Michael Knowles

The effect of driving style on electric vehicle performance, economy and perception

It has long been known that driving style has a major impact on the efficiency of conventional combustion engine powered vehicles. Particular aspects of conventional driving such as harsh acceleration and deceleration and poor anticipation have been demonstrated to be unfavourable for clear technical reasons relating to the efficiency of the internal combustion engine at particular speeds and loads. Furthermore, definite trends have been identified in terms of the relationship between age and driving style for conventional vehicles. Little work has been done in this area using electric vehicles. This paper addresses this by presenting a detailed study of the performance of a number of drivers around a standard route in an electric vehicle. In addition to highlighting how particular aspects of driving style influence power consumption and regeneration. We also look at how the drivers perceived the electric vehicle compared to conventional vehicles of the same class.

Impact of Second Life Electric Vehicle Batteries on the Viability of Renewable Energy Sources

Aims: To investigate the viability of using second life electric vehicle batteries as ‘buffer packs’ for localized, in home storage of renewable energy. To investigate the potential value of such energy storage in terms of reducing energy bills. To investigate the effect, if any, of social group on energy demand patterns and the subsequent effect on buffer packs. Methodology: Energy data was collected from 15 households and a representative daily demand pattern was formed from each. The availability of solar power was calculated based on standard assumptions regarding UK household installations. The flow of energy between the supply grid, renewable sources (i.e. solar panels) and a 10kWh battery storage system (the ‘buffer pack’) were simulated on a minute by minute basis allowing the savings in power drawn from the grid to be calculated. Results: The simulation has indicated that the use of buffer packs has the potential to greatly increase the utilization of locally generated renewable energy. In some cases the stored energy removed the need to draw any power from the grid. The value of energy storage of this type has been estimated at approximately £250. Conclusion: Buffer packs have substantial potential to increase the degree of utilization of renewable energy sources. However the financial viability of such systems remains questionable, even when utilizing second-life electric vehicle batteries. Further work is recommended to address some of these issues and, in particular, to investigate the effects of seasonal variations in energy use and renewable energy availability on buffer pack applications. Keywords : Renewable energy; solar photovoltaic power; electric vehicles; batteries. - See more at: http://www.sciencedomain.org/abstract.php?iid=268&id=5&aid=2103#sthash.i2FboWT8.dpuf

Reinforcement Learning for Scheduling of Maintenance

Improving maintenance scheduling has become an area of crucial importance in recent years. Condition-based maintenance (CBM) has started to move away from scheduled maintenance by providing an indication of the likelihood of failure. Improving the timing of maintenance based on this information to maintain high reliability without resorting to over-maintenance remains, however, a problem. In this paper we propose Reinforcement Learning (RL), to improve long term reward for a multistage decision based on feedback given either during or at the end of a sequence of actions, as a potential solution to this problem. Several indicative scenarios are presented and simulated experiments illustrate the performance of RL in this application.

Maintenance Strategy Development in the UK Food and Drink Industry

The food and drink industry is a key industrial sector employing many thousands of people making a substantial contribution to the global economy. Small to medium enterprises (SMEs) account for a considerable percentage of the sector’s activity. While many manufacturing sectors have embraced and contributed to the development of modern maintenance practices, the food and drink industry is perceived to be falling behind, a trend which is having a negative effect on the productivity of this industry. This paper investigates the barriers, both real and perceived, to the implementation of modern maintenance practices and the opportunities to apply modern technology to support improved efficiency.

Towards Computational Modelling of Neural Multimodal Integration Based on the Superior Colliculus Concept

Information processing and responding to sensory input with appropriate actions are among the most important capabilities of the brain and the brain has specific areas that deal with auditory or visual processing. The auditory information is sent first to the cochlea, then to the inferior colliculus area and then later to the auditory cortex where it is further processed so that then eyes, head or both can be turned towards an object or location in response. The visual information is processed in the retina, various subsequent nuclei and then the visual cortex before again actions will be performed. However, how is this information integrated and what is the effect of auditory and visual stimuli arriving at the same time or at different times? Which information is processed when and what are the responses for multimodal stimuli? Multimodal integration is first performed in the Superior Colliculus, located in a subcortical part of the midbrain. In this chapter we will focus on this first level of multimodal integration, outline various approaches of modelling the superior colliculus, and suggest a model of multimodal integration of visual and auditory information.

MIRA: A Learning Multimodal Interactive Robot Agent

In this paper we present a robotic head MIRA (multimodal interactive robot agent) which has been developed for studying the learning of human robot interaction and improving our understanding of human robot interaction techniques. In this paper we focus on two main aspects of the system; first, we describe how the robot head learns to recognise faces for supporting the interaction process between a human and MIRA. Second, we show how MIRA can learn to identify sound sources of interest and attend to the source location improving the social interaction effect. We propose that there is substantial potential for learning visual and auditory features in order to increase adaptability and robustness of robotic heads.

Converter simulation using SimPowerSystems a comparison of drive cycles and control strategies

This paper describes the setup of a direct current to direct current (DC-DC) converter simulation using Matlab / Simulinks SimPowerSystems (SPS). The simulation is set up to focus on full drive cycle simulation to be able to visualize the effects of a control strategy on a drive cycle over time. The aim of these simulations is to compare the effects of different control strategies and to introduce a modular control strategy. A modular control strategy is proposed and tested and the effects of the control strategies are described and discussed. The load factor of the battery current is calculated to look for improvements in smoothness.

Novel EV drive train topology — A review of the current topologies and proposal for a model for improved drivability

The use of Direct Current to Direct Current (DC-DC) converters in Electric Vehicles (EV) drive trains allows for optimal combination of a power source complementing an energy source. Various topologies have been analysed and compared for their best features. A set of parameters has been created, based on which a new topology has been developed. The new topology is modular and allows for redundancy while improving driveability and has the potential to reduce cost of the drive train, by increasing the life expectancy of the energy sources.

Condition Management of Marine Lube Oil and the Role of Intelligent Sensor Systems in Diagnostics

Failures in marine diesel engines can be costly and can cause extreme inconvenience when they result in ships becoming stranded. Lubricating oil is a crucial component in maintaining engine reliability and so monitoring its condition is essential. Furthermore the lubricating oil offers early indication of various other engine faults. Current approaches to oil-based condition monitoring involve samples being sent for land based testing which involves considerable delay during which the situation could deteriorate further. Furthermore there is a substantial risk of contamination. The POSSEIDON project aimed to address this by developing a system involving real-time condition monitoring sensors observing the properties of the lubricating oil. Novel sensors were developed which address the specific issues associated with the marine environment. Furthermore, to complement the sensor system outputs, specific monitoring and diagnosis software has been developed to support the operation of onboard personnel with specific advice. On-line management of engine and lubricant condition aboard the ship may thus be achieved. In this paper we will describe the progress achieved in this area by the recently completed POSSEIDON project, outline the opportunities for ongoing development in this area and describe the roadmap for future development. The Reliability Centered Maintenance (RCM) paradigm will be applied to identify critical aspects of oil condition and prioritize parameters for measurement. The critical issues for development of the prototype unit into a viable commercial unit will be discussed including hardware design constraints, sensor miniaturization and display optimization. Issues such onboard connectivity, ship to shore communications will also be addressed.

The Hybrid Integration of Perceptual Symbol Systems and Interactive Reinforcement Learning

In order to produce robots which can interact more effectively with humans we propose that it is necessary for their cognitive processes to be grounded in the same perceptual elements as humans deal with. Perceptual symbol systems offer an attractive mechanism for capturing the symbolic properties of the senses and for integrating them into higher level cognitive processes. We have designed a perceptual symbol system where the robot learns about objects through interaction and reinforcement and have carried out experiments to assess the merits of this approach. We show that the use of human perceptual elements combined with interactive reinforcement leads to intuitive learning and interpretable knowledge structures.