Dna Encoding Method Of Weight For Chinese Postman Problem

Aili Han,Daming Zhu

DOI: https://doi.org/10.1109/cec.2006.1688377

2006-01-01

Abstract:We have devised a new DNA encoding method to represent numerical values and apply it to solve the Chinese postman problem, an instance of optimization problems on weighted graphs. For any a weighted, undirected graph G=(V,E), nu(i)is an element of V, 1 <= i <= n, e(j)is an element of E, 1 <= j <= m, where exists a numerical value (weight) w(j) on edge e(j), we convert it into its half-dual graph G'=(V',E'), v(i)'is an element of V', 1 <= i <= m, where v(i)' is mapped from e(i). Our DNA encoding method is based on the half-dual graph G'=(V',E'). For any a vertex nu(i)', we use DNA strand s(i) of length w(i) to encode it. For any an edge nu(i)'nu(j)', we use reverse complementation of the second half of s(i) and the first half of s(j) to encode it. Our work makes weight be easily dealt with and extends the capability of DNA computing to solve NP-hard problems.