Evolutionary Algorithm for Network Reconfiguration in Distribution Systems Considering Thermal Operational Conditions
- PDF / 582,443 Bytes
- 16 Pages / 595.276 x 790.866 pts Page_size
- 82 Downloads / 166 Views
Evolutionary Algorithm for Network Reconfiguration in Distribution Systems Considering Thermal Operational Conditions Juan Carlos Cebrian Amasifen · Antônio Paulo da Cunha · Francisco Pereira Júnior · Benedito Vieira de Mello · Lamberto M. B. Beekhuizen
Received: 14 September 2012 / Revised: 11 January 2013 / Accepted: 20 September 2013 / Published online: 22 November 2013 © Brazilian Society for Automatics–SBA 2013
Abstract This article presents a methodology for the development of an evolutionary algorithm (EA) applied to the reconfiguration of radial distribution systems. The EA uses an integrated thermal and electric model (ITM) to evaluate the fitness function. Initially, the ITM calculates steady-state temperature of each conductor and, consequently, their electric impedance. ITM can be used to assess overhead and underground networks. This assessment is done efficiently to obtain the electric losses in each topology obtained by EA. On the one hand, the ITM is applied to two real distribution networks: an underground of 144 bars and another overhead of 1,311 bars. Results show that there is a divergence from 3 to 17 % when the ITM is used during electric losses evaluation. On the other hand, EA is applied to two overhead distribution networks of 14 and 135 bars. The solutions obtained are feasible and with electric losses approx. 14 % lower than that of the initial topology. Keywords Reconfiguration · Evolutionary algorithms · Thermal assessment · Electric cables · Distribution network planning J. C. C. Amasifen (B) · A. P. da Cunha · F. Pereira Júnior Sinapsis Inovação em Energia. Al. Ribeirão Preto, 130, conj. 42-Bela Vista, São Paulo, SP 01331-000, Brazil e-mail: [email protected] A. P. da Cunha e-mail: [email protected] F. Pereira Júnior e-mail: [email protected] B. V. de Mello · L. M. B. Beekhuizen AES-Eletropaulo, São Paulo, SP, Brazil e-mail: [email protected] L. M. B. Beekhuizen e-mail: [email protected]
123
1 Introduction Reconfiguration of distribution networks is a process carried out by utilities to modify distribution network topologies by changing state (open/closed) of interconnection switches (located at strategic points) with the following objectives: to maintain a radial network, reduce electric losses, improve balancing of loads, improve voltage profiles on consumers, and increase system reliability, among others (Guimaraes et al. 2010; Saffar et al. 2011). However, due to considerable number of bars and lines in distribution networks, reconfiguration becomes a combinatorial optimization problem because there are real and binary variables. This is known in the literature as a mixed-integer nonlinear programming problem (Lavorato et al. 2012). In order to solve this optimization problem, various heuristics and artificial intelligence techniques have been developed by several authors. Heuristic methods (Civanlar and Grainger 1988; Baran and Wu 1989; Raju and Bijwe 2008; Ababei and Kavasseri 2011) have obtained topologies with minimum elec
Data Loading...
