W. B. Langdon. Scheduling maintenance of electrical power transmission networks using genetic programming. In John Koza, editor, Late Breaking Papers 245 at the GP-96 Conference, pages 107--116, Stanford, CA, USA, 28--31 July 1996. Stanford Bookstore.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... achieve better results [2] Genetic and evolutionary algorithms are widely considered in many applications [42] including the multiple knapsack problem [28] project scheduling [9] and many problems related to electrical framework, e.g. unit commitment [15, 17] and unit maintenance scheduling [29, 37]. Unit commitment and maintenance scheduling di er a bit from load clipping but the ideas can be easily adopted to load clipping, as well as the ideas from [9, 28] Kim et al. 29] develop simulated annealing and tabu search methods for unit maintenance scheduling. We present genetic algorithm for ....

....each heuristic is used. 7. 5 Genetic algorithms Genetic algorithms (GA s) are used eciently with many di erent and noteworthy dicult combinatorial problems [42, 50, 28] As we have already pointed out, GA s are used with unit commitment and other problems related to electricity management [15, 17, 37, 29]. The basic idea behind GA is taken from the evolutionary biology [42] Figure 27 contains the fundamental structure of a simple GA. 1) Initialize population (2) while not termination do (3) Produce new individuals (4) Insert new individuals into the population (5) od (6) Report on ....

W. B. Langdon. Scheduling maintenance of electrical power transmission networks using genetic programming. In Late Breaking Papers at the Genetic Programming 1996 Conference, pages 107-115, 1996.

....operations competing for the same machine randomly the authors suggest the use of one of twelve priority rules to resolve these conflicts. Further work on schedule builders includes a method using genetic programming, to evolve the best scheduling heuristics to use within the schedule builder (Langdon, 1996). Also relevant to this area is the Evolving Heuristic Choice (EHC) method (Fang, Ross and Corne, 1994) which includes heuristic rules to be used in the schedule builder as part of the chromosome encoding, allowing them to evolve alongside the schedule, and produces far superior results on a ....

....in real life, or choosing between a large selection of available heuristics to see which allows us most effective use of the GA as an optimisation tool. The actual search for a best rule to use in this situation has been explored previously. Fang, Ross and Corne, 1994; Dorndorf and Pesch, 1995; Langdon, 1996) We wish to examine the effectiveness of using rules against an alternative method. 4.2 Pre expressed Preferences A less rigid decision could be made by the use of pre expressed preferences of which line to use for any given product. By creating a matrix which explicitly expressed a value of ....

[Article contains additional citation context not shown here]

Langdon, W., 1996. Scheduling Maintenance of Electrical Power Transmission Networks Using Genetic Programming, in Late Breaking Papers at the GP-96 Conference, John Koza (ed), Stanford Bookstore.

....Dahal et al. [10] presented a genetic algorithm (GA) in 1997 using a binary representation. Further work [11] showed that an integer representation reduced the search space of the GA and also reduced the execution time of the algorithm. On a related topic, genetic algorithms have also been used [12] to schedule the maintenance of electrical power transmission networks. Also in 1997, the authors [13] produced evidence that a tabu search approach could yield good results compared to a genetic algorithm, a simulated annealing approach and a combination of tabu search and simulated annealing. An ....

....to a more specific problem incorporating Japanese legal requirements into the problem. It would be interesting to see whether the algorithm presented here could be usefully applied to this problem, and also if the technique can produce good results for a related problem in the electricity industry [12]. ....

W. B. Langdon, "Scheduling maintenance of electrical power transmission networks using genetic programming," Research Note RN/96/49, University College London, Gower Street, London WC1E 6BT, UK, 28 June 1996.

....benefits of the memetic algorithm could be very significant indeed. The authors next intend to apply the algorithm to a scaled up problem which involves scheduling maintenance for a significantly larger network consisting of 28 nodes and 42 lines over a period of 53 weeks. In this problem, Langdon [7, 8] is able to find a schedule with a cost of 388MW weeks in 78 minutes although Gordon [4] was less successful on this problem. Preliminary applications of the memetic algorithm to this problem indicate that solutions with a significantly lower cost than 388 MW weeks can be found inside 10 minutes ....

W. B. Langdon. Scheduling maintenance of electrical power transmission networks using genetic programming. Research Note RN/96/49, University College London, Gower Street, London WC1E 6BT, UK, 28 June 1996.

No context found.

W. B. Langdon. Scheduling maintenance of electrical power transmission networks using genetic programming. In John Koza, editor, Late Breaking Papers 245 at the GP-96 Conference, pages 107--116, Stanford, CA, USA, 28--31 July 1996. Stanford Bookstore.

....the global optima were missed in many runs. There are many techniques that can be used to ensure population diversity remains high (and so the search is defocused) such as splitting the population into demes, fitness niches and mutation, some of which were used in [ Langdon, 1995; Langdon, 1996b; Langdon, 1996a ] Techniques based on biased mate selection to preserve diversity are discussed in [ Ryan, 1994 ] Defocusing the search means the search is more random and will take longer, if indeed it succeeds. Other approaches to avoid getting trapped at local optima ( premature convergence ) change the ....

W. B. Langdon. Scheduling maintenance of electrical power transmission networks using genetic programming. In John Koza, editor, Late Breaking Papers at the GP-96 Conference, pages 107--116, Stanford, CA, USA, 28--31 July 1996. Stanford Bookstore.

....Figure 1: South Wales Region High Voltage Network In this paper we report work showing the combination of GAs and greedy optimisers described in [Lan95] being applied to the South Wales region of the National Grid. Other work on the South Wales problem using Genetic Programming is reported in [LT97]. Section 2 describes the South Wales region high voltage network, Section 3 describes the fitness functions used in Sections 4 and 5 in which low cost schedules are produced, firstly without the network resilience requirement and secondly (in Section 5) when including it. Details of the GA ....

....been evolved, we then considered evolving maintenance plans which contain a degree of network resilience (cf. Section 5) In this second problem the true conductor rating were used. Considering potential network faults is highly CPU intensive and so the Genetic programming approach described in [LT97] did not attempt to solve the second version of the South Wales problem. 3 The Fitness Function The fitness function used by both GA and genetic programming (GP) approaches to scheduling maintenance in the South Wales region is based on that used in the four node problem. To summarise; ffl The ....

[Article contains additional citation context not shown here]

W. B. Langdon and P. C. Treleaven. Scheduling maintenance of electrical power transmission networks using genetic programming. In K. Warwick, R. Aggarwal, and A. Ekwue, editors, Applied Artificial Intelligence in Power Systems, chapter 10. IEE, 1997.

....University has diverse interests and international collaborative links. At present the major topics include the theoretical underpinnings of GP [PL97, LP97a, Lan98, LP97b] novel representations and operators (Section 2.2. 1) as well as applications such as medical imaging [Pol96c] and scheduling [LT97] 3.3 Scotland Three Scottish universities, Napier, Edinburgh and Glasgow are very active in the GP field. For example at Napier the fundamental scaling problems associated with solving complex problems are under investigation [ABF97] as are issues concerning generating efficient programs from ....

W. B. Langdon and P. C. Treleaven. Scheduling maintenance of electrical power transmission networks using genetic programming. In Kevin Warwick, Arthur Ekwue, and Raj Aggarwal, editors, Artificial Intelligence Techniques in Power Systems, chapter 10, pages 220--237. IEE, 1997.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC