Algorism

Abstract: The Decision Tree is an important classification method in data mining classification. Aiming at deficiency of ID3 algorism, a new improved classification algorism is proposed in this paper. The new algorithm combines principle of Taylor Formula with information entropy solution of ID3 algorism, and simplifies the information entropy solution of ID3 algorithm, then assigns a weight value N to simplified information entropy. It avoids deficiency of ID3 algorism which is apt to sample much value for testing. The improved algorithm is applied in score analysis and analyzed through experiment. The experiment results show that simplified entropy weight algorism spends decrease 65 Seconds compares ID3 algorithm in building up decision tree, and the accuracy was increased by 3%.

Abstract: appropriate instruction from the ground system as well as the algorism of an on-board software should be well considered to guarantee a sure execution of On-orbit service such as refueling, repairing and re- an on-board vision feed-back under rapidly orbiting is highly required for space activities. For changing on-orbit lighting condition, since the on-orbit service, it is one of the most important and computing power of a space qualified is most difficult technique to capture a “customer limited. satellite” by a robot arm which can move dynamically in a wide range of space. An on-board autonomy “machine vision” is highly required to perform this dynamic task, because a control loop including the ground station tends to be unstable because of the communication time delay between a ground station and a space robot. An experiment was conducted using Japanese Engineering Test Satellite VI1 (ETS-VII) [ 13 to capture a free floating satellite on orbit first time in the world. Successful results of the experiment are presented in this paper.

Abstract: Shipping companies consider most of the ship turnaround time as a critical factor when selecting a calling port for reducing costs. Container terminal operator tries to improve the port quality of service requested by a shipping line to maximize the productivity of container terminals. Therefore, this thesis proposes RFID (Radio Frequency Identification)-based RTLS (Real Time Location System) for enhancing the performance of terminal operation system. The suggest container terminal's yard allocation for ubiquitous computing is based on current yard allocation algorism. What is difference from the existed algorism is that it shall not apply container-loading sequence to each of independent containers, but shall apply to each of group in use the RFID based RTLS. In addition, pre-planning based on ubiquitous computing environment will support working of gantry cranes and yard tractors, and shorten ship turnaround time, more efficiently. It can accomplish the whole lead-time of port logistics.

Abstract: The invention discloses a three-dimensional visualizing method of hyperlarge scale of spatial data under network environment, which comprises the following steps: 1. realizing rapid acquisition of spatial data in the distributed isomerization environment through data allocating model and data conversion model; 2. adopting ellipsoidal four-tine tree and progressive tree technique to realize high-effective index of spatial data; 3. integrating composite technique with high-effective compression and release, dynamic flow media technique, external progressive transmitting algorism to satisfy the need of real-time transportation of large amount of spatial data; 4. making the spatial information form a series of layer again to be simplified into two video relative geometric buffer; displaying dynamic multi-distinguishability in the GPU.