This paper describes `Power Plant Analyzer', an analytical toolbox for power plant dynamic and steady-state performance analysis, and for control system design and test, which has been developed by the Queen's...
详细信息
This paper describes `Power Plant Analyzer', an analytical toolbox for power plant dynamic and steady-state performance analysis, and for control system design and test, which has been developed by the Queen's University Engineering Design Centre. `Power Plant Analyzer' simulates, in an easy-to-use way, fossil power plant systems, such as drum-type, once-through, gas turbine, and combined cycle plants. It provides power engineers and educators with a tool for a deep understanding of the complex and interrelated thermodynamic processes and operating characteristics with manual and automatic control loops. It can also be used for conceptual training of power plant operators, and as an experimental bed for testing control strategies.
Artificial neural networks (ANNs) is made up of many elements with connections between them which influence each other according to the algorithms that the designer employs. Each element is able to influence its outpu...
详细信息
Artificial neural networks (ANNs) is made up of many elements with connections between them which influence each other according to the algorithms that the designer employs. Each element is able to influence its output according to an embedded function whose value is derived from a training phase. Most developments employing ANN rely upon data stored in matrix arrays. It was therefore possible to use matlab that is extremely good at developing matrix data sets to develop ANN structures that could be transformed into formats that could be employed directly on available hardware.
This paper compares the results of four robust multivariable design techniquesapplied to the depth control of a nonlinear Remotely Operated Vehicle (ROV) subjected to parameter perturbations. Three nonlinear control ...
详细信息
This paper compares the results of four robust multivariable design techniquesapplied to the depth control of a nonlinear Remotely Operated Vehicle (ROV) subjected to parameter perturbations. Three nonlinear controltechniques, namely, input-output(i/o) linearization control, input-output linearization control with sliding mode control and input-output linearization control with adaptive fuzzy control are studied. The fourth technique is the H∞ loop shaping design procedure using multiobjective optimization for the simultaneous design of robust stability and performance.
Traveller and driver information systems are usually applied on motorway and high capacity truck road networks, providing information on accidents, queues and congestion levels and recommending diversion routes where ...
详细信息
Traveller and driver information systems are usually applied on motorway and high capacity truck road networks, providing information on accidents, queues and congestion levels and recommending diversion routes where appropriate. using Edinburgh as a case study, this paper describes how traveller information and control systems can be designed as part of an integrated approach to reducing traffic congestion in urban areas. The Edinburgh system illustrates how traveller information and control systems can be applied in an integrated manner within a complex urban environment to tackle the problem of traffic congestion.
using several simple correlation functions, a method has been established for controlling an unmanned vehicle enabling it to either follow a pre-taught path or to follow a computer generated sequence of images equival...
详细信息
using several simple correlation functions, a method has been established for controlling an unmanned vehicle enabling it to either follow a pre-taught path or to follow a computer generated sequence of images equivalent to the pre-taught path. The need for extensive image processing and image recognition is avoided because the images are not processed or interpreted. Furthermore, the method is not specific to unmanned vehicles but can be applied to other areas such as automated assembly. Improvements are still required as the method is particularly susceptible to rotational errors and further work is required to overcome this problem.
Gas is transported through a large network of pipes which are operated over a wide range of pressures. Within these networks are many elements which ensure that all the necessary working conditions are met. An importa...
详细信息
Gas is transported through a large network of pipes which are operated over a wide range of pressures. Within these networks are many elements which ensure that all the necessary working conditions are met. An important role is played by valves which can perform a variety of functions. A valve and actuator arrangement, which is used to throttle gas from streams found on offshore platforms, is considered. A diagnostic system which will detect and isolate specific faults is developed. British Gas supplied data of step tests from four different valves but additional data was required for the diagnostic procedure. A mathematical model of the system was derived using mathematical relationships and established laws which were interpreted in matlab/Simulink.
Increasing levels of congestion in many towns and cities affect bus operations and passenger as well as motorists. For this situation to be contained and reversed, full use has to be made of bus priority systems to pr...
详细信息
Increasing levels of congestion in many towns and cities affect bus operations and passenger as well as motorists. For this situation to be contained and reversed, full use has to be made of bus priority systems to protect buses from congestion, and attract new passengers from private transport. This paper provides an overview of the development and application of current traffic signal controltechniques used in the UK to provide priority for buses in congested situations. The focus is on the Urban Traffic control systems TRANSYT and SCOOT, which are the predominant form of signal control in the UK towns and cities.
The use of neural networks for satisfactory nonlinear depth control of a unmanned underwater vehicle (UUV) was demonstrated. The single hidden-layer, three neuron feedforward network proved to be the best overall cont...
详细信息
The use of neural networks for satisfactory nonlinear depth control of a unmanned underwater vehicle (UUV) was demonstrated. The single hidden-layer, three neuron feedforward network proved to be the best overall controller of the UUV in this application. Performance parameters fell within the required criteria and was robust enough to cope with unfamiliar terrain and tidal-like disturbances. Initial training by backpropagation was improved upon by use of reinforcement training techniques of which the Alopex algorithm proved to be the most effective in finding a minimum global error solution.
This paper addresses the problem of designing guidance, navigation and control systems for Autonomous - Unmanned -Underwater Vehicles (AUVs) to achieve accurate tracking of reference paths defined in a universal refer...
详细信息
This paper addresses the problem of designing guidance, navigation and control systems for Autonomous - Unmanned -Underwater Vehicles (AUVs) to achieve accurate tracking of reference paths defined in a universal reference frame. The theoretical framework adopted is illustrated with a design example in which recent developments in multivariable control theory and classical results in navigation and guidance were applied to the design of a trajectory following system for the AUV MARIUS (Marine Utility Vehicle System). The key ideas in the design methodology are to clearly state performance specifications in the frequency domain, and to use design tools that explicitly address these types of specifications. Thus, the natural constraint that the navigation, control and guidance systems have decreasing bandwidths, can be directly incorporated in the initial phase of the project. Analysis of the integrated system is performed using a simulation package that allows the user to assess the impact of the navigation, guidance, and control algorithms on the dynamic behaviour of the vehicle.
One of the most common database mining tasks is that of inducing relationships which hold between attributes of a database relation. Rules which are discovered to apply to all of the extensional database are called st...
详细信息
One of the most common database mining tasks is that of inducing relationships which hold between attributes of a database relation. Rules which are discovered to apply to all of the extensional database are called strong rules. Strong rules are useful for decision support and are often used in expert systems and expert database systems. The STRIP algorithm (Strong Rule Induction In Parallel), which discovers strong rules within a database relation using parallel techniques, is presented. How the rules induced using STRIP can be used for implementing a semantic query pre-processor among other potential uses is discussed. How parallel processing can be used to speed up database mining operations is also discussed.
暂无评论