This thesis addresses the problem of offline identification of salient patterns in genetic programming individuals. It discusses the main issues related to automatic pattern identification systems, namely that these (a) should help in understanding the final solutions of the evolutionary run, (b) should give insight into the course of evolution and (c) should be helpful in optimizing future runs. Moreover, it proposes an algorithm, Extended Pattern Growing Algorithm ([E]PGA) to extract, filter and sort the identified patterns so that these fulfill as many as possible of the following criteria: (a) they are representative for the evolutionary run and/or search space, (b) they are human-friendly and (c) their numbers are within reasonable limits. The results are demonstrated on six problems from different domains.

A navigation and positioning system for an electric automatic guided vehicle has been designed and implemented on an industrial pallet truck. The system includes an optical sensor mounted on the vehicle, capable of recognizing special markers at a distance of 0.3m. Software implemented in a z-80 microprocessor controls the sensor, performs all data processing and contains the decision making processes necessary for the vehicle to navigate its way to its task location. A second microprocessor is used to control the vehicle's drive motors under instruction from the navigation unit, to accurately position the vehicle at its destination. The sensor reliably recognises markers at vehicle speeds up to 1ms- 1, and the system has been integrated into a multiprocessor controlled wire-guidance system and applied to a prototype vehicle.

Automation in industry provides both cost and management benefits. Within the materials handling field, these benefits are provided by automated guided vehicle systems. Currently available systems suffer from the problems of path and vehicle dependence, i.e. the need for a guide wire and the customised nature of the autanatic equipnent which is only suitable for the vehicles with which it is supplied. Future systems will need to be more complex whilst maintaining the ability to suit the smaller, simpler systems. More interfacing with other automatic manufacturing and warehouse equipnent will be needed and the vehicles themselves will need more intelligence. A free-ranging, autonomous, intelligent, automatic vehicle would overcome the present limitations and simultaneously meet the expected future requirements. This thesis describes a research programme which is adequate to develop such a vehicle. This includes the design of a materials handling system which can operate equally well with both simpler wire guided vehicles and the more complex free­ ranging vehicles. The design is based on the roncept of delegated intelligence where the vehicle sub-systems are all controlled by microprocessors. The major advantage of automated guided vehicle systems is the flexibility they give to the materials handling operation and this is a direct function of the intelligence of the system. The provision of intelligence in the vehicle sub-systems has a maJor impact on the flexibility of the complete system. An automatic vehicle has various sub-systems of which the most immediate is steering. The vehicle steering system is designed in a modular manner. A mathematical model of the steerirg system has been developed and validated, and the building of the steering system on the basis of this model means that future system developnent will be for the sensors only and will be rapid. The control of the steering function by a microprocessor gives the system great flexibility, particularly with the use of software controlled manoeuvres.

No description available.

Computer models, or simulators, are widely used in a range of scientific fields to aid understanding of the processes involved and make predictions. Such simulators are often computationally demanding and are thus not amenable to statistical analysis. Emulators provide a statistical approximation, or surrogate, for the simulators accounting for the additional approximation uncertainty. This thesis develops a novel sequential screening method to reduce the set of simulator variables considered during emulation. This screening method is shown to require fewer simulator evaluations than existing approaches. Utilising the lower dimensional active variable set simplifies subsequent emulation analysis. For random output, or stochastic, simulators the output dispersion, and thus variance, is typically a function of the inputs. This work extends the emulator framework to account for such heteroscedasticity by constructing two new heteroscedastic Gaussian process representations and proposes an experimental design technique to optimally learn the model parameters. The design criterion is an extension of Fisher information to heteroscedastic variance models. Replicated observations are efficiently handled in both the design and model inference stages. Through a series of simulation experiments on both synthetic and real world simulators, the emulators inferred on optimal designs with replicated observations are shown to outperform equivalent models inferred on space-filling replicate-free designs in terms of both model parameter uncertainty and predictive variance.

In this thesis we present an approach to automated verification of floating point programs. Existing techniques for automated generation of correctness theorems are extended to produce proof obligations for accuracy guarantees and absence of floating point exceptions. A prototype automated real number theorem prover is presented, demonstrating a novel application of function interval arithmetic in the context of subdivision-based numerical theorem proving. The prototype is tested on correctness theorems for two simple yet nontrivial programs, proving exception freedom and tight accuracy guarantees automatically. The prover demonstrates a novel application of function interval arithmetic in the context of subdivision-based numerical theorem proving. The experiments show how function intervals can be used to combat the information loss problems that limit the applicability of traditional interval arithmetic in the context of hard real number theorem proving.