total journey time and cost. A new parking model (PM) is then proposed. In the PM, a new vehicle-queueing model at road intersections is also derived, as is a new resource model for total expenditure estimation during the car-parking process. The objective of the developed model is to decrease the time spent on cruising to search for a suitable parking lot.

Healthcare is a ubiquitous theme in industrial research. Seamless access to healthcare data in buildings is a particularly important issue. With this issue in mind, a ZigBee-based feasibility test using a wireless sensor network is proposed and implemented. The packet error rate (PER) of the network is the most critical issue. Additionally, indoor air quality (IAQ) is analyzed, and a novel IAQ control scheme, referred to as a fuzzy genetic multi-layered control scheme (FGMLCS), is developed to reduce the concentration of volatile organic compounds (VOCs) in indoor air. The objective of the scheme is to deliver a comfortable and healthy environment for people in smart buildings.