Intelligent control for near-autonomous aircraft missions

Control of a phosphate processing plant via a synergistic architecture for adaptive, intelligent control

Soft-computing tools such as genetic algorithms and neural networks have found increased use in problems of engineering design. Applications have included the use of neural networks as response surface-like function approximations, and in identifying causality in numerical data. Both neural networks and genetic algorithms have also been used in the solution of generically difficult design optimization problems. A more recent application of these tools is in the field of computational intelligence. The present paper explores the use of a machine learning paradigm referred to as a classifier system, the central building block of which is a genetic algorithm, to provide a rational approach through which to construct neural network based global function approximations. The emphasis of the approach is to determine through computations, an optimal distribution of data which maximally improves the quality of the function approximation available from a back-propagation neural network. Numerical results are presented in support of the proposed scheme.

A control engineering perspective on modeling and representation issues, nonlinear analysis, design implementation, and evaluation of intelligent control systems is developed. The importance of using a control engineering approach to assess the contribution of the field of intelligent control is discussed. The ways in which technology has provided a driving force for the field of intelligent control are explained, and several research directions for control engineers working in the area of intelligent control are presented. The ways in which the control engineer's use of modeling, analysis, design, simulation, implementation, and experimental evaluation provide a sound engineering approach to the development of intelligent control systems.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Low-cost environment for analysis and design of knowledge-based control algorithms

The book chapter provides a comprehensive set of knowledge in the field of intelligent control of generated electrical impulses for wire electrical discharge machining. With the designed intelligent electrical pulse control system, the stability of the electroerosion process, as well as the increased surface quality after wire electrical discharge machining (WEDM), can be significantly enhanced compared to standard impulse control systems. The aim of the book chapter is also to point out the importance of monitoring in addition to the established power characteristics of generated electrical pulses, such as voltage and current, as well as other performance parameters. The research was mainly focused on those parameters that have a significant impact on the quality of the machined surface. The own’s theoretical and knowledge base was designed to enrich the new approach in increasing the geometric accuracy of the machined surface, as well as the overall efficiency of the electroerosion process for WEDM through intelligent control of generated electrical pulses.

Wearable strain sensors have huge potential for applications in healthcare, human-machine interfacing, and augmented reality systems. However, the nonlinear response of the resistance signal to strain has caused considerable difficulty and complexity in data processing and signal transformation, thus impeding their practical applications severely. Herein, we propose a simple way to achieve linear and reproducible resistive signals responding to strain in a relatively wide strain range for flexible strain sensors, which is achieved via the fabrication of Janus and heteromodulus elastomeric fiber mats with micropatterns using microimprinting second processing technology. In detail, both isotropic and anisotropic fiber mats can turn into Janus fiber mats with periodical and heteromodulus micropatterns via controlling the fiber fusion and the diffusion of local macromolecular chains of thermoplastic elastomers. The Janus heterogeneous microstructure allows for stress redistribution upon stretching, thus leading to lower strain hysteresis and improved linearity of resistive signal. Moreover, tunable sensing performance can be achieved by tailoring the size of the micropatterns on the fiber mat surface and the fiber anisotropy. The Janus mat strain sensors with high signal linearity and good reproducibility have a very low strain detection limit, enabling potential applications in human-machine interfacing and intelligent control fields if combined with a wireless communication module.

Bare-bones languages serve small systems with ease: Cushmann, R HEDN Vol 25 No 16 (5 September 1980) pp 171–177

Recent developments in the field of computational intelligence (CI) techniques are helping to solve various problems of water resources modeling and management. These techniques have also shown their potential as an alternative approach to conventional controllers. In this paper, artificial neural networks (ANN) and fuzzy systems (FS) are shown to be efficient alternatives to using optimal control algorithms in real-time control of the polder water system of Overwaard in The Netherlands. The relation between the optimal decision or action and the influencing parameters are learned by ANN and FS and then used to derive the decisions and control actions in real-time. It was possible to reproduce the centralized behavior (in terms of water levels and corresponding discharges) of optimal control action by using easily measurable local information. Moreover, it is demonstrated that model simulation with external intelligent controllers is ten times faster than that with the optimal control.

Dynamic signature verification is an important area of biometrics. In this area methods from the field of computational intelligence can be used. In this paper we propose a new method for genetic selection of the most characteristic descriptors of the dynamic signature. The descriptors are global features of the signature and components created within its partitions. Selection of the descriptors is realized individually for each user of the biometric system. Its purpose is to increase the precision of the biometric system by eliminating the descriptors which do not increase efficiency of verification procedure. Number of descriptors (their combination) can be high, so the use of genetic algorithm to reduce their number seems to be justified. Moreover, reduction of descriptors increases interpretability of fuzzy mechanism for evaluation of signatures’ similarity. Proposed method was tested using known dynamic signatures database-MCYT-100.

The lower limbs kinematic chain (LLKC) is the important part of human body. It contains two joints with four degrees of freedom (DOF). As it is difficult to compute LLKC’s inverse kinematic solution quickly and accurately by using neural network or genetic algorithm (GA) due to its high degree of freedom, we proposed a BP-GA approach by combining BP neural network and GA to resolve it. Firstly, the mathematical model of LLKC was built based on D-H method. Then BP neural network output a local optimal solution, which could be served as an individual of GA initial population. And the searching domain of GA could be determined by local optimal solution. Finally, the high-accuracy solution was searched by using the adaptive GA. The actions of walking, squatting and seated position were simulated and results showed that the precise solution could be calculated efficiently by using the proposed approach with the BP neural network high-speed and GA high-accuracy.

As we all know, big data and intelligent computing have been widely used in information science, life science, computer science and intelligent control, intelligent robots, vehicle networking, space technology, marine development and other fields, especially in the future life science and medical field. This paper is based on a new digital optimal decision-making idea of intelligent signal processing and control. On the basis of extending non-digital signal processing and information prediction to signal processing and information prediction in the digital generalized information measurement space, and making full use of optimization algorithms such as digital filtering and frequency spectrum analysis, this article discusses in depth the intelligent optimal control of a class of non-stationary random processes in artificial intelligence, biomedicine, radio communication, automatic control and other scientific and technological fields. It provides a new theory, a new approach and an effective and reliable new method for the mathematical modeling and optimal control of such processes, and has broad application prospects and value.

A comprehensive overview of wind turbine controller technology: Emerging trends and challenges

Subcutaneous insulin administration by deep reinforcement learning for blood glucose level control of type-2 diabetic patients

The PhD project is by far the largest individual research, writing and software project that I have undertaken. This essay provides an information account of my project management experiences over the course of the project. This project debrief is presented as review of the practices that worked (Section II. ) practices that failed (Section III. ) and practices that will be considered if the project were to be repeated (Section IV. ). Finally the general principles and best practices are considered (Section V. ).The motivation for this work was to articulate and capture the general research project management lessons learned as a guide for future project design and work ethic. This is a personal account meaning that the applicability of the practices are relative to my experience and working style. Before the research project I was a trained software engineer with four years of practical work experience in industry. The research project was completed in the field of Computational Intelligence, a sub-field of Artificial Intelligence with strong themes of abstraction information processing and empirical simulation. The general work ethic involved an approximate 50-hour workweek and explicit separation of work from personal with regard to both physical location and activities undertaken [Introduction].

It’s a new idea to make computers be able to obtain “sensations” from a color image through some unsupervised ways. To let the idea come into true, a granule-based model, based on granular computing(GrC) which is a new way to simulate human thinking to help solve complicated problems in the field of computational intelligence, is proposed for color image processing. First, this paper deems data a hypercube, defines two new concepts, attribute granules(AtG) and connected granules(CoG), and presents the definitions of the granule-based model. Then, in order to fulfill the granule-based model, this paper designs a single attribute analyser(SAA), defines some theorems and lemmas related to decomposition, and describes the processing of extracting all attibute granules. Experimental results on over 300 color images show that the proposed analyser is accurate, robust, high-speed, and able to provide computers with “sensations”.