| Plantation resource is an important part of the forest resource in China,and it plays an irreplaceable role in providing forest products and improving the ecological environment,etc.According to the eighth national forest resources inventory,the area of plantation forests in China is 69.33 million hectares,which ranks the first place in the world,but the tree volume is only 52.76 m3 per hectare and mean DBH(Diameter at Breast Height)is 13.6cm,which is far below the world’s average.How to manage the plantation resource scientifically and effectively is the key to improve forest productivity.Forestry information is an important means in the sustainable development of forestry modernization,and also a component of current national informatization strategy.Larch is an important plant tree species in China,which are mass cultivated in the northeast,Inner Mongolia and other places.However,research of larix plantations management by information technology is not deep enough and comprehensive.It is worthy of studying to scientifically and effectively manage the larix plantations.In this paper,Larix principis-rupprechtii as research object,forestry,information technology and statistical method were utilized for research and develop the forest management decision support system,whose core is the subcompartments management work plans preparation assistant decision.And taking subcompartments tending work plans as the main research content.That provides decision support for the management of Larix principis-rupprechtii.Based on subcompartments survey data of Huamugou forestry farm,in Chifeng City,Inner Mongolia,and the survey data of the temporary standard plot and the survey data of the fixed standard plot,competition index model and crown model of Larix principis-rupprechtii plantations were studied,and the theory and technology of intelligent selection tending subcompartments as well,then it developed a Assistant Decision Support System for Subcompartment Tending Feling Plan Design(TF-ADSS).Main research work and major results for this study are as follows:1.Constructing Crown competition index.Competitions among trees are not only from horizontal extrusion but also mutual occlusion vertically.The simulations of competition between trees are realized by choosing several common competitive indexes,which are dependent on distance and independent on distance,respectively.And based on crown overlap area index AO-CIO,a factor,crown length,was introduced as the influence factor of occlusion between object tree and competition tree on the vertical direction of the canopy.2.Based on Voronoi diagram and 4 trees spatial competition structure units formed,respectively,the selection methods and algorithms of the object tree and competition trees were studied.This thesis put forward quadrant edge correction method which could increase the utilization rate of trees effectively,especially to small plot.These algorithms were designed with the R language.3.Ten competition index models of the above two spatial competition structure units were analyzed and compared with linear expression,exponential expression,logarithm,power function,hyperbola and binomial.The DBH and 20 competition indices were fitted respectively.The results show:(1)when fitting competition index and DBH of Larix principis-rupprechtii,binomial,exponential and power functions performed better.(2)the determination coefficient of Voronoi spatial competition structural unit was larger than that of 4 trees spatial competition structural units generally.Relatively,the Voronoi spatial competition structural unit is more suitable for competition index than 4 trees spatial competitions structural unit.(3)Raising the competition index model of influencing factors in vertical direction of tree crown can obviously improve the influence of competition index on DBH;(4)the competition index models which depend on the distance could show the relationship between DBH and competition index better than those independent of distance.That is to say,it could show competitive capacity among trees better,but the difference was not obvious,relative to competition index model specifically.4.Based on the study of tree competition,in the view of the canopy and spatial,the interaction between forest canopies,trees and tending felling of forest stand was analyzed,which provided forest thinning decision with references.Assuming the row spacing of Larix principis-rupprechtii plantation forest is corresponded,and the crown is projected into a positive circle.The condition of tending felling is analyzed into three aspects as follows:(1)When the sum of the crowns of the two trees is less than or equal to the distance between the two trees,it is not necessary to carry out the tending felling of the two trees.(2)When the distance between the two trees is equivalent to the crown radius,it must be tending felling.(3)When the distance between the two trees between the above two,other ways can be used to determine whether to tending felling.5.Based on the study of the competition between the trees of Larix principis-rupprechtii,the 10 competition indices of the two spatial structural units were used to construct the crown model.The results showed that when the competition model was determined by the Voronoi,the crown performance of CI10 was the best.The crown model with Sum Line Length competition index CI3 is the best when the competition unit is determined by four trees.At the same time,using the stepwise regression analysis method to establish the crown model,the independent variables and dependent variables of several treatment methods were calculated and analyzed,for the independent variables and logarithmic form,both cases are likely to exist in the model,and when they appear together,the model would have better fitting performance.6.Based on the study above,this paper,under the control of the task of forest tending felling,adopted spatial analysis and genetic algorithm,explored the intelligent selection method of forest subcompartments tending felling,and provided decision support for follow-up forest management activities such as tending felling plan design based on subcompartment.There are three main aspects of innovation:(1)Proposed and constructed the evaluation index system of subcompartments selection for forest tending felling,which can also be extended to the intelligent selection of other forest subcompartments.(2)Optimized the point buffer analysis,and designed Annulus Control Algorithm(ACA)which can also be used in other spatial analysis.(3)An improved greedy strategy was utilized to encode the symbolic coding method,combined with the operator of improved genetic algorithm,genetic algorithm is improved and a new genetic algorithm IGSEGA(Improved Greedy Strategy and Encoding for Genetic Algorithm)was formed.It can be applied to other application requirements at the same time.7.Aiming at the problem of designing forest tending felling subcompartments selection and subcompartments tending felling operation management plan assistant design,this thesis studied intelligent selection for subcompartments which need for the design of the working plan,and suitability evaluation and reasoning method of subcompartment management operation scheme.By experts of forest management analyzing on forest management plan knowledge,It constructed the knowledge base and inference engine of the TF-ADSS,and adopted the framework of representation and generative rule representation according to the stand state,site conditions and management objectives and other conditions to determine the types of business knowledge.Plan generation technology was also studied,and two design methods were designed.That is "intelligent design" and "manual aided design".8.Based on the analysis of the knowledge of forest management decision-making experts,the TF-ADSS knowledge base,model base,method base and forestry database were constructed according to the forest management knowledge such as site conditions,tree species,forest management operation methods and the types and methods of tending felling,forest management planning.The data visualization technology is based on.NET platform and R program language were studied.9.On the basis of the above research,the system architecture and function structure of the TF-ADSS were designed according to the forest management decision-making process,which was based on the.NET platform,C#program language,R program language,combined with ArcGIS Engine component development kit.The function of the system is mainly to provide decision support for the making of forest management work plans for tending felling work in forest subcompartments. |
PDF Full Text Request