Font Size: a A A

The Research And Application On The Biological Immunity Based Dynamic Risk Identification Model

Posted on:2012-06-23Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y TaoFull Text:PDF
GTID:1118330335981815Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
As one of many typical biological systems with complex functions in the natural world, biological immune system provides a source of inspiration for effectively solving complex problems. Deeply studying immune theory provides a good research foundation for generating new intelligent algorithms and models. In order to meet the requirement of dynamic risk identification, this paper established biological immunity based dynamic risk identification model (BIDRIM) and related immune algorithms on the basis of understanding and studying operation mechanism of biology immune system. Experiments on the benchmark data and successful application in the Shield and Out of Tunnel Risk Control System show that the proposed model and algorithms have good performance.This paper has carried out the research on the immune identification, immune optimization, immune synergism and many other studies based on analyzing the characteristics of dynamic risk. The research proposed a new design concept for dynamic risk identification.The main works and innovations are as follows:1. After futher studying running mechanisms of biology immune system, immune identification, immune optimization and immune synergism, bionic mechanisms are extracted, and biological immunity based dynamic risk identification model BIDRIM is proposed.2. A residual antigen based dynamic memory algorithm (RADMA) is proposed. RADMA effectively solves identification of the small probability event and improves identification accuracy of data samples containing missing attributes. A cell death recognition theory based dynamic risk identification algorithm (CDDRIA) is proposed. CDDRIA improves the classification accuracy and enhances the learning capacity of the unknown antigen.3. A pareto-based multi-object evolutionary immune algorithm (PMEIA) as well as a dynamic multi-object evolutionary immune algorithm (DMEIA) which is more suitable for tracking environmental change is proposed.4. A mechanism of members synergism based on biological immune system (MMS) and a mechanism of functions synergism based on biological immune system (MFS) are proposed, which realize optimization of system deployment and optimization of functional synergy of system.5. A shield and out of tunnel risk control system (TRCS) which takes BIDRIM as the core is constructed, and TRCS is successfully applied in shield and out of tunnel engineering.
Keywords/Search Tags:Dynamic Risk Identification Model, Immune Identification, Immune Optimization, Multi-object Optimization, Immune Synergism, Shield and Out of Tunnel Risk Control System
PDF Full Text Request
Related items