Memetic algorithms combine evolutionary algorithms with local search heuristics in order to speed up the evolutionary process. The bacterial memetic algorithm applies the bacterial operators instead of the genetic algorithm's crossover and mutation operator. In this paper a novel diversity induction method is proposed for bacterial memetic algorithm. Hibernation of some individuals in the population is applied for avoiding premature convergence to local optima. This operation can help in turning back from unpromising region of the search space by previously found solutions represented in hibernated bacteria. © 2012 IEEE.