This thesis presents an adaptive model reference controller for a highly maneuverable high performance aircraft, in particular, a modified F18. An adaptive controller is developed to maneuver an aircraft at a high angle of attack. Thus, the aircraft is required to fly over a highly nonlinear flight regime. The adaptive controller presented in this thesis can be viewed as a combination of a linear and a nonlinear controller. Around a fixed flight condition the adaptive controller converges to a linear controller; however, the controller remains a nonlinear controller during maneuvers. The contributions of this thesis lie in two areas. The first area is in control. A successful application of linear adaptive control is presented for a highly nonlinear system. A new method is used to generate the reference trajectory. The reference model uses output feedback to improve the reference trajectory. It is shown that this improvement is necessary because of the control limitations. This work is also important to the control of highly maneuverable high performance aircraft. A successful adaptive controller has been developed to rapidly maneuver an aircraft to a high angle of attack. The main focus of this thesis is adaptive control. / Graduation date: 1993

Generative Multimedia (GM) applications are an increasingly popular way to implement interactive media performances. Our contributions include creating a metric for evaluating Generative Multimedia performances, designing a model for accepting perceiver preferences, and using those preferences to adapt GM performances. The metric used is imprecision, which is the ratio of the actual computation time of a GM element to the computation time of a complete version of that GM element. By taking a perceiver's preferences into account when making adaptation decisions, applications can produce GM performances that meet soft real-time and resource constraints while allocating imprecision to the GM elements the perceiver least cares about. Compared to other approaches, perceiver-directed imprecision best allocates impreciseness while minimizing delay.

High performance twin-tail aircraft, like the F-15 and F/A-18, encounter a condition known as tail buffet. At high angles of attack, vortices are generated at the wing fuselage interface (shoulder) or other leading edge extensions. These vortices are directed toward the twin vertical tails. When the flow interacts with the vertical tail it creates pressure variations that can oscillate the vertical tail assembly. This results in fatigue cracks in the vertical tail assembly that can decrease the fatigue life and increase maintenance costs. For many years, research has been conducted to understand this phenomenon of buffet and to reduce its adverse effects on the fatigue life of aerospace structures. Many proposed solutions to this tail buffet problem have had limited success. These include strengthening the tail, modifying the vortex flow, using an active rudder control, and leading edge extensions. Some of the proposed active controls include piezoelectric actuators. Recently, an offset piezoceramic stack actuator was used on an F-15 wind tunnel model to control buffet induced vibrations at high angles of attack. The controller was based on acceleration feedback control methods. In this thesis a procedure for designing the offset piezoceramic stack actuators is developed. This design procedure includes determining the quantity and type of piezoceramic stacks used in these actuators. The changes of stresses, in the vertical tail caused by these actuators during an active control, are investigated. In many cases, linear controllers are very effective in reducing vibrations. However, during flight, the natural frequencies of the vertical tail structural system changes as the airspeed increases. This in turn, reduces the effectiveness of a linear controller. Other causes such as the unmodeled dynamics and nonlinear effects due to debonds also reduce the effectiveness of linear controllers. In this thesis, an adaptive neural network is used to augment the linear controller to correct these effects.

In this dissertation, we present new approaches to improve standard designs in adaptive control theory, and novel adaptive control architectures. We first present a novel Kalman filter based approach for approximately enforcing a linear constraint in standard adaptive control design. One application is that this leads to alternative forms for well known modification terms such as e-modification. In addition, it leads to smaller tracking errors without incurring significant oscillations in the system response and without requiring high modification gain. We derive alternative forms of e- and adaptive loop recovery (ALR-) modifications. Next, we show how to use Kalman filter optimization to derive a novel adaptation law. This results in an optimization-based time-varying adaptation gain that reduces the need for adaptation gain tuning. A second major contribution of this dissertation is the development of a novel derivative-free, delayed weight update law for adaptive control. The assumption of constant unknown ideal weights is relaxed to the existence of time-varying weights, such that fast and possibly discontinuous variation in weights are allowed. This approach is particularly advantageous for applications to systems that can undergo a sudden change in dynamics, such as might be due to reconfiguration, deployment of a payload, docking, or structural damage, and for rejection of external disturbance processes. As a third and final contribution, we develop a novel approach for extending all the methods developed in this dissertation to the case of output feedback. The approach is developed only for the case of derivative-free adaptive control, and the extension of the other approaches developed previously for the state feedback case to output feedback is left as a future research topic. The proposed approaches of this dissertation are illustrated in both simulation and flight test.

The objective of this thesis is to develop a method for the reconfiguration of three-dimensional modular robots. A modular robot is composed of simple individual building blocks or modules. Each of these modules needs to be controlled and actuated individually in order to make the robot perform useful tasks. The presented method allows us to reconfigure arbitrary initial configurations of modules into any pre-specified target configuration by using graph grammar rules that rely on local information only. Local in a sense that each module needs just information from neighboring modules in order to decide its next reconfiguration step. The advantage of this approach is that the modules do not need global knowledge about the whole configuration. We propose a two stage reconfiguration process composed of a centralized planning stage and a decentralized, rule-based reconfiguration stage. In the first stage, paths are planned for each module and then rewritten into a ruleset, also called a graph grammar. Global knowledge about the configuration is available to the planner. In stage two, these rules are applied in a decentralized fashion by each node individually and with local knowledge only. Each module can check the ruleset for applicable rules in parallel. This approach has been implemented in Matlab and currently, we are able to generate rulesets for arbitrary homogeneous input configurations.

Originally presented as the author's Thesis (doctoral)--Technische Universität München. / "FGV-TUM." Includes bibliographical references (p. 147-156).

This dissertation investigates the problem of developing verifiable stable control architectures for gas turbine engines. First, a nonlinear physics-based dynamic model of a twin spool turboshaft engine which drives a variable pitch propeller is developed. In this model, the dynamics of the engine are defined to be the two spool speeds, and the two control inputs to the system are fuel flow rate and prop pitch angle. Experimental results are used to verify the dynamic model of JetCat SPT5 turboshaft engine. Based on the experimental data, performance maps of the engine components including propeller, high pressure compressor, high pressure, and low pressure turbines are constructed. The engine numerical model is implemented using Matlab. Second, a stable gain scheduled controller is described and developed for a gas turbine engine that drives a variable pitch propeller. A stability proof is developed for a gain scheduled closed-loop system using global linearization and linear matrix inequality (LMI) techniques. Using convex optimization tools, a single quadratic Lyapunov function is computed for multiple linearizations near equilibrium and non-equilibrium points of the nonlinear closed-loop system. This approach guarantees stability of the closed-loop gas turbine engine system. To verify the stability of the closed-loop system on-line, an optimization problem is proposed which is solvable using convex optimization tools. Through simulations, we show the developed gain scheduled controller is capable to regulate a turboshaft engine for large thrust commands in a stable fashion with proper tracking performance. Third, a gain scheduled model reference adaptive control (GS-MRAC) concept for multi-input multi-output (MIMO) nonlinear plants with constraints on the control inputs is developed and described. Specifically, adaptive state feedback for the output tracking control problem of MIMO nonlinear systems is studied. Gain scheduled reference model system is used for generating desired state trajectories, and the stability of this reference model is also analyzed using convex optimization tools. This approach guarantees stability of the closed-loop gain scheduled gas turbine engine system, which is used as a gain scheduled reference model. An adaptive state feedback control scheme is developed and its stability is proven, in addition to transient and steady-state performance guarantees. The resulting closed-loop system is shown to have ultimately bounded solutions with a priori adjustable bounded tracking error. The results are then extended to GS-MRAC with constraints on the magnitudes of multiple control inputs. Sufficient conditions for uniform boundedness of the closed-loop system is derived. A semi-global stability result is proven with respect to the level of saturation for open-loop unstable plants, while the stability result is shown to be global for open-loop stable plants. Simulations are performed for three different models of the turboshaft engine, including the nominal engine model and two models where the engine is degraded. Through simulations, we show the developed GS-MRAC architecture can be used for the tracking problem of degraded turboshaft engine for large thrust commands with guaranteed stability. Finally, a decentralized linear parameter dependent representation of the engine model is developed, suitable for decentralized control of the engine with core and fan/prop subsystems. Control theoretic concepts for decentralized gain scheduled model reference adaptive control (D-GS-MRAC) systems is developed. For each subsystem, a linear parameter dependent model is available and a common Lyapunov matrix can be computed using convex optimization tools. With this control architecture, the two subsystems of the engine (i.e., engine core and engine prop/fan) can be controlled with independent controllers for large throttle commands in a decentralized manner. Based on this D-GS-MRAC architecture, a "plug and play" (PnP) technology concept for gas turbine engine control systems is investigated, which allows us to match different engine cores with different engine fans/propellers. With this plug and play engine control architecture, engine cores and fans/props could be used with their on-board subordinate controllers ready for integration into a functional propulsion system. Simulation results for three different models of the engine, including the nominal engine model, the model with a new prop, and the model with a new engine core, illustrate the possibility of PnP technology development for gas turbine engine control systems.

[Truncated abstract] Fuzzy inference systems (FIS) are information processing systems using fuzzy logic mechanism to represent the human reasoning process and to make decisions based on uncertain, imprecise environments in our daily lives. Since the introduction of fuzzy set theory, fuzzy inference systems have been widely used mainly for system modeling, industrial plant control for a variety of practical applications, and also other decisionmaking purposes; advanced data analysis in medical research, risk management in business, stock market prediction in finance, data analysis in bioinformatics, and so on. Many approaches have been proposed to address the issue of automatic generation of membership functions and rules with the corresponding subsequent adjustment of them towards more satisfactory system performance. Because one of the most important factors for building high quality of FIS is the generation of the knowledge base of it, which consists of membership functions, fuzzy rules, fuzzy logic operators and other components for fuzzy calculations. The design of FIS comes from either the experience of human experts in the corresponding field of research or input and output data observations collected from operations of systems. Therefore, it is crucial to generate high quality FIS from a highly reliable design scheme to model the desired system process best. Furthermore, due to a lack of a learning property of fuzzy systems themselves most of the suggested schemes incorporate hybridization techniques towards better performance within a fuzzy system framework. ... This systematic enhancement is required to update the FIS in order to produce flexible and robust fuzzy systems for unexpected unknown inputs from real-world environments. This thesis proposes a general framework of Adaptive T-S (Takagi-Sugeno) type Rough-Fuzzy Inference Systems (ARFIS) for a variety of practical applications in order to resolve the problems mentioned above in the context of a Rough-Fuzzy hybridization scheme. Rough set theory is employed to effectively reduce the number of attributes that pertain to input variables and obtain a minimal set of decision rules based on input and output data sets. The generated rules are examined by checking their validity to use them as T-S type fuzzy rules. Using its excellent advantages in modeling non-linear systems, the T-S type fuzzy model is chosen to perform the fuzzy inference process. A T-S type fuzzy inference system is constructed by an automatic generation of membership functions and rules by the Fuzzy C-Means (FCM) clustering algorithm and the rough set approach, respectively. The generated T-S type rough-fuzzy inference system is then adjusted by the least-squares method and a conjugate gradient descent algorithm towards better performance within a fuzzy system framework. To show the viability of the proposed framework of ARFIS, the performance of ARFIS is compared with other existing approaches in a variety of practical applications; pattern classification, face recognition, and mobile robot navigation. The results are very satisfactory and competitive, and suggest the ARFIS is a suitable new framework for fuzzy inference systems by showing a better system performance with less number of attributes and rules in each application.

Bibliography : leaves 238-262. / xxiii, 262 leaves : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.)--University of Adelaide, Dept. of Electrical and Electronic Engineering, 1995

The quality of Machine Translation (MT) can often be poor due to it appearing incoherent and lacking in fluency. These problems consist of word ordering, awkward use of words and grammar, and translating text too literally. However we should not consider translations such as these failures until we have done our best to enhance their quality, or more simply, their fluency. In the same way various processes can be applied to touch up a photograph, various processes can also be applied to touch up a translation. This research outlines the improvement of MT quality through the application of Fluency Enhancement (FE), which is a process we have created that reforms and evaluates text to enhance its fluency. We have tested our FE process on our own MT system which operates on what we call the SAM fundamentals, which are as follows: Simplicity - to be simple in design in order to be portable across different languages pairs, Adaptability - to compensate for the evolution of language, and Multiplicity - to determine a final set of translations from as many candidate translations as possible. Based on our research, the SAM fundamentals are the key to developing a successful MT system, and are what have piloted the success of our FE process.