Abstract:

In order to solve large-scale 0-1 nonlinear problem，particle filter is designed and implemented. Particle filter using particle set to indicate the probability can be used in any form of state space model. The core idea is to express the random state particles extracted from the posterior probability distribution，which is a sequential importance sampling method. In solving large-scale nonlinear 0-1 programming problem，the solution is divided into M intervals，uniformly distributed random decimal number particles in each interval，the conversion of the interval length binary number is used to get the initial feasible solution，calculating the mean and variance of the initial particles feasible solution in each interval. Then normal distribution iterations are used to produce feasible
solution particles，so that the particle distribution of the feasible solutions gradually approaches or is equal to 0-1 nonlinear programming and get the optimal solution.

Key words: 0-1 nonelinear programming, particle filter, probability distribution