Study On Site Quality Evaluation And Afforestation Model Design Technology Of Chinese Fir In Fujian

With the deepening of ecological civilization construction and the development of "land greening action",afforestation has become the primary task of forestry development in China.The core of afforestation design is to compile afforestation models for reproducing area in afforestation projects.However,at present,the experts and technicians of afforestation design in grass-roots units of our country are seriously inadequate and rely on production experience to afforestation and lack of scientific management.This situation has caused China's forestry for a long time to be in a state of low intensification,extensive cultivation and management,and backward management.In this study,the site quality of plantation forest was evaluated by machine learning method.On this basis,the woodland was selected by intelligent selection algorithm for afforestation project,and the design technology of afforestation model was studied and the design of information system assisted afforestation model was established.This paper studies the key technologies of site quality evaluation,afforestation field intelligence selection and afforestation pattern design in the process of afforestation model design,and designs and realizes the afforestation model auxiliary design system,which provides service for afforestation pattern design,and has certain practical value.The main research contents of the paper are as follows:1.The site quality evaluation method based on machine learning is studied.By quantifying the site factors,using the quantitative theory I method and the radial basis neural network method to predict the dominant height of the afforestation land,the status index is obtained by using the reference age of Chinese fir,and the site quality of the forestland is evaluated.When using the quantitative theory I method to evaluate the status index,the factors with higher scores in the elevation factor category are elevation,soil thickness and geomorphology.Through the correlation matrix,it can be seen that the higher correlation between the average height of the dominant wood and the 7 site factors is the slope,the geomorphology and the soil thickness.On the basis of the RBF neural network model,the optimization of RBF neural network model by genetic algorithm is discussed,and the prediction accuracy of the neural network model is improved.The RBF neural network optimized by genetic algorithm has better prediction accuracy of dominant height.The site quality classification and evaluation technology based on BP_Adaboost algorithm and support vector machine is studied.On the basis of establishing the site quality classification evaluation model based on support vector machine,the kernel function parameter g and the penalty parameter C are optimized by cross validation,and the precision contour map is plotted.Through the experimental analysis,the improved radial basis neural network model based on genetic algorithm is more obvious compared with other methods.It is in good agreement with the evaluation results of the Chinese fir status index table.It provides a feasible site quality evaluation method for afforestation tree species without the status index table.2.On the basis of the study of site quality evaluation,this study proposes an intelligent selection algorithm for afforestation land based on ant colony algorithm and an intelligent selection algorithm for afforestation site based on watershed selection.Selecting reasonable and suitable forestation sites for afforestation projects is also the core issue in afforestation planning and design,and is the precondition for designing afforestation operation method.The algorithm of intelligent selection in afforestation area solves the problem of tree selection and land selection in afforestation project.The aim of this study is to obtain the best afforestation site selection scheme,which considers the multiple factors such as the site quality,area,the distance from the afforestation project center or the forest farm field,and abstracts the intelligent selection of the forestland into the backpack problem,and solves the optimal forestland selection scheme of the afforestation project by ant colony algorithm.In addition to the ant colony algorithm,this study also proposed a watershed selection algorithm,on the basis of site quality evaluation,by adaptively changing the basin buffer radius to select forestland.Experimental results show that watershed selection algorithm outperforms traditional selection algorithm,but it has no advantage compared with ant colony algorithm.Compared with the traditional selection algorithm,the ant colony algorithm has produced an average output of about 9%.The influence of the site quality and area on the value of the final output is more obvious.Ant colony algorithm is more inclined to choose the afforestation land with good site quality and large area,which is also in line with the experience of the Afforestation Planning and design experts.3.On the basis of the research on the intelligent selection of afforestation land,the related technology of afforestation assisted design was studied,and the auxiliary afforestation method was compiled,which provided technical support for the design and Realization of the afforestation method aided design system.Based on the results of site quality evaluation,on the basis of the reliability of the production rules and afforestation models,the afforestation model is calculated by using the repeatability and the knowledge credibility of the site string sequence,and the appropriate afforestation model is recommended for the afforestation site according to the recommended degree.According to the model of stand growth and harvest,the expected profit(stand diameter,tree height and volume)of the afforestation method is calculated and simulated directly in the computer.It is expressed in the form of two dimensional visualization,such as numeric and chart,and the assistant afforestation method is compiled.The design techniques of afforestation plan and afforestation plan are studied and studied.The afforestation operation method is designed and the afforestation work method is exported.4.The knowledge base construction,afforestation knowledge acquisition and forestation knowledge retrieval technology were studied.In this study,we set up a knowledge acquisition module for afforestation,using the theme crawler of afforestation to capture the literature and knowledge about Afforestation on the Internet,and to carry out knowledge de-noising,knowledge extraction,knowledge evaluation,knowledge removal,and enrich the knowledge base of afforestation,and provide reference for the design of afforestation method.The index database of afforestation knowledge is established,and fast retrieval of forestation knowledge base is achieved by segmentation algorithm and full-text search algorithm.5.On the basis of the above research,the overall architecture,functional structure and workflow of the afforestation assistant design system are designed,the model base and the method library are constructed.On the basis of the model base building,the IronPython combined with the inverse Poland formula is used to analyze the model.On this basis,the prototype of the auxiliary design system for afforestation operation method is realized and the corresponding operation example is given.