Progress 10/01/10 to 09/30/11
Outputs
Progress Report Objectives (from AD-416) The objectives of this Specific Cooperative Agreement are to characterize cotton fiber length distribution, to establish the relationship between the original fiber length distribution and the length distribution measured by High Volume Instrument (HVI), to explore non-traditional cotton length parameters that may lead to better prediction of cotton quality, and to study fundamental theories of fiber length determination from rapid testing by use of fiber beards. Approach (from AD-416) In order to obtain accurate measurements of the fiber length distribution, the relationship between the length of specimen fibers actually measured by the instrument and the original sample population has to be established. Three length distributions from specimen fibers, namely fibers selected by the HVI clamp, fibers held in the clamp (invisible for the instrument to measure) and fibers projecting from the clamp (actually measured by the instrument) will be obtained by AFIS (Advanced Fiber Information System) instrument. The relationships among the three distributions and the distribution of the original sample will be analyzed. Various non-traditional statistical parameters of cotton fiber length distributions will be obtained. Algorithm will be developed to obtain Lower Half Mean Length (LHML) that is a good parameter alterative to short fiber content (SFC). The effectiveness to predict cotton processability and yarn properties by use of these non-traditional length parameters will be studied. A new mathematical modeling of cotton fiber length distributions was successfully established by using the mixed Weibull function. This model was shown to fit the cotton fiber length distributions well, based on the data of fiber length distributions measured by using the Advanced Fiber Information System (AFIS). Such a model also enables us to study the length distributions of cotton samples used in the beard test method such as the High Volume Instrument (HVI). Non-linear regression models with different theoretical distributions such as normal distribution were constructed, and then the Gauss-Newton algorithm and the least squares principle were employed to solve the models and search for the probability density functions (PDF) that match the PDFs of the AFIS measured data. Computation results showed that a mixture of two two- parameter Weibull distributions fits the data very well. The Kolmogorov- Smirnov goodness-of-fit test was performed to verify fitness of the mixture Weibull distributions to the experiment data. The results showed that the parameter obtained from the experimental data and that obtained from the mixed Weibull distribution matched well. Based on the work in modeling cotton fiber length distributions by the mixed Weibull function, algorithms were developed to use the Partial Least Squares (PLS) regression to obtain the original fiber length distribution. In fiber length measurement by the rapid method of testing fiber beards instead of testing individual fibers, only the fiber portion projected from the fiber clamp can be measured. The length distribution of the projecting portion is very different from that of the original sample. The original fiber length distribution of a cotton was modeled by a five-parameter mixed Weibull distribution, and the length distribution of the projecting portion was also modeled by a five- parameter mixed Weibull distribution with parameters different from the original sample. A new approach was used to infer the distribution of the actual fiber length based on the length distribution of the observed projecting portion. The approach is the Partial Least Squares (PLS) regression. PLS is a recent statistics technique, which is less restrictive and can better handle situations such as small sample size and multi-collinearity. Length distributions obtained by the PLS algorithm showed a good match with experimental data. Comparisons of some commonly used length quality parameters between experimental data and PLS results also provide a good support of the proposed approach as well. The methods used to monitor activities for this agreement were annual reports, technical visits/e-mails/interactions, presentations at scientific and industry meetings, and publications.
Impacts
(N/A)
Publications
|
Progress 10/01/09 to 09/30/10
Outputs
Progress Report Objectives (from AD-416) The objectives of this Specific Cooperative Agreement are to characterize cotton fiber length distribution, to establish the relationship between the original fiber length distribution and the length distribution measured by High Volume Instrument (HVI), to explore non-traditional cotton length parameters that may lead to better prediction of cotton quality, and to study fundamental theories of fiber length determination from rapid testing by use of fiber beards. Approach (from AD-416) In order to obtain accurate measurements of the fiber length distribution, the relationship between the length of specimen fibers actually measured by the instrument and the original sample population has to be established. Three length distributions from specimen fibers, namely fibers selected by the HVI clamp, fibers held in the clamp (invisible for the instrument to measure) and fibers projecting from the clamp (actually measured by the instrument) will be obtained by AFIS (Advanced Fiber Information System) instrument. The relationships among the three distributions and the distribution of the original sample will be analyzed. Various non-traditional statistical parameters of cotton fiber length distributions will be obtained. Algorithm will be developed to obtain Lower Half Mean Length (LHML) that is a good parameter alterative to short fiber content (SFC). The effectiveness to predict cotton processability and yarn properties by use of these non-traditional length parameters will be studied. A mathematical model (the mixed Weibull function) was shown to fit the cotton fiber length distributions well, based on the data of fiber length distributions measured on the Advanced Fiber Information System (AFIS). Weibull distribution is a continuous probability distribution named after Waloddi Weibull. A traditional Weibull distribution has two parameters, a shape parameter and a scale parameter. This model also enables us to study the length distributions of cotton samples used in the beard (tapered fiber bundle) test method such as the High Volume Instrument (HVI). Non-linear regression models with different theoretical distributions were constructed, and then the least squares principle was employed to solve the models and search for the probability density functions (PDF) that match the PDFs of the AFIS measured fiber length data. Computation results showed that a mixture of two two-parameter Weibull distributions fits the data very well. A goodness-of-fit statistical test was performed to verify fitness of the mixture Weibull distributions to the experiment data. The results showed that the parameter obtained from the experimental data and that obtained from the mixed Weibull distribution matched well. Based on the work in modeling cotton fiber length distributions by the mixed Weibull function, algorithms were developed to use the Partial Least Squares (PLS) regression to obtain the original fiber length distribution. In fiber length measurement by the rapid method of testing fiber beards instead of testing individual fibers, only the fiber portion projected from the fiber clamp can be measured. The length distribution of the projecting portion is very different from that of the original sample. The original fiber length distribution of a cotton was modeled by a five-parameter mixed Weibull distribution, and the length distribution of the projecting portion was also modeled by a five-parameter mixed Weibull distribution with parameters different from the original sample. A new approach was used to infer the distribution of the actual fiber length based on the length distribution of the observed projecting portion. The approach is the Partial Least Squares (PLS) regression. PLS is a recent statistics technique, which is less restrictive and can better handle situations such as small sample size and independent variables that may be correlated with each other. Length distributions obtained by the PLS algorithm showed a good match with experimental data. Comparisons of some commonly used length quality parameters between experimental data and PLS results also provide a good support of the proposed approach as well. The methods used to monitor activities for this agreement were onsite visits, email and telephone communications, progress reports, presentations at scientific and industry meetings, and publications.
Impacts
(N/A)
Publications
|
Progress 10/01/08 to 09/30/09
Outputs
Progress Report Objectives (from AD-416) The objectives of this Specific Cooperative Agreement are to characterize cotton fiber length distribution, to establish the relationship between the original fiber length distribution and the length distribution measured by High Volume Instrument (HVI), to explore non-traditional cotton length parameters that may lead to better prediction of cotton quality, and to study fundamental theories of fiber length determination from rapid testing by use of fiber beards. Approach (from AD-416) In order to obtain accurate measurements of the fiber length distribution, the relationship between the length of specimen fibers actually measured by the instrument and the original sample population has to be established. Three length distributions from specimen fibers, namely fibers selected by the HVI clamp, fibers held in the clamp (invisible for the instrument to measure) and fibers projecting from the clamp (actually measured by the instrument) will be obtained by AFIS (Advanced Fiber Information System) instrument. The relationships among the three distributions and the distribution of the original sample will be analyzed. Various non-traditional statistical parameters of cotton fiber length distributions will be obtained. Algorithm will be developed to obtain Lower Half Mean Length (LHML) that is a good parameter alterative to short fiber content (SFC). The effectiveness to predict cotton processability and yarn properties by use of these non-traditional length parameters will be studied. Significant Activities that Support Special Target Populations The methods used to monitor activities for this agreement were onsite visits, email and telephone communications, progress reports, presentations at scientific and industry meetings, and publications. As a natural product, cotton fiber length has a distribution. If the entire length distribution of a sample is known, any desired quality parameter can be calculated. This research is to obtain the length distribution from the rapid method of testing fiber beards. Statistical models have been established to infer the length distribution from beard test data. Statistical length parameters calculated from the inferred distribution showed good agreements with those from the test data. A larger sample set has been selected and experiments are in progress. The new data will be used to refine and validate the model. The algorithms that are in our on-going research include Multivariate Multiple Regression and Structural Equation Modeling mathematics models.
Impacts
(N/A)
Publications
|
Progress 10/01/07 to 09/30/08
Outputs
Progress Report Objectives (from AD-416) This is a component of a Cooperative Agreement funded by Cotton Incorporated. The objectives of this Specific Cooperative Agreement are to characterize cotton fiber length distribution, to establish the relationship between the original fiber length distribution and the length distribution measured by High Volume Instrument (HVI), to explore non-traditional cotton length parameters that may lead to better prediction of cotton quality, and to study fundamental theories of fiber length determination from rapid testing by use of fiber beards. The Agreement will be reviewed for possible extension in 2007. Approach (from AD-416) In order to obtain accurate measurements of the fiber length distribution, the relationship between the length of specimen fibers actually measured by the instrument and the original sample population has to be established. Three length distributions from specimen fibers, namely fibers selected by the HVI clamp, fibers held in the clamp (invisible for the instrument to measure) and fibers projecting from the clamp (actually measured by the instrument) will be obtained by AFIS (Advanced Fiber Information System) instrument. The relationships among the three distributions and the distribution of the original sample will be analyzed. Various non-traditional statistical parameters of cotton fiber length distributions will be obtained. Algorithm will be developed to obtain Lower Half Mean Length (LHML) that is a good parameter alterative to short fiber content (SFC). The effectiveness to predict cotton processability and yarn properties by use of these non-traditional length parameters will be studied. Significant Activities that Support Special Target Populations The methods used to monitor activities for this agreement were onsite visits, email and telephone communications, progress reports, presentations at scientific and industry meetings, and publications. If the entire length distribution of a sample is known, any desired quality parameter can be calculated. Currently, the High Volume Instrument (HVI) reports Upper Half Mean Length (UHML) and Uniformity Index (UI), but not the length distribution. Cooperatively with the Mathematic Department of University of New Orleans, we are continuing our efforts to obtain the length distribution from the HVI test results. The HVI does not measure fiber from one end to the other; instead, it only measures the fiber portion projecting from the test beard. Currently we are using various statistical methods to infer the Weibull distribution parameters of original cotton from the Weibull parameters of the HVI sampled cotton fiber. (Weibull distribution is named after its inventor Waloddi Weibull and is a widely used probability distribution function). The algorithms that are in our on-going research include Multivariate Multiple Regression and Structural Equation Modeling.
Impacts
(N/A)
Publications
|
Progress 10/01/06 to 09/30/07
Outputs
Progress Report Objectives (from AD-416) This is a component of a Cooperative Agreement funded by Cotton Incorporated. The objectives of this Specific Cooperative Agreement are to characterize cotton fiber length distribution, to establish the relationship between the original fiber length distribution and the length distribution measured by High Volume Instrument (HVI), to explore non-traditional cotton length parameters that may lead to better prediction of cotton quality, and to study fundamental theories of fiber length determination from rapid testing by use of fiber beards. The Agreement will be reviewed for possible extension in 2007. Approach (from AD-416) In order to obtain accurate measurements of the fiber length distribution, the relationship between the length of specimen fibers actually measured by the instrument and the original sample population has to be established. Three length distributions from specimen fibers, namely fibers selected by the HVI clamp, fibers held in the clamp (invisible for the instrument to measure) and fibers projecting from the clamp (actually measured by the instrument) will be obtained by AFIS (Advanced Fiber Information System) instrument. The relationships among the three distributions and the distribution of the original sample will be analyzed. Various non-traditional statistical parameters of cotton fiber length distributions will be obtained. Algorithm will be developed to obtain Lower Half Mean Length (LHML) that is a good parameter alterative to short fiber content (SFC). The effectiveness to predict cotton processability and yarn properties by use of these non-traditional length parameters will be studied. Significant Activities that Support Special Target Populations This report serves to document research conducted under a Reimbursable Cooperative Agreement between Agricultural Research Service (ARS) and University of New Orleans. Additional details of research can be found in the report for the in-house project 6435-44000-069-00D, �Improved Cotton Quality Measurements.� The methods used to monitor activities for this agreement were onsite visits, email and telephone communications, progress reports, presentations at scientific and industry meetings, and publications. Cooperatively with the Mathematic Department of University of New Orleans, we searched for the mathematic functional description for cotton fiber length distribution. Eight cotton samples with a wide range of fiber length are selected and tested on the Advanced Fiber Information System (AFIS) instrument. The measured single fiber length data are used for finding the underlying theoretical distribution of the cotton. It is found that none of the known distributions yielded satisfactory fit, but a mixture of two Weibull distributions fits the data very well. (Weibull distribution is named after its inventor Waloddi Weibull and is a widely used probability distribution function). Fiber length distributions by number and by weight are studied separately, and in both cases a mixed Weibull distribution shows a good fit to the data. Numerical comparisons for various parameters between the original distribution from the data and the fitted distribution were performed. We are applying this technique to a larger data set to explore the relationships among the parameters of the Weibull distributions and the qualities of cotton fiber and yarns.
Impacts
(N/A)
Publications
|
|