Abstract: To improve the shifting comfort, the peak of the turbine speed change rate, the average rate of the turbine speed change and gear are determined as three objective evaluation indices, which are defined in detail, and an objective description of shift quality evaluation for tracked vehicle equipped with integrated automatic transmission is given using back propagation neural network (BPNN) according to its basic principle. A BPNN model used for evaluating shift quality is established and trained by using the BPNN toolbox in Matlab/Simulink software, and an objective evaluation shift software is designed specially using GUI box in Matlab environment. The comparison of the test results with the subjective ratings shows that the developed evaluation method has a better consistency with the traditional one, which can also enhance the objectivity, pertinency and flexibility of this system.

Key words: ordnance science and technology, integrated transmission, shifting evaluation, back propagation neural network