The dynamics of traffic assignment and traffic control: A theoretical study

The paper considers traffic assignment, with traffic controls, in an increasingly dynamic way. First, a natural way of introducing the responsive policy, P_o, into steady state traffic assignment is presented. Then it is shown that natural stability results follow within a dynamical version of this static equilibrium model (still with a constant demand). We are able to obtain similar stability results when queues are explicitly allowed for, provided demand is constant. Finally we allow demand to vary with time; we consider the dynamic assignment problem with signal-settings now fixed. Here we assume that vehicles are very short and that deterministic queueing theory applies, and show that the time-dependent queueing delay at the bottleneck at the end of a link is a monotone function of the time-dependent input profile to the bottleneck. We have been unable to obtain results when dynamic demand and responsive signal control are combined.     

Analysing queueing at toll plazas using a coupled,multiple-queue,queueing system model: application to toll plaza design

A vehicle approaching a toll plaza observes the queues at each of the available toll-lanes before choosing which to join. This choice process, the arrival process of vehicles and the service characteristics of the toll-booths, affect the queues and delay the drivers. In this paper, queueing at a toll plaza is modelled as a multiple-queue queueing system where the arrival process to a queue (toll-lane) is dependent on the state of all the queues. In the past, such systems have been modelled mathematically only for two queues and are not applicable for toll plazas with three or more toll-lanes. The proposed model determines the steady-state probability density function (pdf) for the queues at large toll plazas. This study is used to determine the number of toll-lanes or the length of the upstream queueing area required to achieve certain user-specified levels-of-service. Expected delay and maximum queue length are used as level-of-service measures. Indicative design charts are also provided.     

Characterization and evolution of spatial density patterns in urban areas

This paper examines the spatial patterns of population density, household automobile ownership and other socio-demographic variables that affect urban travel, as a function of distance from the central city core. Spatial density functions provide a useful characterization of urban structure, and of its evolution when taken at different time intervals. Analysis of the data from four case cities (Austin, Atlanta, Dallas, Phoenix) for 1960, 1970 and 1980 reveals continuing overall dispersion away from the traditional central core, accompanied by the densification of formerly low-density suburbs. This presents implications for high congestion levels in the densifying suburban communities, comparable to those typically associated with the CBD. In addition, the analysis has captured the continuing growth of average household automobile ownership and revealed a distinct spatial pattern that seems to be robust across the case areas considered, as well as within radial corridors in the one case that was so analyzed (Austin).List of symbols Y gross population density in census tractt - X _t distance in miles from the center of tract to the CBD - parameter representing the central business district density - y density gradient parameter - X₀, X₁, X2, and X3 locations of the knots in a three segment division of the x-axis - D_i dummy variable defined for thei-th segment - ₁ normally distributed disturbance term,a _i,b _i,c _i,d _i,i = 1, 2, 3 — parameters to be estimated - A _t area of census tractt     

Approximate queueing models for analyzing harbor terminal operations

The purpose of this paper is to present approximate queueing models to help assess the impacts of tug services on congested harbor terminals. The models are intended for harbors in which tug shortages are rare. A congested harbor terminal is modelled as a queueing system with m identical tugs (servers) and n identical berths (customers), and with general probability distributions of tug service time and berth cargo-handling time. The distribution of the number of berths in the system, a basic element to analyze the system performance, was established for two cases. For large m, the distribution was approximated by a binomial model and the respective accuracy tested. For small m, an approximate model for this distribution was developed. Particular emphasis was given to developing explicitly the probability of having one berth in queue and establishing the remaining probabilities of the distribution approximately. The model for small m was validated by means of simulation for various cases of harbor terminal operations exhibiting different ranges of the coefficient of variation of tug service time. The models were found to be reasonably accurate within a certain range covering situations in which tug shortages are in the order of 10% of the time or less.     

Incorporating Uncertain and Incomplete Subjective Judgments into the Evaluation Procedure of Transportation Demand Management Alternatives

This paper proposes a methodology for evaluating transportation demand management (TDM) alternatives in the context of multi-criteria decision making (MCDM). The proposed approach takes into account not only quantitative criteria (i.e. transportation and environmental impacts) but also qualitative criteria (i.e. social impacts) which are intrinsically uncertain and subjective. The transportation impacts of the TDM alternatives are estimated by TRIPS¹ program, while the MOBILE5a² is employed in order to estimate environmental impacts in terms of NO_X, CO and Hydrocarbon. The social impacts of the TDM alternatives are estimated by interviewing relevant experts. Consequently, the uncertain subjective judgements were quantified by the evidential reasoning (ER) approach based on decision theory and Dempster–Shafer theory of evidence. In order to measure the weights of criteria, analytical hierarchy process (AHP) is adopted. As a last step, the CODASID³ method based on a complete concordance and discordance analysis is used to rank alternative TDM schemes. The proposed approach is demonstrated by ranking 14 TDM alternatives, which are chosen for the central business area in Bangkok, Thailand of 22-square-kilometers.     

The construction and use of continuous autoregressive models for traffic indices

This article discusses the use of continuous autoregressive models to describe the behavior of traffic indices. From discretely sampled data, second-order differential equation models are constructed to represent dynamic traffic fluctuations as the response of a linear system to a stochastic forcing function. The results are compared to the more common M/G/∞ queueing model approach, and the analysis is demonstrated on time series of aircraft concentration in thirty-one enroute air traffic control sectors.     

Stochastic dynamic itinerary interception refueling location problem with queue delay for electric taxi charging stations

A new facility location model and a solution algorithm are proposed that feature (1) itinerary-interception instead of flow-interception; (2) stochastic demand as dynamic service requests; and (3) queueing delay. These features are essential to analyze battery-powered electric shared-ride taxis operating in a connected, centralized dispatch manner. The model and solution method are based on a bi-level, simulation–optimization framework that combines an upper level multiple-server allocation model with queueing delay and a lower level dispatch simulation based on earlier work by Jung and Jayakrishnan. The solution algorithm is tested on a fleet of 600 shared-taxis in Seoul, Korea, spanning 603 km², a budget of 100 charging stations, and up to 22 candidate charging locations, against a benchmark “naïve” genetic algorithm that does not consider cyclic interactions between the taxi charging demand and the charger allocations with queue delay. Results show not only that the proposed model is capable of locating charging stations with stochastic dynamic itinerary-interception and queue delay, but that the bi-level solution method improves upon the benchmark algorithm in terms of realized queue delay, total time of operation of taxi service, and service request rejections. Furthermore, we show how much additional benefit in level of service is possible in the upper-bound scenario when the number of charging stations is unbounded.     

Behavior of the cell transmission model and effectiveness of ramp metering

The paper characterizes the behavior of the cell transmission model of a freeway, divided into N sections or cells, each with one on-ramp and one off-ramp. The state of the dynamical system is the N-dimensional vector n of vehicle densities in the N sections. A feasible stationary demand pattern induces a unique equilibrium flow in each section. However, there is an infinite set—in fact a continuum—of equilibrium states, including a unique uncongested equilibrium n^u in which free flow speed prevails in all sections, and a unique most congested equilibrium n^con. In every other equilibrium n^e one or more sections are congested, and n^u  n^e  n^con. Every equilibrium is stable and every trajectory converges to some equilibrium state.Two implications for ramp metering are explored. First, if the demand exceeds capacity and the ramps are not metered, every trajectory converges to the most congested equilibrium. Moreover, there is a ramp metering strategy that increases discharge flows and reduces total travel time compared with the no-metering strategy. Second, even when the demand is feasible but the freeway is initially congested, there is a ramp metering strategy that moves the system to the uncongested equilibrium and reduces total travel time. The two conclusions show that congestion invariably indicates wastefulness of freeway resources that ramp metering can eliminate.     

A moving bottleneck

Suppose that a vehicle or convoy enters a two-lane unidirectional roadway and travels at a velocity v^* less than the prevailing traffic. This moving bottleneck may cause a queue to form as vehicles try to pass the obstruction. It is shown that by going to a moving coordinate system traveling at velocity v^* the analysis of this can be transformed into a corresponding analysis of flow past a stationary bottleneck. The theory is then extended to investigate possible consequences of trucks on grades.     

Walking-distance introduced queueing model for pedestrian queueing system: Theoretical analysis and experimental verification

We have introduced the effect of delay in walking from the head of a queue to the service windows in the queueing model and obtain a suitable type of queueing system under various conditions by both computational simulation and theoretical analysis. When there are multiple service windows, the queueing theory indicates that mean waiting time in a fork-type queueing system (Fork), which collects pedestrians into a single queue, is smaller than that in a parallel-type queueing system (Parallel), i.e., queues for each service window. However, in our walking-distance introduced queueing model, we have examined that mean waiting time in Parallel becomes smaller when both the arrival probability of pedestrians and the effect of walking distance are large. Moreover, enhanced Forks, which shorten waiting time by reducing the effect of walking distance, are considered, and parts of our results are also verified by real queueing experiments.     

Well-to-wake energy and greenhouse gas analysis of SOX abatement options for the marine industry

This paper investigates the well-to-wake energy consumption and greenhouse gas emissions of several key SO_X abatement options in marine transportation, ranging from the manufacture of low sulfur fuels to equipping the vessel with suitable scrubber solutions. The findings suggest that a scrubber system, used with current heavy fuel oils, has the potential to reduce SO_X emissions with lower well-to-wake energy consumption and greenhouse gas emissions than switching to production of low sulfur fuels at the refinery. A sensitivity analysis covering a series of system parameters shows that variations in the well-to-tank greenhouse gas emissions intensity and the energy efficiency of the main engine have the highest impacts in terms of well-to-wake emissions.     

Generalised time and the problem of equity in transport studies

The generalised time expenditure on travel,g _t, may be defined asg _t =t +m/λ where t is the amount of time spent, m is the amount of money spent, and λ is the value of time. Generalised time is expressed in units of time, unlikeg _c, the generalised cost (g _c =m + λt), which measures the same quantities in units of money. Data on all trips from two studies are analysed to show that, as household income increases, the money spent per mile travelled increases, but the time spent per mile decreases. The use of generalised time gives a different picture of the relationship between income and the total time and money spent on travel to that given by the use of generalised cost. The choice between using generalised time and generalised cost in evaluation is fundamentally a choice between assuming that the marginal utility of time and that of money is constant. The procedures at present recommended by the Department of the Environment in U.K. have elements of both assumptions, with some loss of consistency. There are some a priori reasons for expecting constant marginal utility of time to be a more plausible assumption than constant marginal utility of money. Time is, by its nature, equitably and unalterably distributed, not subject to the accidents of inheritance, theft, chance, inflation, social system or government decree. Everybody starts off with 24 hours a day. Although the amount of “free” time varies, of course, it probably varies within a much smaller range than the amount of wealth, certainly for the employed population. In allocating time between various activities, the use of words likespend, save, waste, lose, gain, and so on is a reflection of how deeply rooted in language and thought is the concept of time as a fundamental currency. This approach is strengthened by recent developments in two areas where generalised cost has been found to be a useful tool of analysis — (a) in explaining and predicting the behaviour of travellers, and (b) in evaluating the social costs and benefits arising from transport projects. Only non-working time will be considered here. It is suggested that in some circumstances generalised time would allowbehaviourally correct relationships between non-working time and money to be used in evaluation. This paper — which is a revised version of “A Hypothesis of Constant Time Outlay on Travel” (Paper F29: Planning and Transport Research and Computation Ltd. Sussex, June 1973) — is based upon work carried out with the financial support of the Social Science Research Council.     

MANTO — a research project of telecommunication applications for the future information society

The impact of new technologies has been mostly under-estimated in the past and, as a result, the transitions from one era to another have occurred more or less by chance. The MANTO research project is intended to avoid mischances. So it indicates potential developments in the telecommunications sector and, in particular, investigates their effects upon transport and settlement and also upon society, the economy and the environment. Nevertheless, MANTO does not merely take a look at the future, like the Greek oracle from whom it takes its name; by means of concrete recommendations, MANTO is intended to provide persons in responsible positions with a tool with which they can consciously organise that future. Risky developments are to be precluded in advance while desirable consequences are to be promoted. Explanation of the used term: MANTO was the daughter of the famous prophet Teiresios of Theben and a seer herself too. As a (German) abbreviation MANTO also stands for: human being (Mensch), offer (Angebot) and needs (Nachfrage) for transportation and telecommunication (Transport und Telekommunikation), considering economy and ecology (Oekonomie und Oekologie).     

Delay processes at unsignalised junctions: The interrelation between geometric and queueing delay

The interrelation between geometric delay—the delay caused by the need for isolated vehicles to slow down to negotiate a junction—and the queueing delay due to vehicle-vehicle interactions is considered. The delay elements present in measurements of geometric delay are identified, and some overlap is seen with those normally included in the service mechanisms in queueing models. The total delay per vehicle is not, therefore, the sum of the measured geometric delay and the queueing delay. Although the geometric delay can be redefined so as to eliminate the overlap, it is not then measurable. A framework is developed in which the relationships between the “pure” and measured geometric delay and the queueing delay are expressed for the population mean values. Approximate expressions are developed for the elements of delay. The framework is extended to allow queueing delays to be represented by time-dependent functions allowing approximately for non-randomness in arrival and departure patterns and service time differences between queueing and non-queueing vehicles. Numerically, geometric delay elements constitute a significant proportion of total delay except at traffic intensities approaching unity. In this region time-dependent effects dominate the queueing process.     

The viability of pure vegetable oil as an alternative fuel for large ships

Heavy fuels are likely to remain the dominant fuel source for two-stroke, low-speed diesel engines for large ship propulsion for the next decade or more. There is however, potential for increased use of pure vegetable oils (PVO) as an alternative and, by emitting lower levels of several pollutants, this can help the attainment of Annex VI of the MARPOL 73/78 convention aimed at large ships using fuels with less than 4.5% sulphur or 1.5% sulphur in SO_X emission control areas The use of alternative fuels can also influence the attainment of the Kyoto protocol that requires greenhouse gas emissions to be reduced by 5% by 2010 compared to 1990. This paper analyses the physical and chemical properties of various pure vegetable oils as an alternative to heavy fuel oil for large ship propulsion.     

The convergence of diagonalization algorithms for asymmetric network equilibrium problems

We provide a sufficient condition for the convergence of diagonalization algorithms for equilibrium traffic assignment problems with asymmetric Jacobian matrix B(v) of the link user cost mapping s(v) of the flow v. When

Full-size image

, where D(v^*) > 0 is the diagonal of B(v^*) and v^* is the equilibrium flow, we demonstrate a local convergence theorem for nonlinear cost functions. The implication of this result for practical applications of the model are outlined.     

Comparing air pollution from ferry and landside commuting

Recent and anticipated growth in passenger ferry service has been complicated by concerns about air pollution from marine engines that are only starting to be regulated. While marine engines are known to be a significant and growing source category in some locations, sparse data and analytical difficulties have prevented rigorous comparisons of marine and on-land passenger travel. Using data gathered in the San Francisco Bay Area, we model emissions from three passenger ferries and the matching on-land travel that would be used by commuters if ferry service were not available. The results are analyzed parametrically for levels of ridership and induced travel demand, and for new technologies, including selective catalytic reduction and natural gas fuel. Results indicate that under some conditions, passenger ferries reduce some emissions (including particulate matter emissions) relative to the matching on-land service but increase others. Emissions of NO_X are particularly problematic––all the technologies examined lead to increased NO_X emissions due to ferry commuting. Some of the emissions comparisons are sensitive to mode split, ridership, or induced travel demand. However, NO_X emissions are not––ferry commuting always raises NO_X emissions, even with the most advanced technologies. Implications for local air quality regulators and for technology development are discussed.     

Estimation of nitrogen oxides emissions from petrol and diesel passenger cars by means of on-board monitoring: Effect of vehicle speed,vehicle technology,engine type on emission rates

NO_X emission rates of 13 petrol and 3 diesel passenger cars as a function of average speed from 10 to 120 km/h, emission class (pre-Euro 1 – Euro 5), engine type were investigated by on-board monitoring on roads and highways of St. Petersburg using a portative Testo XXL 300 gas analyzer. The highest level of NO_X emission 0.5–2.5 g/km was inherent to old pre-Euro 1 petrol cars without a catalytic converter. NO_X emissions rates of Euro 1 and Euro 2 petrol cars changed within 0.15–0.9 g/km, Euro 3 – 0.015–0.27 g/km, Euro 4 – 0.013–0.1 g/km, Euro 5 – 0.002–0.043 g/km. Euro 3 – Euro 4 petrol cars generally satisfied corresponding NO_X Emission Standards (ES), except cold-start period, Euro 5 petrol cars did not exceed ES. Warmed, stabilized engines of Euro 3 – Euro 5 petrol cars showed 5–10 times lower NO_X emission rates than corresponding ES in the range of speed from 20 to 90 km/h. NO_X emission rates of diesel Euro 3 and Euro 4 cars varied from 0.45 to 1.1 g/km and from 0.31 to 1.1 g/km, respectively. Two examined diesel Euro 3 and one Euro 4 passenger vehicles did not satisfy NO_X ES at real use. Euro 3 diesel cars showed 28.9 times higher NO_X emissions than Euro 3 petrol cars and Euro 4 diesel car demonstrated 17.6 times higher NO_X emissions than Euro 4 petrol cars at warmed and stabilized engine at a cruise speed ranging from 30 to 60 km/h.     

The design of single allocation incomplete hub networks

The hub location problem deals with finding the location of hub facilities and allocating the demand nodes to these hub facilities so as to effectively route the demand between any origin–destination pair. In the extensive literature on this challenging network design problem, it has widely been assumed that the subgraph induced by the hub nodes is complete. Relaxation of this basic assumption constitutes the starting point of the present work. In this study, we provide a uniform modeling treatment to all the single allocation variants of the existing hub location problems, under the incomplete hub network design. No network structure other than connectivity is imposed on the induced hub network. Within this context, the single allocation incomplete p-hub median, the incomplete hub location with fixed costs, the incomplete hub covering, and the incomplete p-hub center network design problems are defined, and efficient mathematical formulations for these problems with O(n³) variables are introduced. Computational analyses with these formulations are presented on the various instances of the CAB data set and on the Turkish network.     

Further developments in the pricing of local public transport

Abstract

This short paper supplements and updates the writer's Recent Developments in the Pricing of Local Public Transport Services, Transport Reviews, 1 (2), pp. 127–150 (1981). That paper described the situation as at 1980 in Britain, with some examples from other countries. This paper updates the position with regard to four main issues:

(a) Impact of the Transport Act 1980 in Britain, which has virtually removed fares control from stage carriage services, and removed all price and quantity control from express services by road.

(b) Further development of zonal pricing, travelcards and off‐bus fare collection on urban systems, and the effects of price reductions (and subsequent increases) in London, West Midlands, and Merseyside.

(c) Problems concerning the legality of financial support for fares policies in British cities.

(d) Following a presentation of work arising from the earlier paper at the meeting of the Transportation Research Board in Washington, D.C. in January 1983, the contrast between British and American policy is reviewed, with the writer's personal view on current policy there.

In addition, evidence on certain other points in the text is updated.