Transport demand forecast is of paramount significance to transport planners and decision makers as they depend heavily on these forecasting to predict future demand for transport.
There are two main types of forecasting techniques: Qualitative and Quantitative with the Quantitative techniques used in this assignment. Quantitative forecasting technique involves the use of numerical data and the application of mathematical procedure or set of procedures to predict or forecast the quantity of a variable into the future.
Under the Quantitative technique, two forecasts are used in this assignment to forecast tram boarding (millions) for the year 2010-11, 2011-2012, 2012-2013 – the Time series forecasting and the Regression Forecasting.
Charts and visual aids are used in observing and analysing trends and relationships in all the variables in the model coupled with equations to help choose the best methods in forecasting trams demand for the next 3 years.
The results produced have shown that all charts have a common characteristic. There is indeed a linear relationship between the X axis (horizontal) and Y axis (vertical). X2b (Average yearly cost/ litre lagged 6 months) and X3 (Total employed persons (full and part time) MSD (000s) 4) are the two independent variables chosen because they have a higher Adjusted R2 and a lower SEE as compared to the rest of the remaining independent variables.
Whilst comparing the Linear/Polynomial time series forecast, the Multiple Linear regression model and the Power function mode, it can be concluded that Multiple Linear regression provides the best prediction with trams boarding forecast results showing as 172.5024 (2010-11), 176.6695 (2011-12), and 180.3989 (2012-13) respectively.
Besides the independent variables listed in this assignment, there are many other independent variables that need to be considered when it comes to estimating Melbourne’s tram demand, as measured by millions of boarding per year. Some of the other variables include service quality and reliability and parking price.
The objective of this assignment is to forecast transport demand estimation. The forecasting can be achieved by using quantitative forecasting technique models such as Time-Series forecasting and Regression forecasting, (otherwise known as explanatory or causal) and then comparing these two models to derive at the best forecast estimation of transport demand possible.
The subject of study is based on the number of boarding (in millions) of passengers on Melbourne trams from year 1983-84 to 2009-2010 being the Total Tram Boarding as the dependent variable influenced by other independent variables. These independent variables include price of zone 1 full fare ticket, petrol price, employment population, weekly earnings (persons) in Victoria, housing disposable income and the resident population. These independent variables are denoted as X1 to X6 whilst the dependent variable is denoted as Y. Excel spread sheet is used to facilitate in forecasting the transport demand estimation.
The relationship between the dependent and independent variables were exhibited through charts by inputting data into Excel.
Based on the observation and analysis of each graph, it can be said that a common characteristic can be established – there is indeed a linear relationship between the dependent (vertical axis) and independent variables (horizontal axis). For instance, we can see that there is a direct relationship between the total trams boarding and the fiscal year as these two variables were constantly increasing in the same direction for most years with a significant high jump in year 2007-08 to 2008-09 (see Appendix, Figure 1). This would suggest that more residents were taking public transport over the years.
Another pattern of the linear correlation illustrated is the