Springer - unmanned surface vehicle

Artificial intelligence techniques such as artificial neural networks and fuzzy logic are now essential tools in the design of a variety of control systems. These techniques are being employed either to form an integral part of a hybrid control strategy or in the formulization of a stand alone robust control algorithm. For several years control schemes based on fuzzy rule bases have proven to be very successful in several fields of study from a practical viewpoint. However, the major drawbacks with this approach are in the development of the linguistic fuzzy rule bases and their optimization. Heuristic, self-organizing and adaptive search methodologies have been exploited by the AMS Research Group to overcome these shortcomings in the design of fuzzy controllers.

The performance of any control scheme, of course, is governed by the quality of the input signal it receives. For many years linear and extended Kalman filters have been the popular choice for obtaining the best signal estimate from sensor data that has been contaminated with noise. Kalman filters are also widely used in the formation of multi-sensor data fusion (MSDF) algorithms.

In the design of conventional Kalman filters significant difficulties can arise owing to incomplete a priori knowledge concerning their process noise covariance and measurement noise covariance matrices. For most practical applications, these matrices are initially estimated and fixed.

Therefore, the problem which has to be overcome is the optimality of the estimation algorithm in a Kalman filter setting as it is closely connected to the quality of a priori information about the process and measurement noise. It has been shown that insufficiently known a priori filter statistics can reduce the precision of the estimated filter states or introduce biases in their estimates. In addition, incorrect a priori information can also lead to instability in the filter algorithm. To surmount such problems, AMS has developed adaptive Kalman filter designs using fuzzy logic and genetic algorithms that regularly update the aforementioned matrices.

Adaptive fuzzy Kalman filters are also being developed by AMS for use in MSDF algorithms. MSDF may be described as the acquisition, processing and the synergistic combination of sensor information to provide a more comprehensive understanding of the phenomenon being monitored. Owing to the availability of data from multiple sensor sources, uncertainty can be reduced, noise rejected and sensor failure tolerated as a result of built in redundancy.

All the above techniques have been and are being actively researched and applied in a number of marine and industrial applications.

Research grants

A.Khan, S.Sharma and M.Gianni: AB Precision, Poole - £183,271 (KTP awarded 2019-2021)

J.Wan SoE small research grant: £2500 (2017-18)

A.Khan,G. Masala & S. Sharma, "Modifying an Existing Electric-Car with Wireless Connectivity for Safe and Autonomous Driving" SoENG research grant £2500 (2017-18)

A. Khan, "Funding support towards conference participation in Nice, France, 23-27 July 2017", SoENG travel grant £1000 (2017-18)

Sharma SK "The Royal Academy of Engineering (RAE) Newton Research Collaboration award with NIT, Rourkela, India", £11.9K (2015-2016)

Sharma SK "The Royal Society Grant for International Exchanges in conjunction with NSFC (China)", £12k (2013-2015)
Sharma SK and Sutton R "ESF-CUC PhD studentship", £60K (2012-2015)

EPSRC Grant EP/I012923/1 An Intelligent Integrated Navigation and Autopilot System for Uninhabited Surface Vehicles £354K, 2011-2014

TSB - Offshore Wave Energy Limited Marine Demonstrator (Optimal Control Strategy for IT Power’ OWEL WEC) £2.5M (UoP £47K, 2010 – 2013)

Sharma SK and Summerscales J. "To design and manufacture a state of the art resin processing control and delivery system to manufacture composite materials predominantly for the aerospace industry" KTP grant, Oct 2007, £111,672, (over 2 years)

Ming Dai Y and Summerscales J. " Evaluation of the cormarent tidal stream energy device to include concept validation" DTI grant, Oct 2007, £134,431, (over 1 year)

Sutton R. Wine Fermentation Process Monitoring System, EC CRAFT Award, July 2004, 267,280 Euros (over 2 years)

Sutton R. and Chudley J. “An Unmanned Surface Vehicle with Pollutant Tracking and Surveying Capabilities”, EPSRC , April 2004, £250,746 (over 3 years)

Sutton R., Chudley J. and Burns R.S. “An Autonomous Underwater Vehicle with Adaptive Tracking and Navigation Capabilities” EPSRC, September 2001, £218336 (over 3 years). Joint project with Cranfield University – total value £428,480

Sutton R. “IMPROVES” EPSRC, April 2001, £143,099 (over 3 years). Joint project with University of Southampton and University of Wales College Newport – total value £596,200

Khan A "Vehicle-2-Vehicle Connectivity", Santander Seed-corn Research Scholarship, £5K (2016-17)

Khan A, Sharma S & Wan J "Technology Enhanced Learning: Use of ‘Clickers’ and On-line Web Page in Teaching to Enhance Student Participation, Engagement, Learning and Experience", Marine Science & Engineering Small Teaching Grant, 3K (2016-17)

Khan A, Pemberton P & Howell K "Underwater Wireless Sensor Network for Remote Instrumentation and Monitoring Purposes", Marine Science and Engineering Small Research Grant, £2.5K (2016-17)

Khan A, Pemberton P & Schoenborn P "Delivering Impact via Student Research:Sharing Best Practice", Pedagogic Research and Teaching Innovation Fund, £6.5K (2017-18)

Wan J, Santander Seed-Corn Research Scholarship 2016-17: £5000

Wan J, Distinguished Visiting Fellowships Royal Academy of Engineering Round 7: £5000

Wan J, EPSRC First Grant (Hybrid set-theoretic approaches for constrained control and estimation with applications to autonomous sailing boats): £90,722

Jian Wan has been successful in winning a seed funding grant of £20,000 from the Royal Academy of Engineers Frontiers of Engineering, following a symposium in Pretoria, South Africa.

Project: Autonomous Catamarans for Marine Observation and Coastal Surveillance.

Springer 2 - unmanned surface vehicle