But for large scale or complex problems, the conventional monolithic evolving methods show poor performance in learning speed and scalability.

The paper analyzes the existing algorithms and their strengths,weaknesses and the problems,depending on the learning accuracy and learning speed;elaborates the quota status of the incremental learning and gives the following research direction.

The comparison experiments show that the proposed algorithm is as good as the learning algorithm of the traditional SVM in accuracy,but much better in learning speed.

Selectivity obstacle_avoidance research based on relative speed

By using auxiliary triangle method,analytic result for the relative speed on four quadrants of the sphere and the computer program were brought out,then 3D curved faces were plotted to illustrate the speed distribution.

Based on the experimental data processed by the MATLAB language, the relative speed distribution curve and vectorgraph are plotted.

A car-following model with stochastically considering the relative velocity in a traffic flow;

The essential view is that the relative velocity of the burning gas directs at the rocket after the burning gas has ejected, rather than the rocket before the burning gas has ejected.

The relative attenuation coefficient and relative phase velocity are investigated for the microstructures with different content of pearlite,ferrite and upper bainite.