www.gp-field-guide.org.uk
Contents Top Previous Next

12.6 Industrial Process Control

There is evidence that GP is frequently used in industrial process control, although, of course, most industrialists have little time to spend on academic reporting. A notable exception is Dow Chemical, where a group has been very active in publishing results (Castillo, Kordon, and Smits2006aCastillo, Kordon, Smits, Christenson, and Dickerson2006bJordaan, den Doelder, and Smits2006Kordon, Castillo, Smits, and Kotanchek2005Kotanchek et al.2006Mercure, Smits, and Kordon2001). Kordon (2006) describes where industrial GP stands now and how it will progress.

Another active collaboration is that of Kovacic and Balic, who used GP in the computer numerical control of industrial milling and cutting machinery (Kovacic and Balic2003). The partnership of Deschaine and Francone (Francone and Deschaine2004) is most famous for their use of Discipulus (Foster2001) for detecting bomb fragments and unexploded ordnance (Deschaine2006). Discipulus has also been used as an aid in the development of control systems for rubbish incinerators (Deschaine, Patel, Guthrie, Grimski, and Ades2001).

One of the earliest users of GP in control was Willis' Chemical Engineering group at Newcastle, which used GP to model flow in a plasticating extruder (?). Other GP applications in the plastics industry include (Brezocnik, Balic, and Gusel2000). McKay, Willis, Searson, and Montague (2000) also modelled extruding food. Searson, Montague, and Willis (1998) modelled control of chemical reactions in continuous stirred tank reactors. Marenbach (1998) investigated GP in the control of biotech reactors. ? surveyed GP applications, including in the area of control.

Lewin, Lachman-Shalem, and Grosman (2006) and Dassau, Grosman, and Lewin (2006) applied GP to the control of an integrated circuit fabrication plant. Domingos worked on simulations of nuclear reactors (PWRs to be exact) to devise better ways of preventing xenon oscillations (Domingos, Schirru, and Martinez2005). GP has also been used to identify the state of a plant to be controlled (in order to decide which of various alternative control laws to apply). For example, Fleming's group in Sheffield used multi-objective GP (Hinchliffe and Willis2003Rodriguez-Vazquez, Fonseca, and Fleming2004) to reduce the cost of running aircraft jet engines (Arkov, Evans, Fleming, Hill, Norton, Pratt, Rees, and Rodriguez-Vazquez2000Evans, Fleming, Hill, Norton, Pratt, Rees, and Rodriguez-Vazquez2001).

Alves da Silva and Abrao (2002) surveyed GP and other AI techniques applied in the electrical power industry.


www.gp-field-guide.org.uk
Contents Top Previous Next