Review of O-D Estimation Methods

✅ Paper Type: Free Essay	✅ Subject: Transportation
✅ Wordcount: 14052 words	✅ Published: 18th May 2020

Reference this

Share this: Facebook Twitter Reddit LinkedIn WhatsApp

Part 1: review of O-D Estimation Methods

Introduction:

Transport planners must provide comprehensive plans to mitigate the adverse effects of rapid urbanization and traffic congestion. Planning and managing transportation systems require precise and reliable estimations of O-D models are fundamental in planning and managing transportation systems.. In order to estimate the traffic demand within an origin-destination node it requires a plethora of contextual predictions including future economic and population growth, the effects of land use and travel needs, potential transport alternatives and detailed changes to infrastructure.

In this Literature review paper, an overview of four O-D Matrix estimations methods including; Fuzzy logic model, Kalmans Filtering, Neural network and Bayesian Inference will be evaluated:

Fuzzy Logic Model

Derived from the Fuzzy Set Theory by Lotfi Zadeh, the Fuzzy Logic Model is a method of providing a degree of truthiness to imprecise and uncertain data inputs. In comparison the backbone of the Fuzzy logic Model is similar to the way the human brain operates in performing results, this is achieved by further aggregating truths from a prior group of truths which exceed a certain threshold.

(https://searchenterpriseai.techtarget.com/definition/fuzzy-logic)

The Fuzzy logic model requires 3 processes. Beginning by ‘fuzzifying’ all input values into fuzzy membership functions. These fuzzy sets are a set of elements without a crisp (clear defining boundary) with a degree of membership which can be defined as either a triangle or trapezoid shaped curves.

These curves are called membership functions which allow every input space to be represented with a degree of membership, from a range of 0 to 1. In sequence, defuzzification occurs which is the process that maps a fuzzy set to a crisp set. These allows an interpretation of the degree of membership of the elements into a real value or specific decision.

Advantages:

- Fuzzy logic is conceptually easy to understand, and its functioning can be related to the human brain.

- Fuzzy logic is great in interpreting results when input data are assumed to be incomplete and imprecise. In example, for O-D Matrix estimations when input entries such as link counts, number of departures from origins and number of arrivals at destinations are scarce Fuzzy logic is capable of making an estimation.

Disadvantages:

- All factors are taken with the same value of importance. In the context of ODM estimation, parameters equally affect a travels route choice.

- New variables and systems cannot be directly inputted into Fuzzy logic models, a new Fuzzy logic model is required to incorporate and execute these data.

Case Study: A Fuzzy Logic set approach was utilised to estimate OD matrices in congested Brazilian Traffic networks.

Kalman Filtering

The Kalman filter is an optimal estimator algorithm capable of inferring parameters of interest (traffic flow, passenger flow, etc) from imprecise, indirect and uncertain observations that contain statistical noise and other inaccuracies. As a dynamic approach it useful for short-term strategies such as route guidance and traffic management.

Kalman filtering is a two-step process involving the initial prediction step then followed with an updated filtering stage. During the initial stages, estimates of the current state variables are produced along with their uncertainties. To compare between the outputs of a measurements, a weighted average calculated from covariance allows a determination of uncertainty of the estimates present within the system’s state. This process is constantly repeated and iterated to achieve a better estimate that lies between the predicted and measured state. The iteration is constantly using the previous best estimated state and current measurement, which models the Kalman filtering method as a recursive estimator.

Advantages:

- Kalman filtering is capable of inferring data from uncertain and imprecise data sources. It can filter out unrequired data measurements.

- Kalman filtering is recursive, new measurements are capable of being processed immediately.

Disadvantages:

- It assumes both the observations and system models are both realistic, however in real world situations it is not realistic.

- It assumes noise within the system is Gaussian distributed.

Case Study:

Kalman filtering based dynamic OD matrix estimation and prediction for traffic systems.

Neural network

A neural network is an adaptive connectionists model which estimates optimal outputs through a training process. The objective of the training process is to minimize the deviations between the model outputs and the target values (Yaldi, Taylor and Yue (2011)).

During the first input layer, neurons take data into the network which flows along the connections to be scaled by connection weights which are initially estimated. An internal transfer function usually a logistic function is utilised to compute the outputs from the previous scaled inputs (Yaldi et al.,2011). The output layer represents applying an activation function to transform the estimated output, which is the total addition of the values of all the nodes in the hidden layer.

Advantages:

- Neural networks have the ability to learn themselves and produce the output that is not limited to the input provided to them.

- Since the inputs in neural network are stored within a network not a database, an output is still produced regardless if a neuron is missing or faulty.

- The neural network requires less statistical training and can identify complex relationships between variables

- Since neural network learn through examples during the training process, they can be related to real life events.

- They can perform tasks in parallel without affecting the system performance.

Disadvantages:

- A shortcoming of the neural network is its ‘black box’ nature. This issue is that an approximation derived from the neural network will not provide any insight into the form of the function as if no relationships exist between the network weights and the property being modelled. This sometimes leaves the network unexplained.

- This process is computationally heavy requiring appropriate hardware and is also time consuming.

- The model is susceptible to overfitting. This is a phenomenon in modelling when the approximated function is too closely fitted to a limited set of data points. This causes poor predictive abilities and an inability to incorporate additional data.

Case Study:

A Neural Network model was used to estimate the Origin-Destination (OD) matrix from the link volumes of transportation network.

Bayesian Inferences

Bayesian inferences is a statistical inference method based on Bayes’ rule, which constantly provides updates on the probability on a hypothesis through model structure and prior data. It considers the prior distribution and posterior distribution, which allows probability distribution to be modelled and updated each time with a new observation. Prior distributions represent our beliefs and the distribution of parameters before any data is observed whereas posterior distributions are our beliefs and the distribution of parameters after observing data. In the context of O-D matrices, Bayesian inferences can combine current and past information on OD matrix.

Advantages:

- Progressively provides more precise predictions with informative data

- Any changes to data can be easily implemented, which complies with O-D matrix estimations as traffic flow is random.

- Results are conditional and exact hence there isn’t really difference between a large and small sample size. This is extremely cost efficient as it is expensive to obtain data for a large number of traffic flows on specific set of network links.

Disadvantages:

- If the prior information is incorrect, this can lead inference results as unreliable and misleading

- There is a high computational costs for a large set of parameter.

Part 2: a) Review of Light Rail Transit (LRT)

Introduction

Metropolitan regions experiencing rapid urbanization need to emphasize on transit-oriented development (TOD) to prosper social, economic and environmental benefits. Environmental problems continue to be exacerbated consequent of increasing population densities and transportation by-products; focus has been placed to implement cleaner forms of transport infrastructure such as the Light Rail Transit (LRT) system. The LRT system is a popular strategy in integrating public transport networks to ensure better management of traffic congestion, improving accessibility and mobility within high density regions and fostering economic growth. In this literature review, its positive and negative externalities are analysed with a cost-benefit analysis to ensure these projects are economically viable in certain regions.

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!

Essay Writing Service

Impacts of Accessibility, Land use, Gentrification and Property Values

The rationale in developing an LRT system is to foster economic developments within the vicinity by greater land use and accessibility, however the indirect effects of gentrification need to be mitigated. Dziauddin, Alvanides and Power (2013) utilize the concept of accessibility to understand the correlation between transportation and land use. In denser metropolitan regions, an LRT system would significantly promote the spatial interactions between activities and land uses. This would spur economic activity, as evident in Fan, Guthrie and Levinson (2012) reports on the effects of the introduction of the Hiawatha light-rail line on lower wage workers. According to the National Household Transportation Survey, these lower wage household had relatively less car ownership rates when compared to other households. Furthermore, the regression analysis deduced positive benefits for improving accessibility by 30-40% for transit dependent workers and the significant impact of a fully integrated transit network. (Fan et al (2012); Johnson, Ercolani and Mackie (2017)) Improvements in accessibility allowed areas with shorter public transport times to be associated with higher employment levels and economies. Comparatively, a case study undertaken in the Klang Valley region in Malaysia had undergone major economic growth and rapid urbanizations had led to a steady increase of vehicles ownerships and traffic congestion levels as a result (Dziauddin et al (2013)). The implementation of an LRT improved accessibility due to congestion relief and better regional mobility, reduced roadway costs, improved traffic safety and air qualities which will be a focus later in this review. In contrast to a greater interconnectivity within the city region and promotion of land use, several studies have confirmed significant or marginal impacts on gentrification and property values within the vicinity of LRT infrastructure. Gentrification is a controversial topic in urban planning, it is process of changing the character of a neighbourhood when an influx of higher income residents displaces and moves into a low income or working-class community (Kennedy and Leonard (2001)). Increased property values in proximity of LRT stations are indicative of transit-induced gentrification, a by-product of TOD; Cavers and Patterson (2015) and Immergluck (2009) analysed that transit infrastructure planning illustrated positive effects of gentrification and property value from urban transits. However, LRT systems tend to be located within regions amenable for development hence it is ambiguous to assume that the LRT is the sole reason for changes in land usage and values. Parsons, Brinckerhoff, Quade and Douglas (1998) elucidated that he level of accessibility and land use development are not solely dependent and are effected by many factors, including the current level of accessibility, state of regional economy, the availability of services such as sewer, drainage and water and desirability for land development. In addition, this is manifested in the introduction of the LRT in Manila, there were only marginal impacts on property and land values (Pacheco-Raguz (2010)). This is a result of existing transport infrastructure with good accessibility prior to LRT and the lack of complementary TOD policies. Furthermore, to mitigate the dilemma of gentrification, transport planners must consider the rising costs of living for existing residents within the community by providing affordable housing options and financial aid. This was manifested in the Metro TOD Program in Portland, investments with the aid from local governments were made to provide affordable housing development for moderate- and lower-income households living near the LRT regions. Urban planners must consider effects on surrounding areas when implementing an LRT system, however by improving the level off accessibility with densely populated regions economic development is fostered.

Cost-benefit analysis, Project Implementation and the state of economy

The implementation of large-scale infrastructure projects requires extensive planning and a cost benefit analysis to determine if its economically feasible. TOD’s including the introduction of an LRT system would provide a framework for future populations and employment growth and alleviate problems that arise from urbanization; environmental sustainability, traffic congestion and overcrowding. As an infrastructure innovation, the implementation of the LRT systems is susceptible in incurring costs overbudget due to unpredicted and delays mainly affecting their capital expenditure (CAPEX). This was the case for the current construction of Sydney’s CBD and South East Light Rail network, which was initially budgeted at $1.6 billions and a final estimation of the project exceeded $3 billion due to mispricing and omissions in the business case. Initially, a benefit cost ratio (BCR) of 0.8 was determined, indicating that the costs of construction and operations exceeded the revenue required for this project. However, the BCR was recalculated and an economic appraisal was made where the LRT in Sydney was forecasted to provide $4 billion in total economic benefits. This calculated a BCR of 2.4 during the initial budget of $1.6 billion. Prolonged delays, misconduct and legal processes reduced the BCR to 1.4, however it the project continued as benefits were still provide as the BCR were greater than 1.0. Moreover, to avoid major delays within projects these considerations must be achieved for a large-scale project to be implemented, this is illustrated in the table below. The LRT project development involves four major steps (IUT, (2013));

Major steps in LRT Project Implementation	Course of Action
Transaction Structuring	It involves the selection of the Public-Private Partnerships (PPP) partner to perform deliveries and services related to the implementation of the project.
Final location survey and layout drawings	Solving with nearby city agencies and the planning of the layout and location of the project
Choice of mode/technology	The technology used are considered, this is also dependent on the context of the cities needs and state of economy.
Design of Civil Engineering structures	This step involves the design and modelling of technical aspects Including the station, routes, etc.

Major steps in the Implementation of LRT

During the initial stages of project development, feasibility reports in the ‘Transaction structuring’ step are undertaken including a cost-benefit analysis to ensure if the project is necessary, beneficial and financially adequate. To finance for large scale infrastructure plans such as the LRT systems, Public Private Partnerships (PPP) are vital in development, as the government pays the private sector to perform and deliver the services required for the infrastructure. These payments are usually sourced from the public sector in the form of tax revenues.

The overall budget for LRT’s system is mainly contributed to its CAPEX, however its operating expenditure (OPEX) when compared to other modes of transport is relatively low as it is energy efficient and capable of carrying many passengers for a given trip. Reiterating, that the cost performance of LRT systems are susceptible in incurring further costs in CAPEX, discredits the innovation of large infrastructure project with regards to the justification and adoption of LRT (Love, Ahiage-Dagbui, Welde and ODeck (2017)). Illustrated in Table, mean overbudget of 42% were experienced by LRT projects in US cities from 1988 to 2013. Furthermore, they also mentioned on future proofing which would improve the cost effectiveness of the LRT by mitigating strategic misrepresentation and optimism bias. Many governments and companies are working to increase the capital efficiency of infrastructure procurements, by maximizing the productivity of assets and minimizing the costs associated with infrastructures (EIU (2014)). However, the CAPEX and OPEX costs are high to maintain, the LRT system provides many benefits that usually exceeds the costs in the long run. It is capable of stimulating investment by revitalizing commercial and residential development, grater productivity due to better mobility within the CBD and many other social and environmental benefits. However, the economic viability of the cost-benefit analysis also dependent on the locational contexts and the opportunity cost of spending capital on other sources or different modes of transport including the BRT. The cost per kilometre(km) for LRT varies significantly depending on the location of cities and countries, evident in the table below.

Location	Cost of LRT per kilometre(km)
Canberra	$58-$70m
Gold Coast	$18-$22m
Sydney	$15m
Manchester (UK)	$24m

LRT comparisons with bus rapid transit (BRT)

An alternative to an LRT system in tackling traffic congestion and improving transport capacity is the bus rapid transit (BRT).

Comparison

BRT and LRT

Operating Costs

LRT systems are more cost effective and have lower operating and

maintenance costs compared to the BRT as they are more durable.

Operating costs for LRT systems per user are approximately 16%

less than BRT’s due to its ability to transport a large capacity of

passengers per trip. Hence, LRTs require less subsidies in the long time

due to its higher influx of passengers.

Environmental

Sustainability

Reduction in carbon emissions and improvements in air quality

especially in the regions around the CBD can be significantly reduced e if the source of LRTs electricity is derived from renewable resources.

In the case of BRTs, most buses are diesel-fuelled which contributes

to carbon footprint and greenhouse gases.

According to Transport for NSW, LRT systems use 10 times less

energy than BRTs for the same travel distance and the system is

expected to reduce carbon footprint by 663,000 tons over the next

3 years.

However, a rollout of electric buses which utilize renewable resources

would have similar impacts to the environment as the LRT. But a

downfall is that buses contribute and exacerbate congestion on roads

which indirectly affects emission from other surrounding vehicles.

CAPEX

BRT systems are typically 4 to 20 times less than an equivalent LRT

system. The capital costs per mile of BRT systems are substantially

lower than LRT systems as in the figure!!. Since the construction

of its infrastructure is much simpler and easier to integrate with the

transport network.

Public disruptions

(noise, vibrations)

Both transit systems are culprits of noise pollution and vibrations,

policies have been implemented to minimize vibrational annoyance to

a certain limit. These disturbances are present during the vehicle

functioning and especially during the construction phases.

Sourced from the U.S Federal Transit Administration Figure !!,

depicts LRTs are more disruptive in terms of vibration when

compared to BRTs at any corresponding speed.

Accessibility and

Flexibility

BRTs can provide service over a larger geographical area. BRT routes

and station stops are much more flexible and adequate compared to

LRT which is designated on a fixed route. Also, it would be costly to

further trackwork and stations at the construction of the LRT.

Transportation

LRTs usually operate at higher speeds and offer greater passenger

capacity compared to BRTs. Buses are more likely to be subjected to

traffic congestion and are required to be constantly changing speeds at

stops and intersections than LRTs. In Australia, LRTs on average

travel at 35km/h compared to BRTs at 30km/h. Furthermore, LRTs

have higher ridership and passenger capacity per vehicle than BRTs.

The main advantage of BRT systems is its flexibility and accessibility which makes it preferable for smaller-medium sized cities. In contrast, LRT systems yield greater capacity of passengers and alleviate traffic congestion allowing it to be more suitable for larger and denser cities.

Reliability, user satisfaction, service quality, emissions, health and noise.

Reliability is a measure of punctuality and the maintenance of proper function, these issues are prevalent and can be categorised as operational, planning and technical. These instances arise from schedule development, malfunctions, bad weather and accidents. Furthermore, factors of unreliability usually occur within the CBD due to pedestrian movements, several signalized intersections and higher dwell times as a result of higher demands (Sirisoma, Wirasinghe (2008)). The LRT is usually perceived as a reliable mode of transport due to its frequency, accessibility and less susceptibility to delays due to its designated and fixed routes. Moreover, LRTs are attractive hallmarks of a strong economy that can foster economic development and providing benefits for citizens including better ridership experiences and reduced carbon footprint. In addition, improvements or modification on existing infrastructure are relatively low costs for LRTs as operating expenses are usually within the budget of the public sector. When properly implemented, with the correct placement of stations and routes, better interconnectivity within city regions would substantially provide region development of rising property values and de-densify road networks to encourage transit use. Another incentive to implement LRT systems is to strategize environmental sustainability and urban renewal, a core desire of developed cities. There is a direct correlation of the impacts of public transport and transit-oriented development to the social, economic and environmental aspects of cities (Cervero 2005). Ultimately, as a locomotive based on electricity would improve air quality within the epicentre of cities and influence transit use by removing cars of the road. Several studies have illustrated the burden on health of car pollutants, in addition the reduced travel time promotes physical activity and lessens the effectiveness of seating. These results are emphasized if the electricity sourced for the LRT is entirely sourced from renewable resources such as wind, solar and hydroelectricity. Evident within a study, Macdonald, Stokes. Choen. Kofner and Ridgeway (2010) investigated the effects of LRT facilitating physical activity and changes in BMI. Consequently, LRT users commuting to work were 81% less likely to become obese over time, the results depicted the publics use of LRT could lead to health improvements for millions of individuals. Several studies have illustrated positive relationships between LRT systems and reducing regional carbon and greenhouse gas emissions. (Andrade and D’Agosto (2015)) forecasted forgone emission subsequent of the Rio de Janeiro Metro line 4 from 2016-2040 as shown in figure !!!. Furthermore, calculations of avoided emissions were calculated in major cities as results of Metro or Light Rail line implementation.

Acceptability, Noise and Safety

The implementation of LRT systems causes social impacts that may conflict with individuals. Safety is a major concern; certain LRT systems are incorporated into existing road with traffic where pedestrians are exposed to unprotected rails. (Currie and Reynolds (2010)) demonstrated methods in improving pedestrian safety within the surroundings of light rails which include; improving pedestrian awareness of crossings and approaching trains, informative educational programs about hazards and greater sight distance of approaching vehicles. Furthermore, analysis displayed that after new stops were implemented with a safety audit approach, light rail and pedestrian incidents declined by 10% per annum. In addition, the negative externalities that arise from the LRT system include the noise that is generated during construction and operation and its general acceptability especially by non-users. Vogiatzis and Vanhonacker (2016) investigated track-based solutions and the source of noise to mitigate the railway rolling noise, results related to applying noise absorbent barriers and installing rail dampers. During the development of the LRT, many indirect implications that impact surrounding residents and commuters due to road and pathway blockages, noise and vibration pollution and irritating sight. Reiterating, LRT systems are susceptible to prolonged delays and disruptions which exacerbate the negatives of construction, traffic congestion and increased density of pedestrians are more prevalent due to reduced pathways and detours. However, people are usually acceptable of the implementation of LRTs as they would provide further benefits in the long run.

(Ercolani, 2017)

b) Bibliometric analysis

Bibliometric analysis is a statistical method in quantitively analysing the academic quality of publications.

A bibliometric analysis was performed on the Light Rail Transit to highlight its evolution of research, indicate which countries are taking the most interest in this field and important topic within this research area. In recent years, there has been an 2-12% increase in article publication from the past decade. This signifies a growing interest and research on light rail transit systems, consequent of implementing cleaner and more effective transportation methods to mitigate the negative effects of globalisation. USA has shown greatest interest in this field at 54% of all publications, as the strongest economy in the world USA has steady resources to supply for large infrastructure project such as the LRT system. The Top 5 countries in record count contribute to approximately 75% of all publications; All of these countries are either developed or have a high population density such as the Peoples Republic of China, urbanization is a common factor and incentives to integrate an LRT system would alleviate overcrowding and reduce vehicle usage.. Furthermore, LRT system are a hallmark of a developed economy state that can contribute in fostering economic development, aiding traffic congestion and improving environmental sustainability. Moreover, Environmental sustainability has become a focal point especially in developing countries to reduce carbon footprint and greenhouse gases. In addition, these developed countries want to provide a future framework for their cities by improving accessibility and travel time. In Figure, the bibliometric analysis represents important topics in the LRT research which consist of benefits, environment, accessibility and development.

Mackie, E. a., 2017. Econometric analysis of the link between public transport accessibility and employment.

References