Abstract: Some improvements were presented for a basic evolving recurrent neural network system．Firstly‚a switchable fitness evaluating function was proposed to maintain the sensitivity of fitness function in training er-ror‚and ensure the exact and effective duplication of fine individuals by selecting operation．Next‚aiming at the slight mutation intensity of uniform mutation on individuals‚a concentrative mutation method with variable neighborhood length was introduced to improve the system abilities to keep population diversity and to find fine individuals．Then‚based on the relationship between the fitness of individual and the mean fitness of population‚an adaptive adjustment strategy of mutation step size was proposed．Finally‚by applying the Hamming distance among individuals‚the elitist keeping strategy was also improved to stop the fittest individual from reduplicating itself in population．Simulation results show that the evolving recurrent neural network system with all these improvements is of more excellent performance．

Key words: evolving recurrent neural network, fitness evaluating function, mutation