Design And Implementation Of Molecular Computation Algorithm To Several Typical Problems

Leonard Adleman's paper about solving the Hamilton Path Problem with molecular biological method on Science in 1994, created a new era of molecular computation. From then on, this new computing mode which is largely different from traditional electronic computer has been developed swiftly.There are many nonlinear problems and NP complete problems in the morden complicated engineering system. The electronic computer, because of low processing speed and very limited memory, usually spends exponential time to solve any one of them. With advantages of high parallel processing and the huge storage, molecular computation is apparently efficient on dealing with those kind of problems. But, the usual implement of molecular computation, biochemical experiment, has many disadvantages such as, strict environmental requirement, very complicated operations, hard to maintain accuracy and bad human-computer interface.The paper introduced the solutions, based on basic concepts of molecular computation, to several typical hard problems and simulated the algorithm of molecular computation with the software method of electronic computer.Satisfiablity(SAT) problem is a classic NP complete problem. The paper introduced a simulating experiment of solution to SAT problem and managed to get the whole solution set of 6 and 20 variables example. The result validated the correctness and operaterablity of the algorithm.Average distribution problem is a simple combination optimizing problem. The paper solved 5 and 20-scale examples, with algorithm of molecular computation.0-1 planning problem and knapsack problem are very important in operational research. The paper analyzed their principle and tried to solve a typical multidimension knapsack with molecular algorithm. The simulating experiment gets the example's best solution, which validates the fitness of algorithm.Besides solving nonlinear problems with molecular algorithm, the paper also tried to optimizing the continuous function. With an example of multimodal continuous function, the simulating experiment of molecular algorithm can find its both local and global optimal solutions. It proves the use of molecular algorithm can be expanded.