• Entrar
    Ver item 
    •   Página inicial
    • BIBLIOTECA DIGITAL: Teses & Dissertações
    • Teses & Dissertações
    • Ver item
    •   Página inicial
    • BIBLIOTECA DIGITAL: Teses & Dissertações
    • Teses & Dissertações
    • Ver item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Métodos usuais e propostos para expressar perfis dendrométricos e determinar volumes individuais

    Thumbnail
    Visualizar/Abrir
    R - T - ARTUR LOURIVAL DA FONSECA MACHADO.pdf (2.284Mb)
    Data
    2009
    Autor
    Machado, Artur Lourival da Fonseca
    Metadata
    Mostrar registro completo
    Resumo
    Resumo: Neste trabalho foram propostas alternativas à construção de funções de afilamento para árvores individuais, visando perfis monotônicos ou com forma preservada por funções spline cúbica, potência com expoente variável e segmentada de Max e Burkhart além de novas fórmulas para cubagem de toras por sólidos de revolução de polinomiais parabolóide (Artur-VP2), cúbico (Artur-VP3), com diâmetros de toras adjacentes (Artur - VP3A e AP3A), além de alternativas de médias geométricas às fórmulas de Newton e de Smalian. Testes foram realizados para dados de 1367 toras de 1m de comprimento com volumes reais obtidos por Xilômetro para 52 árvores de Pinus elliottii L. coletados na Estação Experimental do Canguiri da Universidade Federal do Paraná. Os métodos foram avaliados por meio do erro absoluto percentual médio (EAPM) na estimativa de volumes. A ANOVA foi aplicada para os fatores tipo de volume, método e comprimento de tora. Os métodos de Huber e de Huiquan Bi foram classificados pelo teste de Scott-Knott como de melhor ranking em todos os agrupamentos, com EAPM para: todos os volumes, de 4,14% e 4,33%; laminação (diâmetro mínimo de 25 cm na ponta fina), de 5,61% e 5,57%; serraria (15 cm), de 4,26% e 3,78%; polpa (8 cm), de 5,91% e 5,86%; volumes totais, de 3,28% e 3,16% e totais comerciais, de 3,26% e 3,04%, respectivamente; foi incluída a fórmula geométrica ponderada, para volumes totais e, além desta, a função spline de Lagrange, para totais de volumes comerciais; para tais métodos as variações de EAPM foram menores que 1,8% para diferentes comprimentos de tora e tipos de volume. Para serraria e polpa, nove e doze métodos foram considerados homogêneos para EAPM variando de 3,54% a 4,59% e de 4,55% a 6,5%, respectivamente. A função de Huiquan Bi estima exatos a altura total e o diâmetro à altura do peito (DAP). A função spline de Lagrange, ajustada com correção nas extremidades das classes de sortimento, exceto para volumes totais, foi classificada entre primeiro e segundo grupos de melhor desempenho. A fórmula Artur-VP3A, sem usar ponto médio, e a fórmula Artur-VP2 foram equivalentes ao método de Newton, de forma que se espera o mesmo da Artur - VP3A em relação a Artur-AP3A, com redução do número de termos. Em geral, as propostas de médias geométricas como modificações às fórmulas de Smalian e de Newton apresentaram melhor desempenho ou menores valores de EAPM que estas, exceto para volumes destinados a polpa, onde a geométrica teve menor desempenho que a de Smalian e a geométrica ponderada, maior EAPM que a de Newton. As fórmulas propostas, exceto a geométrica, foram classificadas no grupo de segundo melhor desempenho, para todos os tipos de volumes, exceto serraria e polpa, no grupo de melhor desempenho. A fórmula Artur - VP5 proporciona técnica de integração numérica para afilamento por polinomial do quinto grau. As funções spline cúbica de Wolberg e Alfy e not-a-knot representaram um terceiro grupo de desempenho, exceto serraria e polpa. A primeira deve ser a preferida, por apresentar preservação de forma e controle de oscilações. Em geral, os métodos superestimaram volumes destinados à laminação e subestimaram volumes destinados à serraria e polpa. A seleção de métodos adquire maior importância para volumes totais e laminação. Os parâmetros da função de Wolberg e Alfy foram estimados por meio de Programação Linear e Não-Linear e as funções potência e de Max e Burkhart, com o método de Levenberg-Marquardt. Os métodos foram implementados nos softwares Maple e R.
     
    Abstract: In the present work we proposed alternative taper functions for individual trees, emphasizing shape preserving profiles, through cubic spline functions, power functions with variable-exponent and the Max and Burkhart taper function; and new formulas for log volume evaluation built through solid of revolution of polynomial functions as paraboloid (Artur - VP2), cubic (Artur - VP3), and cubic using two adjacent logs diameters (Artur - VP3A and AP3A), as well as alternative geometric average for the Smalian and weighted geometric average for the Newton formula. Tests were accomplished for a dataset with1367 logs of 1m length and available real volumes obtained from Xylometer for 52 trees of Pinus elliottii L. collected in the Canguiri Experimental Station of the Federal University of Paraná. The methods were evaluated through the mean absolute percent error (EAPM) in volume estimating; the Analysis of Variance were applied for the factors method, log length and volume type. The Huber and Huiquan Bi methods were ranked better in all groupings through the Scott-Knot test, with EAPM of 4.14% and 4.33% for all volume types; 5.61% and 5.57% for veneer (25 cm minimum diameter); 4.26% and 3.78% for sawtimber(15cm); 5.91% and 5.86% for pulpwood(8cm); 3.28% and 3.16% for all volumes and 3.26% and 3.04% for total of commercial volumes, respectively. The weighted geometric formula was included in the better group for total volumes and, this and the Lagrange spline function for total commercial volumes. The EAPM range for the 4 log lengths and all volume types was less than 1.8%. For sawtimber and pulpwood there were nine and twelve methods ranked in a better homogeneous group, with EAPM varying between 3.54% to 4.59% and 4.55% to 6.5%, respectively. The Huiquan Bi function provided exact estimates for the total height and for the breast height diameter (BHD), for relative diameters being used. The Lagrange spline function was fitted with slope correction at every height of changing assortment class, except for total volumes, and classified among first and second performance groups. As the Newton formula performed close to the Artur-VP2, its solid of revolution version, the same is expected to the Artur-AP3A to perform close to the Artur-VP3A, with its advantages and smaller number of terms. In general, the proposed alternatives of geometric average and the weighted geometric average performed better than the Smalian formula and the Newton formula, respectively, except for pulpwood volumes, with the geometric in a second group. All the proposed formulas, except the geometric one, were classified in the second better performance group, for all the volume types, except for sawtimber and pulpwood, in the first group. The Artur-VP5 formula provides a numerical integration method for fifth degree polynomial taper functions. The Wolberg and Alfy and the not-a-knot cubic spline functions were ranked in a third performance group, except for the sawtimber and pulpwood volumes; the first must be preferred among the traditional spline functions, for its shape preserving and illogical oscillation control. In general, the methods overestimated the veneer volumes and underestimated the sawtimber and pulpwood volumes. The selection of methods acquires importance mainly for veneer and total volumes. The parameters of the Wolberg and Alfy spline function were fitted through Linear and Non-linear Programming techniques. The power function and the Max and Burkhart function were fitted with the Levenberg-Marquardt method. The methods were implemented mainly in the Maple software and R software.
     
    URI
    https://hdl.handle.net/1884/35973
    Collections
    • Teses & Dissertações [10564]

    DSpace software copyright © 2002-2022  LYRASIS
    Entre em contato | Deixe sua opinião
    Theme by 
    Atmire NV
     

     

    Navegar

    Todo o repositórioComunidades e ColeçõesPor data do documentoAutoresTítulosAssuntosTipoEsta coleçãoPor data do documentoAutoresTítulosAssuntosTipo

    Minha conta

    EntrarCadastro

    Estatística

    Ver as estatísticas de uso

    DSpace software copyright © 2002-2022  LYRASIS
    Entre em contato | Deixe sua opinião
    Theme by 
    Atmire NV