Tableau d'honneur de la Faculté des études supérieures et postdoctorales, 2006-2007 / Un système de tenségrité correspond à un assemblage de composants qui sont chargés de manière axiale. Ce faisant, des câbles ou des ressorts peuvent être utilisés pour les composants en tension ce qui réduit la masse et l’inertie du système. Par conséquent, des mécanismes de tenségrité sont introduits dans cette thèse comme des alternatives aux mécanismes plus conventionnels pour certains types d’application. Les objectifs principaux de la thèse sont le développement et l’analyse de nouveaux mécanismes. La nécessité pour les mécanismes de tenségrité de toujours se retrouver dans des configurations où leurs câbles et leurs ressorts sont soumis à des forces de tension complique passablement leur développement. Par conséquent, une approche novatrice fondée sur l’utilisation de ressorts est proposée pour surmonter cette difficulté. Pour faciliter l’utilisation de cette approche, des règles qui s’appliquent à la quantité de ressorts utilisée dans une architecture sont présentées. À partir de ces règles et en s’inspirant de systèmes de tenségrité existants, deux mécanismes plans, trois mécanismes spatiaux et deux mécanismes modulaires sont développés. Chaque mécanisme est modélisé dans le but d’analyser sa cinématique, sa statique et sa dynamique. Étant donné la présence de degrés de liberté non contraints dans les architectures des mécanismes, les relations entre leurs variables d’entrée et de sortie dépendent des chargements externes, gravitationnels et inertiels qui leur sont appliqués. En supposant un régime quasi-statique, de telles relations peuvent être calculées avec une approche numérique. Toutefois, pour le cas spécifique où les mécanismes ne sont pas soumis à des chargements, des solutions analytiques sont trouvées. Ces solutions sont ensuite exploitées dans le calcul des frontières des espaces atteignables des mécanismes. Les degrés de liberté non contraints des mécanismes de tenségrité leur permettent de se déformer sous l’application de chargements externes. Une attention particulière est alors portée au calcul de la raideur des mécanismes ainsi que des limites des chargements pouvant être résistés sans perte de stabilité ou de tension dans les câbles. Des observations sont également faites concernant l’amortissement des vibrations des mécanismes dans les directions des degrés de liberté non contraints. / A tensegrity system corresponds to an assembly of components that are subjected only to axial loads. As a consequence, cables or springs can be used for the tensile components thus considerably reducing the mass and inertia of the system. With the goal of benefiting from these interesting properties, this thesis introduces tensegrity mechanisms as alternatives to more conventional type mechanisms for certain types of applications. The main objectives of the thesis are the development and analysis of novel mechanisms. The need for tensegrity mechanisms to remain in configurations where their cables and springs are subjected to tensile loads complicates their development signi- ficantly. Consequently, a new approach based on the use of springs is proposed to overcome this difficulty. In order to facilitate the use of this approach, certain rules pertaining to the quantity of springs used in a given architecture are formulated. Based on these rules, two planar mechanisms, three spatial mechanisms and two modular mechanisms are developed using existing tensegrity systems. Each new mechanism is modeled with the goal of analyzing its kinematics, statics and dynamics. Due to the presence of unconstrained degrees of freedom in the mechanisms’ architectures, relations between their input and output variables are influenced by any external, gravitational or inertial loads that might be acting. By assuming a quasi-static regime, such relations can be computed using a numerical approach. However, for the specific case where the mechanisms are not subjected to any loads, analytical solutions are found. These solutions are then exploited in order to compute the boundaries to the mechanisms’ workspaces. The mechanisms’ unconstrained degrees of freedom allow them to deform under the application of external loads. Special attention is thus given to the stiffness of the mechanisms as well as the limits of the external loads that they may resist without losing their stability or the tension in their cables. Observations are also made regarding the damping of the mechanisms’ vibrations along the unconstrained degrees of freedom.

Dans le domaine de la robotique les robots parallèles sont contraints à des tâches très spécifiques en comparaison avec leurs homologues sériels à cause de leur espace de travail plus restreint. L'objet de ce sujet de maitrise est de présenter une architecture de robot parallèle dont l'espace de travail est agrandi par l'introduction d'une redondance cinématique dans son architecture pour le cas plus particulier des robots sphériques. La première partie du mémoire présente l'architecture du robot sphérique. Son modèle cinématique est décrit puis les variables et paramètres qui seront utiles pour les calculs et analyses sont introduits. Les relations cinématiques entrée-sortie sont présentées au travers des matrices Jacobiennes du robot et le concept de singularité est discuté à la fin de cette partie. Les deux phénomènes limitant l'espace de travail sont principalement les interférences mécaniques, ce qui se comprend facilement, et les singularités. Ce qu'on appelle singularité dans le contexte des architectures robotiques sont des configurations du robot où des mobilités entrantes ou sortantes sont inopérantes. Puis, dans la seconde partie, on s'intéresse à l'équivalent architectural du robot présenté dans la première partie auquel la partie redondante a été retirée dans le but de mesurer l'impact de l'introduction de la redondance cinématique sur l'espace de travail du robot. L'architecture du robot simple est alors brièvement présentée, puis les configurations singulières de ce robot sont analysées par deux méthodes différentes : une méthode numérique et une méthode géométrique. Ensuite dans la troisième partie une analyse des lieux des singularités de l'architecture redondante est présentée et quelques contraintes architecturales intéressantes au regard de l'analyse des singularités précédente sont discutées. Les effets de l'introduction de la redondance cinématique sur l'espace de travail sont discutés à la fin de cette partie en analysant les résultats précédents. Dans la quatrième partie on s'intéresse à la résolution du problème géométrique inverse du robot. La redondance cinématique fait que pour une pose spécifique de l'effecteur du robot il existe une infinité de solutions pour les coordonnées articulaires. Dans ce chapitre on explore plusieurs choix possibles pour les coordonnées articulaires, les avantages et inconvénients de plusieurs méthodes sont évalués, puis un exemple de suivi de trajectoire est présenté. Enfin le problème des interférences mécaniques est discuté, notamment son influence sur l'espace de travail du robot. La dernière partie concerne l'élaboration d'un prototype de test pour le robot sphérique redondant. Les différents choix de designs faits, le dimensionnement des actionneurs ainsi que plusieurs problèmes spécifiques sont présentés.

Ces présents travaux de recherche portent sur le développement de nouveaux capteurs de déplacements à fibre optique monomode a modulation d'intensité causée par la présence de désalignements entre les fibres. À partir d'un fondement théorique rigoureux, des nouveaux principes de fonctionnement de capteurs sont développes. Des modèles analytiques d'efficacité de couplage optique entre faisceaux gaussiens sont utilisés. Un modèle basé sur une expression plus juste du mode de propagation fondamental de la fibre monomode et sur le phénomène de propagation d'un champ optique dans l'air libre est aussi développé à l'aide de la méthode de l'intégrale de recouvrement. Les performances expérimentales de deux capteurs sont évaluées et des directives de conception sont proposées. Ces capteurs sont caractérisés par un comportement linéaire, des possibilités de mesure multidimensionnelle, des performances adaptables a différentes applications et l'évitement des désavantages classiques des capteurs à modulation d'intensité. Une étude principalement théorique de quelques autres solutions est également réalisée. / The present research work relates to the development of novel fiber optic displacement sensors based on intensity modulation caused by the presence of misalignments between single-mode fibers. Starting from a rigorous theoretical base, working principles of new sensors are developed. Analytical models for the optical coupling efficiency of gaussian beams are used. A model based on a better expression of the fundamental propagation mode of single-mode fibers and on the optical field propagation phenomenon is developed using the overlap integral method. The experimental performances of two sensors are evaluated and design guidelines are proposed. These sensors are characterized by a linear behavior, possibilities of multidimensional measurement, adaptable performances to various applications and the avoidance of the classical drawbacks of intensity modulation based fiber optic sensors. A mainly theoretical study of other sensors is also achieved.

Les préhenseurs robotiques sont largement utilisés en industrie et leur déploiement pourrait être encore plus important si ces derniers étaient plus intelligents. En leur conférant des capacités tactiles et une intelligence leur permettant d’estimer la pose d’un objet saisi, une plus vaste gamme de tâches pourraient être accomplies par les robots. Ce mémoire présente le développement d’algorithmes d’estimation de la pose d’objets saisis par un préhenseur robotique. Des algorithmes ont été développés pour trois systèmes robotisés différents, mais pour les mêmes considérations. Effectivement, pour les trois systèmes la pose est estimée uniquement à partir d’une saisie d’objet, de données tactiles et de la configuration du préhenseur. Pour chaque système, la performance atteignable pour le système minimaliste étudié est évaluée. Dans ce mémoire, les concepts généraux sur l’estimation de la pose sont d’abord exposés. Ensuite, un préhenseur plan à deux doigts comprenant deux phalanges chacun est modélisé dans un environnement de simulation et un algorithme permettant d’estimer la pose d’un objet saisi par le préhenseur est décrit. Cet algorithme est basé sur les arbres d’interprétation et l’algorithme de RANSAC. Par la suite, un système expérimental plan comprenant une phalange supplémentaire par doigt est modélisé et étudié pour le développement d’un algorithme approprié d’estimation de la pose. Les principes de ce dernier sont similaires au premier algorithme, mais les capteurs compris dans le système sont moins précis et des adaptations et améliorations ont dû être appliquées. Entre autres, les mesures des capteurs ont été mieux exploitées. Finalement, un système expérimental spatial composé de trois doigts comprenant trois phalanges chacun est étudié. Suite à la modélisation, l’algorithme développé pour ce système complexe est présenté. Des hypothèses partiellement aléatoires sont générées, complétées, puis évaluées. L’étape d’évaluation fait notamment appel à l’algorithme de Levenberg-Marquardt.

As a result of the remarkable demands on electronic and other portable compact devices, the need to produce various miniaturized parts, particularly those made from metallic sheet is growing. In other words, in order for manufacturing companies to stay in competition, they are required to develop new and innovative fabricating processes to produce micro components with more complex features and a high standard of quality and functionality. Microforming is a micro fabrication process that can be employed efficiently for mass production with the advantages of greatly minimizing material waste and producing highly accurate product geometry. However, since the clearance between the rigid tools, i.e. punch and die, utilized in microforming techniques is relatively very small, there is a high risk of damaging the tools during the forming operations. Therefore, the use of forming tools made of flexible materials in sheet metal forming processes at micro scale has powerful potential advantages. The main advantages include a reduction in the production cost, eliminating the alignment and mismatch difficulties, and also the creation of parts with different geometrical shapes using the same flexible tool. As the workpiece is in contact with a flexible surface, this process can significantly improve the quality of the obtained products. Despite these clear advantages, micro flexible forming techniques are currently only utilized in very limited industrial applications. One reason for this is that the deformation behaviour and failure mode of sheet metals formed at micro scale are not yet well understood. Additionally, the experience-based knowledge of the micro-forming process parameters is not sufficient, particularly when flexible tools are used. Hence, to advance this technology and to improve the production quality of formed micro parts, more investigation of the key process parameters related to the material deformation are needed. The main contribution of this work is the development of a novel technique for achieving micro deep drawing of stainless steel 304 sheets using a flexible die and where an initial gap (positive or negative) is adopted between the blank holder plate and an adjustment ring utilized in the size-scaled forming systems developed for this purpose. The interesting point here is that this study presents the first attempt of employing flexible material as a forming die tool in the micro deep drawing technology to produce micro metallic cups at different scaling levels. Polyurethane rubber materials are employed in this study for the forming flexible die with various Shore A hardness. Also, the stainless steel 304 sheets utilized for the workpieces have different initial thicknesses. Various parameters that have a significant influence on the sheet formability at micro scale are carefully considered, these include initial gap value, rubber material properties, initial blank thickness, initial blank diameter, friction coefficients at various contact interfaces, diameter and height of the rubber die and process scaling factor. The size effect category of process dimension was also taken into account using similarity theory. Three size-scaled micro deep drawing systems were developed correspondingly to three different scaling factors. In each case, finite element simulations for the intended micro drawing systems are performed with the aim of identifying optimum conditions for the novel forming methodology presented in this thesis. The numerical models are built using the known commercial code Abaqus/Standard. To verify the microforming methodology adopted for the proposal technique as well as to validate the predictions obtained from simulations, an appropriate number of micro deep drawing experiments are conducted. This is achieved using a special experimental set up, designed and manufactured to fulfil the various requirements of the micro-forming process design procedure. The new knowledge provided by this work provides, for the first time, a predictive capability for micro deep drawing using flexible tools that in turn could lead to a commercially viable production scale process.

Ultrasonic cutting devices have been successfully used in several industries, especially the food industry. This knowledge, developed for industrial procedures, has been exported to other areas where it is having great impact. In medicine, during the last 30 years, di↵erent ultrasonic devices have been designed for a wide variety of surgical procedures involving soft tissue, and even more recently for cutting of bone. The increasing numbers of surgeons adopting ultrasonic devices as the device of choice has in turn increased the demand for devices which are able to be used increasingly in new procedures with more dicult to access surgical sites. Currently, ultrasonic cutting devices consist of a Langevin piezoelectric transducer attached to a cutting blade both tuned to resonate in a longitudinal mode at a low ultrasonic frequency, usually in the 20-50 kHz range. The first commercial ultrasonic devices for bone cutting applications, designed by the Italian company Mectron and called Piezosurgeryr, were based on a Langevin piezoelectric transducer. Langevin transducers incorporate a piezoceramic stack capable of delivering a few microns of vibration amplitude, and therefore the transducer and the device as a whole must be resonant to achieve the required ultrasonic displacement amplitude at the cutting tip. Because the ultrasonic blade is a tuned component its length must be a half-wavelength or a multiple of the half-wavelength at the driving frequency. Also, because Langevin transducers can only deliver a few microns of vibration amplitude, the blade profile must be carefully designed to provide sucient vibration amplitude gain to meet the requirements of the material to be cut. Therefore the cutting blade itself incorporates high amplitude gain, which can lead to very high stresses, and the design of the blade geometry is somewhat restricted by the requirement for resonance. These two geometry requirements can be very restrictive in the design of devices; a half- wavelength at a low ultrasonic frequency leads to quite a large cutting device and profiling for high gain leads to very high stresses. This thesis investigates adapting the class V flextensional ‘cymbal’ transducer for power ultrasonic applications. The cymbal transducer consists of piezoelectric rings bonded to two end-caps with truncated conical shape. When the ring contracts radially under an AC voltage, the end-caps flex providing an amplified motion normal to the cap surfaces. This thesis introduces a new prototype of an ultrasonic cutting device for bone surgery based on a cymbal transducer, optimised for use in power ultrasonics applications, which removes many of the geometrical restrictions on the cutting tip. For the proposed application, a cutting blade is attached to one of the vibrating end-caps with little e↵ect on the operational frequency. Thus, the blade behaves nearly as a rigid body, without the need to be a tuned component of the device. The enormous benefit of this technology is that the cutting blade design can focus more closely on delivering the best interaction between the blade and bone to provide a highly accurate cut, and also the ultrasonic device can be miniaturised to allow the design of devices for delicate orthopaedic procedures involving minimal access surgery. The results show how the cymbal transducer can excite suciently high vibration displacement amplitude at lower driving voltages, by adapting the configuration of the cymbal to remove the problem of epoxy layer debonding and by optimising the cymbal end-caps and geometry through finite element modelling supported with experimental vibration characterisation. Preliminary trials of the resulting prototype ultrasonic bone cutting device, which operates near to 25 kHz, are presented to illustrate the success of this novel device design.

Numerous testing regimes have been used in vitro to assess the fatigue behaviour of acrylic bone cements. While some attempts have been made to introduce an optimal protocol that measures the fatigue life of bone cement under similar stress conditions to those exist in vivo, the effects of specific testing variables such as test specimen specification and stress parameters are still questionable. These factors can be important since inconsistency in results have been reported regarding the precise effects of other variables such as the mixing method of cement components and the resultant porosity. For a given series of testing variables; namely, specimen cross sectional shape, surface production method and stress type and level (herein collectively termed testing regime), this study investigates the effect of each variable on both the fatigue life and the fatigue crack propagation properties (fatigue behaviour) of bone cement. Testing was constantly performed in 37˚C saline under stress-controlled conditions at a frequency of 3Hz (2Hz for the CT specimens). All specimens were produced after vacuum mixing of the cement components and soaked in 37˚C saline for 1- 6 weeks. Specimens were manufactured with two cross sectional shapes: rectangular (ISO 527-2) and circular (ASTM F2118), using two production methods: direct moulding or machining. Two different bone cements were used: SmartSet GHV and CMW1. For each specimen type, at least 10 specimens were fatigued to failure at a maximum stress of 20 MPa applying either fully reversed tension-compression (R= –1) or tension-tension (R= 0.1) loading, followed by Weibull analysis. For the fully reversed loading only, at least 5 specimens were tested for each group at other three levels: ±12.5, ±15 and ±30 MPa and the four stresses were compared using S-N curves. Behaviour of fatigue cracks were assessed based on the cyclic stress-strain responses. CT specimens were used to measure the crack growth rates in the two cements. The findings of this study have emphasised the important role of the set of a testing regime variables included in testing and identified the influence of each testing variable on the fatigue behaviour of bone cement. Machining of test specimens and applying high stress levels, in particular, can lead to irrelevant findings when considering the in vivo conditions, depending also on the cement composition. While these “inappropriate” testing variables can be considered as possible reasons for the variations in fatigue results reported in previous work, it is suggested to consider the effects of these variables in future work.

In recent years, there has been notable interest in the integration of smart and active materials, such as shape memory alloys, in the design of tuneable and multiple frequency devices. There is a growing desire to be able to tune transducers for a range of applications. As an example, surgical procedures could be enhanced by using an ultrasonic device whose performance could be tailored to penetrate more than one material, such as bone and soft tissue. Research conducted on cymbal transducers, a type of Class V flextensional transducer developed at Pennsylvania State University in the early 1990s, has been largely limited to low power applications, such as for hydrophone systems, and their performance in high power applications has only recently been studied. As such, the integration of smart materials to expand the useful applications of this type of transducer has not been fully explored. In this investigation, a shape memory alloy (SMA) called nickel-titanium, or Nitinol, has been adopted in two forms, one being superelastic and the other shape memory, as the end-cap material in the classical cymbal transducer configuration. The resonant frequencies of these transducers can be tuned by changes to the temperature of the Nitinol, which alters the microstructure, and the modulus, of the material. The microstructure of Nitinol can also be controlled by changes in applied stress. The phases present in the Nitinol microstructure are relatively hard cubic austenite and comparably soft monoclinic martensite. An intermediate phase, called the R-phase, can also appear. This is a rhombohedral distortion of austenite, and has been known to be a source of inconvenience for those who wish to avoid multiple stage transformations. An advantage of using Nitinol end-caps in the classical cymbal transducer configuration is that they are very small, hence minimal thermal energy is required to generate a phase transformation. Also, cymbal transducers are very simple and inexpensive to fabricate. The first part of this research focuses on the development of a dual resonance cymbal transducer using steel and titanium as the end-cap materials. Dynamic analysis techniques comprising electrical impedance measurements, experimental modal analysis (EMA) and vibration resonance response characterisation (VRRC) using laser Doppler vibrometry are introduced and form the dynamic characterisation process. The experimental data is supported in part by finite element analysis (FEA). It is demonstrated that a major problem in cymbal transducer fabrication is the difficulty in controlling the deposition of epoxy resin which is used to create the mechanical coupling in the transducer. This means that the bond layers in a transducer will likely be dissimilar, thereby introducing asymmetry into the transducer. This asymmetry can contribute to the dual resonance in a cymbal transducer. The cymbal transducer is designed to be actively tuneable by the incorporation of Nitinol end-caps in the transducer assembly. The characterisation of Nitinol transducers is performed using the dynamic characterisation methods in conjunction with differential scanning calorimetry (DSC). This is a thermoanalytical technique which has been adopted to estimate the transformation temperatures of Nitinol, and hence the temperatures at which each transducer must be driven to generate the desired operating frequencies. It is demonstrated that in certain cases, particularly with respect to superelastic Nitinol, the estimations of the transformation temperatures from the DSC analysis of Nitinol can be misinterpreted. The dynamic performance of Nitinol vibrating at ultrasonic frequencies has not before been the subject of detailed investigation, including the influence of superelasticity on the vibration response of an ultrasonic transducer. Superelasticity occurs in the austenite phase of Nitinol, where austenite reorients to martensite after a characteristic stress threshold is passed, thereby accommodating very large strains. The results show that whilst Nitinol can be used to fabricate cymbal transducers with tuneable resonant frequencies, there is no evidence that superelasticity contributes to the vibration response of the transducers. The incorporation of shape memory Nitinol in a simple prototype actuator device is also considered, where it appears that the transformation of the shape memory Nitinol is affected by the affixed cylinders used to create the device.

Electrosurgical vessel sealing devices have been demonstrated to reduce patient blood loss and operative time during surgery. Whilst the benefits of such devices are widely reported there is still a large variation in the quality of the seal produced, with factors such as vessel size known to effect seal quality. The study aimed to investigate parameters affecting device performance and improve the seal quality. The burst pressure test was used to assess the seal quality and tissue adhesion was measured using a peel test. Additionally histology techniques were used to quantify vessel morphology and found that with an increase in elastin content there was a reduction in seal quality. A number of device modifications were made, testing a selection of non-stick coatings and surface features of the shims. No coating reduced the level of tissue adhesion to the device, but results found that with a greater level of adhesion there was a reduction in seal quality. Considering the different surface features one design, a combination of longitudinal and transverse grooves, resulted in a seal failure rate of 0.0%, a significant improvement in device performance. Two FEM’s were produced to further investigate the device modifications; one in FEBio investigating the mechanical aspects of vessel sealing and the second a multiphysics model to investigate the thermal aspects of vessel sealing. Results from both FEM’s showed a difference in shim performance, with the addition of surface features effecting the stress distribution within the vessel wall and the heat distribution. Additionally DIC was used to capture the vessel sealing process, with results showing each seal was produced in a different way with different levels of tissue contraction. Research conducted demonstrated a number of significant relationships between seal quality and vessel properties, but did not find an explanation for all variation occurring.

The demand for better indoor environment has led to a wide use of heating, ventilating and air conditioning (HVAC) systems. Employing advanced HVAC control strategies is one of the strategies to maintain high quality indoor thermal comfort and indoor air quality (IAQ). This thesis aims to analyse and discuss the potential of using advanced control methods to improve the indoor occupants’ comfort. It focuses on the development of controllers of the major factors of indoor environment quality in buildings including indoor air temperature, indoor humidity and indoor air quality. Studies of the development of control technologies for HVAC systems are reviewed firstly. The problems in existing and future perspectives on HVAC control systems for occupants’ comfort are investigated. As both the current conventional and intelligent controllers have drawbacks that limit their applications, it is necessary to design novel control strategies for the urgent issue of indoor climate improvement. Hence, a concept of designing the controllers for indoor occupants’ comfort is proposed in this thesis. The proposed controllers in this research are designed by combining the conventional and intelligent control technologies. The purpose is to optimize the advantages of both conventional and intelligent control methods and to avoid poor control performance due to their drawbacks. The main control technologies involved in this research are fuzzy logic control (FLC), proportional-integral-derivative (PID) control and neural network (NN). Three controllers are designed by combining these technologies. Firstly, the fuzzy-PID controller is developed for improvement of indoor environment quality including temperature, humidity and indoor air quality. The control algorithm is introduced in detail in Section 3.2. The computer simulation is carried out to verify its control performance and potential of indoor comfort improvement in Section 4.1. Step signal is used as the input reference in simulation and the controller shows fast response speed since the time constant is 0.033s and settling time is 0.092s with sampling interval of 0.001s. The simulating result also proves that the fuzzy-PID controller has good control accuracy and stability since the overshot and steady state error is zero. In addition, the experimental investigation was also carried out to indicate the fuzzy-PID control performance of indoor temperature, humidity and CO2 control as introduced in Chapter 5. The experiments are taken place in an environmental chamber used to simulated the indoor space during a wide period from late fall to early spring. The results of temperature control show that the temperature is controlled to be varying around the set-point and control accuracy is 4.4%. The humidity control shows similar results that the control accuracy is 3.2%. For the IAQ control the maximum indoor concentration is kept lower than 1100ppm which is acceptable and health CO2 level although it is slightly higher than the set-point of 1000ppm. The experimental results show that the proposed fuzzy-PID controller is able to improve indoor environment quality. A radial basis function neural network (RBFNN) PID controller is designed for humidity control and a back propagation neural network (BPNN) PID controller is designed for indoor air quality control. Then, in order to further analyze the potential of using advanced control technologies to improve indoor environment quality, two more controllers are developed in this research. A radial basis function neural network (RBFNN) PID controller is designed for humidity control and a back propagation neural network (BPNN) PID controller is designed for indoor air quality control. Their control algorithms are developed and introduced in Section 3.3 and Section 3.4. Simulating tests were carried out in order to verify their control performances using Matlab in Section 4.2 and Section 4.3. The step signal is used as the input and the sampling interval is 0.001s. For RBFNN-PID controller, the time constant is 0.002s, and there is no overshot and steady state error. For BPNN-PID controller, the time constant is 0.003s, the overshot percentage is 4.2% and the steady state error is zero based on the simulating results. Simulating results show that the RBFNN-PID controller and BPNN-PID controller have fast control speed, good control accuracy and stability. The experimental investigations of the RBFNN-PID controller and BPNN-PID control are not included in this research and will carried out in future work. Based on the simulating and experimental results shown in this thesis, the indoor environment quality improvement can be guaranteed by the proposed controllers.