Peter Young: The Art and Craft of DBM Modelling

Peter Young Profile

Peter C. Young: Emeritus Professor of Environmental Systems, Lancaster Environment Centre and an Associate of the Data Science Institute, Lancaster University.

Over the last 67 years, I have been extremely fortunate, both in my life and career. Together with my long-time partner Wendy, we have led a wonderful life; and my career over this same period has been both stimulating and rewarding, involving moves around the World that neither of us could have anticipated   But this all started with experience as an apprentice in industry.

A Five Year Apprenticeship in the Aircraft Industry, 1958-1963 Followed by M.Sc Research at Loughborough University 1963-1965 

In 1958, I was accepted for a five-year student apprenticeship in aeronautical engineering with English Electric Aviation Ltd., working at locations close to Lancaster; first in Preston and then at Warton Aerodrome near Lytham St. Annes. The Company sent me to Loughborough University (at that time Loughborough College of Advanced Technology) where, in 1962, I was awarded a B.Tech. degree. After the final year of my apprenticeship, I was approached by the Head of the Department of Transport Technology at Loughborough and invited to pursue an M.Sc. degree by research.  I accepted this offer and married Wendy in 1963, just before we moved to Loughborough. Wendy, who has been my partner for the past 67 years is a constant source of encouragement and inspiration, as well as being an accomplished craftswoman, particularly in the area of embroidery and textiles. My research at Loughborough was concerned with adaptive control using a ‘hybrid’ (analogue/digital) computer being developed in the Electrical Engineering Department by a research student there, Colin Dwyer. I worked with Colin for two years and we were awarded the PACE Prize by the EAL computer company for the work. This was fortuitous because it helped to establish the next step in my career.

Ph.D Research on Data-Based Modelling of Dynamic Systems at Cambridge University, 1965-1967

In 1965  I was was awarded a Whitworth Fellowship (WhF), first introduced by the industrialist Sir Joseph Whitworth to provide support for 30 former apprentices to go to University. By this time, however, only two annual awards were offered to cover Ph.D research. I was fortunate to receive one of these and it led to me starting a Ph.D studentship in the Control Division of the Engineering Department at Cambridge University . Professor John Coales, the Head of the Division, was about to have a hybrid computer installed and wanted me to carry out research on the data-based modelling and control of stochastic dynamic systems. Unfortunately, however,  there were difficulties with the installation of the computer and so I had to carry out my experiments on an analogue computer, with the modelling carried using time series data sampled from the analogue computers on paper tape and then processed on the digital computer! Despite these difficulties, my research went well and I developed novel recursive estimation procedures based on an instrumental variable approach to statistical estimation. During this time, our first son, Timothy, was born in 1966, followed by Melanie in 1967; and I was awarded MA and Ph.D degrees for my research in 1970.

A Two-Year Sojourn in California, 1968-1969

Between early 1968 and late 1969, I was employed as a civilian research engineer at a US Navy research station in the Mojave Desert of California, working mainly on the adaptive control of airborne vehicles (see e.g. Young, 1980) and strapped down inertial guidance. This stay in California covered an important period in the history of the USA, with Apollo landing on the Moon in 1969. It also introduced us to the natural wonders of California and experiences of wild nature that influenced our life from that time forward.

Back to Cambridge as a lecturer in the Engineering Department, 1970-1975

In 1970, I received a surprise invitation from Professor Coales to return to Cambridge as a Lecturer in the Control Division of the Engineering Department. A few months after taking up this position, I was elected as an  Official Fellow of Clare Hall Cambridge (at that time the graduate part of Clare College) . These appointments allowed me to start serious research on the data-based modelling of dynamic systems, recursive estimation and time-series analysis. During this time, I started work on the development of a Computer Aided Program for Time series Analysis and Identification of Noisy systems (CAPTAIN), a computer-based `toolbox’ that included the novel algorithms I was developing as tools in an interactive software system (see Young and Shellswell, 1973). Our second son, Jeremy, was born in 1973, two years before I made an important decision that changed my academic career.

This decision was prompted by my wish to be involved in multi-disciplinary R&D, guided by a feeling that most of the important problems faced by the human race are indeed multi-disciplinary. And it was triggered by two main experiences:

  1. First, when I arrived back in Cambridge, I was almost immediately and unexpectedly introduced to the world of macro-economics. This was because Professor Coales, was friendly with some of the most famous economists in Cambridge at that time and, as a result, I was fortunate enough to meet and talk to some of them. They included Professors Joan Robinson and James Meade, younger colleagues of the most famous Cambridge economist John Maynard Keynes; and Nicholas Kaldor, who later became an advisor to the UK Labour Government. Not surprisingly, I was immediately fascinated by macro-economics and with some difficulty, as I recall, acquired and read Keynes’s enormously influential treatise on the macro-economics of the western capitalist economies: The General Theory of Unemployment, Employment, Interest and Money, first published in 1936 (Keynes, 1936). These encounters and consequent enthusiasm led, in turn, to some joint work with students in the Division at the time, resulting in an Engineering Department report CUED/B-Control/TR48 (1973) entitled Macro-economic Modeling – A Case Study, which was published later that year (Young et al., 1973). This was, in effect, a control and systems interpretation of main elements in the ‘Keynesian Economy’.
  2. Second, a Ph.D student, Bruce Beck (who later had a distinguished career in Europe and the United States, becoming Professor of Environmental Systems Analysis in the Warnell School of Forest Resources, University of Georgia, Athens, USA ) asked if he could do research in the environmental area. This led to a study of modeling water quality in the River Cam between Baits Bite and Bottisham (see e.g. Beck and Young, 1975) , based on daily measurements of Biochemical Oxygen Demand (BOD) and Dissolved Oxygen (DO); and, later, collaboration with the Great Ouse River Authority on the Bedford-Ouse Study (see e.g. Young and Whitehead, 1975). It also led to my long association with research in the area of hydrology, as recounted recently to Professor Keith Beven on https://youtu.be/2fZJZEV99yw.

A Research Chair at the Australian National University, 1975-1981 

To my surprise and apparently influenced by my multi-disciplinary research at Cambridge, I was approached by Professor Frank Fenner and invited to take up the position of Professorial Fellow and Head of Environmental Systems in the new  Centre for Resource and Environmental Studies (CRES) that Professor Fenner had established at the Australian National University in Canberra, Australia. It was sad to leave Cambridge but my desire to pursue multi-disciplinary studies overcame any reservations Wendy and I might have had at the time. And it was aided by our friendship with Neville Rees, then Professor of Engineering at the University of NSW, and his wife Bettye, whom we had met whilst working in California. Whilst in Australia, as with our previous experience in California, we were able to enjoy the wonders of the Australian natural environment, aided by research studies I pursued whilst in CRES. For instance, we were involved in a study of degraded water quality in the Peel Inlet of Western Australia (see e.g. Humphries et al, 1980) that allowed our whole Family to drive across Australia via the Nullabor Plain and the cliffs of the Great Australian Bite, camping each night either in the ‘bush’ or near the Ocean. And we were able to return by the famous Indian-Pacific Railway. Another study concerning the potential problems associated with waste water from proposed uranium mines in the Northern Territory allowed me to visit what was to become the Kakadu National Park (see e.g. Young, 1980). Whilst in Australia my research on the data-based identification and estimation of Transfer Function (TF) models continued, with a definitive article on the optimal Refined Instrumental Variable (RIV) method (Young, 1976), culminating in a large 3-part series of papers (Young and Jakeman, 1979-80). The last part of these showed how this RIV estimation procedure could be extended to continuous-time TF models, i.e. differential equation models with multiple inputs; models that are particularly important to the idea of Data-Based Mechanistic (DBM) modelling (see e.g. Young 1993), where the model is interpreted in physically-meaningful terms, something that is made much simpler if the model is in such a linear of nonlinear differential equation form.

Lancaster University, 1981-present

I returned to the UK in 1981 to become Professor of Environmental Systems and Head of the Environmental Science Department. In 1987, I was instrumental in establishing the Institute of Environmental and Biological Sciences at Lancaster; and I helped to set up the Lancaster Centre for Forecasting with Prof. Robert Fildes in 1992. I then directed the Centre for Research on Environmental Systems and Statistics until my full retirement in 2005. A good idea of my research over this period is given in the ResearchGate link at the bottom of this page  It includes writing or editing several books, including latterly: Recursive Estimation and Time Series Analysis (Young, 2011) and True Digital Control (Taylor, Young and Chotai, 2013). After retirement, I have continued to pursue research on data-based mechanistic modelling, forecasting and control. This has included theoretically-oriented  publications, (e.g. Young, 2015; Chen and Young 2022); and applications in numerous areas of research and development. For instance, in addition to my current research, as recounted in my latest book The Art and Craft of Data-Based Mechanistic Modelling, which deals with global climate warming, the COVID-19 pandemic and macro-economics in the USA (see the link from the main page of this site), I have also worked on applications including the modelling and control of robotic systems (e.g. Janot, Young and Gautier, 2017); efficient parameterisation of nonlinear dynamic systems (e.g. Young and Janot, 2018); wireless power transfer (Chen et al, 2020); and industrial system control (Wang et al, 2023).

 

References

  1. P. C. Young and S. H. Shellswell. Captain – a computer aided program for time series analysis and the identification of noisy systems. In Proc. IEE Conference on Computer Aided Control System Design, London, 1973. IEE.
  2. M. B. Beck and P. C. Young. A dynamic model for BOD-DO relationships in a non-tidal stream. Water Research, 9:769–776, 1975.
  3. P. C. Young and P. G. Whitehead. Water quality in river systems: Monte carlo analysis and planning. Water Resources Research, 15:451–459, 1975
  4. R. B. Humphries, P. C. Young, and T. Beer. Systems analysis of an estuary. Technical Report Bull. No. 100, Western Australian Department of Conservation and the Environment, Perth, Western Australia, 1980.
  5. P. C. Young. Mining and the natural environment – systems analysis and mathematical modelling. In S.F.Harris, editor, Social and Environmental Choice: The Impact of Uranium Mining in the Northern Territory, pages 64–78, Canberra, 1980. ANU Press.
  6. P. C. Young. Some observations on instrumental variable methods of time-series analysis. International Journal of Control, 23:593–612, 1976.
  7. P. C. Young and A. J. Jakeman. Refined instrumental variable methods of time-series analysis: Parts I, II and III. International Journal of Control, 29, 1-30; 29, 621-644; 31, 741-764, 1979–1980.
  8. P. Young and A. Jakeman. The development of captain: A computer aided program for time-series analysis and identification of noisy systems. IFAC Proceedings Volumes, 12(7):391–400, 1979 (IFAC Symposium on computer Aided Design of Control Systems, Zurich, Switzerland, 29-31 August).
  9. P. C. Young. Data-based mechanistic models. In P. C. Young, editor, Concise Encyclopedia of Environmental Systems, pages 137–142, Oxford, 1993. Pergamon Press.
  10. P. C. Young. Recursive Estimation and Time-Series Analysis: An Introduction for the Student and Practitioner. Springer-Verlag, Berlin., 2011.
  11. C. J. Taylor, P. C. Young, and A. Chotai. True Digital Control. Wiley, Chichester, 2013.
  12. P. C. Young. Refined instrumental variable estimation: Maximum likelihood optimization of a unified Box-Jenkins model. Automatica, 52:35–46, 2015.
  13. A. Janot, P. C. Young, and M. Gautier. Identification and control of electro-mechanical systems using state-dependent parameter estimation. International Journal of Control, 90(4):643–660, 2017
  14. P. C. Young and A. Janot. Efficient parameterisation of nonlinear system models: a comment on N ̈oel and Schoukens (2018). International Journal of Control, (DOI: 10.1080/00207179.2018.1521008), 2018.
  15. F. Chen, P. C. Young, H. Garnier, Q. Deng, and M. K. Kazimierczuk. Data-driven modeling of wireless power transfer systems with multiple transmitters,. IEEE Transactions on Power Electronics, 35(11):11363–11379, 2020.
  16. F. Chen and P. C. Young. A simple robust method of fractional time-delay estimation for linear dynamic systems. Automatica, 137:110117, 2022.
  17. L. Wang, C. T. Freeman, E. Rogers, and P. C. Young. Disturbance observer-based repetitive control system with nonminimal state-space realization and experimental evaluation. IEEE Transactions on Control Systems Technology, 31(2):961–968, 2023.
  1. A full listing of my 529 publications, including 7 books, is available on ResearchGate (7,575 Research Interest Score; 19,428 Citations; 49,381 Reads; 75 h-index):  https://www.researchgate.net/profile/Peter-Young-19?ev=hdr_xprf