Mathematical/simulation model construction
Accurate data collection is essential for appropriate performance analysis and is the key in decision-making. However in
reality, due to a variety of reasons, accurate data are not always readily available or are very difficult to capture. Other difficulties,
such as the non-existence of processes in the “to-be” scenarios also pose great challenges for performance analysis.
Parameters provided by the industry or public domain data sometimes have to be used during the data collection phase. This
gives no guarantee of the accuracy of the performance analysis due to the difference in terms of operational environment,
operational efficiency, data consistency and even the accuracy of the data available. Sometimes, the data sources provided do
not differentiate the types of equipment used and might be at a high aggregated level. One example is in Roso's (2007) study
where the CO2 emission for trucks is estimated as 1 kg/km; while in the study by Liao, Tseng, and Lu (2009), CO2 emission is
estimated as 155 g/tonne km. Clearly, the second data source provides a more accurate measurement since it gauges both
weight and distance.
To remedy such disadvantages brought by these types of data, mathematical or simulation models can be constructed to
enable further investigation of the impacts of different variables. If we take container distribution as an illustrative example,
we can see that in order to transport a certain number of containers, different types of trucks can be used (e.g., B-double,
Super B-double, Super B-triple, etc.). It is possible to work out the number of trips that need to be taken to completely
transport these containers and then treat the CO2 emissions as a variable in mathematical models. This will facilitate the
analysis of the impact of the variable on the overall system.
Sometimes, it is not enough to only look at the mathematical model, since a lot of other factors are also changing. Again
using the container distribution example, it is important to take factors such as the volatile business environment (and
therefore not constant container flows), fast moving fuel prices (therefore no fixed cost for each trip), seasonality demand of
transport equipment (therefore a limitation to the available transport options) into consideration and assess the impacts of
these factors on the decision making process of transport network design. Mathematically modelling these types of changes
would almost be impossible since the level of complexity increases significantly if such uncertainties are to be taken into
account. Simulation models can be a substitute for mathematical models and allow the decision makers to play with a lot of
combinations of different parameters. Variables can still be embedded into the simulation models to enable further analysis
once the simulation runs are completed.
Mathematical/simulation model constructionAccurate data collection is essential for appropriate performance analysis and is the key in decision-making. However inreality, due to a variety of reasons, accurate data are not always readily available or are very difficult to capture. Other difficulties,such as the non-existence of processes in the “to-be” scenarios also pose great challenges for performance analysis.Parameters provided by the industry or public domain data sometimes have to be used during the data collection phase. Thisgives no guarantee of the accuracy of the performance analysis due to the difference in terms of operational environment,operational efficiency, data consistency and even the accuracy of the data available. Sometimes, the data sources provided donot differentiate the types of equipment used and might be at a high aggregated level. One example is in Roso's (2007) studywhere the CO2 emission for trucks is estimated as 1 kg/km; while in the study by Liao, Tseng, and Lu (2009), CO2 emission isestimated as 155 g/tonne km. Clearly, the second data source provides a more accurate measurement since it gauges bothweight and distance.To remedy such disadvantages brought by these types of data, mathematical or simulation models can be constructed toenable further investigation of the impacts of different variables. If we take container distribution as an illustrative example,we can see that in order to transport a certain number of containers, different types of trucks can be used (e.g., B-double,Super B-double, Super B-triple, etc.). It is possible to work out the number of trips that need to be taken to completelytransport these containers and then treat the CO2 emissions as a variable in mathematical models. This will facilitate theanalysis of the impact of the variable on the overall system.Sometimes, it is not enough to only look at the mathematical model, since a lot of other factors are also changing. Againusing the container distribution example, it is important to take factors such as the volatile business environment (andtherefore not constant container flows), fast moving fuel prices (therefore no fixed cost for each trip), seasonality demand oftransport equipment (therefore a limitation to the available transport options) into consideration and assess the impacts ofthese factors on the decision making process of transport network design. Mathematically modelling these types of changeswould almost be impossible since the level of complexity increases significantly if such uncertainties are to be taken intoaccount. Simulation models can be a substitute for mathematical models and allow the decision makers to play with a lot ofcombinations of different parameters. Variables can still be embedded into the simulation models to enable further analysisonce the simulation runs are completed.
การแปล กรุณารอสักครู่..
Mathematical/simulation model construction
Accurate data collection is essential for appropriate performance analysis and is the key in decision-making. However in
reality, due to a variety of reasons, accurate data are not always readily available or are very difficult to capture. Other difficulties,
such as the non-existence of processes in the “to-be” scenarios also pose great challenges for performance analysis.
Parameters provided by the industry or public domain data sometimes have to be used during the data collection phase. This
gives no guarantee of the accuracy of the performance analysis due to the difference in terms of operational environment,
operational efficiency, data consistency and even the accuracy of the data available. Sometimes, the data sources provided do
not differentiate the types of equipment used and might be at a high aggregated level. One example is in Roso's (2007) study
where the CO2 emission for trucks is estimated as 1 kg/km; while in the study by Liao, Tseng, and Lu (2009), CO2 emission is
estimated as 155 g/tonne km. Clearly, the second data source provides a more accurate measurement since it gauges both
weight and distance.
To remedy such disadvantages brought by these types of data, mathematical or simulation models can be constructed to
enable further investigation of the impacts of different variables. If we take container distribution as an illustrative example,
we can see that in order to transport a certain number of containers, different types of trucks can be used (e.g., B-double,
Super B-double, Super B-triple, etc.). It is possible to work out the number of trips that need to be taken to completely
transport these containers and then treat the CO2 emissions as a variable in mathematical models. This will facilitate the
analysis of the impact of the variable on the overall system.
Sometimes, it is not enough to only look at the mathematical model, since a lot of other factors are also changing. Again
using the container distribution example, it is important to take factors such as the volatile business environment (and
therefore not constant container flows), fast moving fuel prices (therefore no fixed cost for each trip), seasonality demand of
transport equipment (therefore a limitation to the available transport options) into consideration and assess the impacts of
these factors on the decision making process of transport network design. Mathematically modelling these types of changes
would almost be impossible since the level of complexity increases significantly if such uncertainties are to be taken into
account. Simulation models can be a substitute for mathematical models and allow the decision makers to play with a lot of
combinations of different parameters. Variables can still be embedded into the simulation models to enable further analysis
once the simulation runs are completed.
การแปล กรุณารอสักครู่..