Spread Simulation Of Forest Disease And Pest Based On Multi-agent System

Forest disease and pest has caused enormous losses to forestry, ecology, society. On the basis of study on the multi-agent simulation theory and methodology, proceeding from bio-ecological characteristics of forest pests and according to the occurrence and spread of disaster and complex process of its impact on forest ecosystem, this paper constructs a spatial prediction and dynamic simulation model of disaster spread based on multi-agent, and uses NetLogo software platform to implement time-space prediction and spread process simulation of disaster. The research will rise disaster prediction from traditional time scale to space scale and from static analysis to dynamic tracking simulation, solve time-space scale effects problem of disaster spread, expand the application of multi-agent technique to forest visualization, and develop a new method for forest pests prediction.The main research contents and results in this paper are the following:(1) According to the occurrence and spread of disaster and complex process of its impact on forest ecosystem, the impact factors of disaster's occurrence and spread are analyed and summarized, including bio-ecological characteristics, topography, stand structure, climate, time and human activities. Research finds that, the occurrence and spread of disaster is affected more of less by these factor but dominant factors are different(2) On the basis of studying and exploring the multi-agent simulation theory and methodology, combined with the impact factors of the occurrence and spread of forest pests, a multi-agent simulation model framework of forest pests spread is constructed, and the model is implemented by NetLogo software platform.(3) Taking the Bursaphelenchus xylophilus of Sanming city in Fujian province as the study object, determine the condition and spatial distribution of disaster's occurrence, find spread rules and dynamically simulate disaster spread process. Finally, analyse the simulation results. The results indicate:first, vector insect, of which ability to fly is limited, can't carry disease to isolated Pinus massonian Lamb; second, the spread of Bursaphelenchus xylophilus Nickle is related to temperature and rainfall, positive correlative with temperature and negative with rainfall, which shows high temperature and dry contributes to spread of disease; simultaneously, multi-agent system is fit for simulation of forest complicated system, and thus a new realistic technique is provided for insect disease prevention.