The paper develops an equilibration algorithm that can be applied to several traffic assignment models. The algorithm aims to equalize the journey times of the used paths and to minimize them, by transferring flow from the loaded path with the longest travel time to the shortest path. Then it is necessary to store the paths.The procedure of equalization by transfer (PET) was introduced by Gibert to solve the Wardrop-Beckmann model; we extend it to address elastic demand, or the linear logit model, or a dual criteria model with continuously distributed values-of-times. The extension principle lies on linking to each path a generalized journey impedance which stands as a counterpart of the journey time considered in the Wardrop-Beckmann model.A PC-based computational experiment is carried out to demonstrate the applicability of the PET for medium-sized networks (eg. a 140 zone, 2,000 link network); a comparison with the traditional Frank-Wolfe procedure is included, showing that a dual criteria model may be solved at a cost comparable to that of the Wardrop-Beckmann model.
Since its first achievement in France in 1981, railway high speed has progressively spread in Europe. This paper shows this evolution with some high speed maps for France, Germany, Great Britain, Italy, Spain, Sweden...It underlines the fact that this development has been permitted by the construction of 2 500 kilometers of new railway lines. But the high speed trains are also using 14 000 kilometers of standard lines. It is on these standard lines that the pendular technology offers interesting prospects. The pendular trains allow an increase in the commercial speed while few investments are necessary to improve infrastructure. So, this new technology constitutes a fundamental component for a new deal in railway high speed.
Is the railway passengers traffic, and in particular the high speed train traffic, going to gain ground in France by 2015 ? Facing the bad results observed at the beginning of the 90’, one could be pessimistic and consider that it will be impossible to reach again the growth rate of the 1975-1992 period. Nevertheless, the M.A.T.I.S.S.E. model can help us to find out on what conditions the pursuit of the former trends is possible. Two key variables should be taken into account : the first one is the economic growth rate and its effects on travel demand ; and the second one, the development of new high speed train supply (East, Méditerranée, Rhin-Rhône, Lyon-Turin, Languedoc Roussillon...). The achievement of all of these projects would allow an increase from 2.1 to 2.6% of railway passengers traffic, that is to say one point more than between 1975 and 1992. But this result would be mainly due to the international traffic since the most important national lines have already been carried out.
Given the French official projects of high speed train, several new lines should have been built during the 90’. That has not been the case because many factors of sticking have arisen. It is now obvious that the construction of important length of new lines leads to a dead end. The development of railway high speed has to seek another solution. One can be found thanks to the pendular technology which offers interesting prospects. Joined with new methods of control and signal (European Train Control System and European Rail Traffic Management System), the pendular technology allows an optimization of the whole railway network: both of the new and standard lines. The most important advantage of this new technology lies in the feasible staggering of investments in new infrastructures. So, with the same amount of capital, and in a perspective of fitting out the territory, a number of projects can be developed simultaneously.
The disappointing commercial results of the northern high speed train, associated with a global crisis of railway transport in France at the beginning of the 90', have led to call into question all the projects of new high speed train lines. All of them have been revalued in the Rouvillois report which concluded to the impossibility of building new high speed lines without public finances support. Given the fact that the most profitable lines are already built, private funding alone is not sufficient. But this obviousness must not lead to abandon all the projects. Some of them have indeed a good financial profitability and even a very good social return. But in order not to penalize the whole high speed train system, priority must not be given to the less profitable lines. And it would be better to avoid a zero sum game between the owner of infrastructure and the running company.
The purpose of this paper is to exhibit the main changes in mobility revealed by the comparison of the results of the two waves of survey. It begins with a brief presentation of the main characteristics of the survey. Then, the results are split according to 3 geographical areas a significantly different distance from Paris. Within each of them, the three main purposes of travel (business, commuting, and private) are successively considered. Last, the conclusion relates the observations carried out on the Atlantic-TGV to the knowledge already gathered concerning long-distance mobility and high-speed rail services. It basically confirms the characteristics of " high speed mobility " as well as the fact that many elements outside the sphere of transport have a definite influence on mobility dynamics.
This paper leans on the J.M. Fourniau analysis regarding to the relationship between the decision making, the involvement and the sharing of power. It questions the J.M. Fourniau model from theoretical, heuristic and empirical point of views. Firstly, it shows that the thesis of J.M. Fourniau refers both to Pluralist and Corporatist theories of the state. Such a position reveals the interest of a global approach as the " policy networks " approach suggest it. Secondly, regarding to its too high level of aggregation, the heuristic dimension of the J.M. Fourniau model is critically examined. Thirdly, an empirical analysis of the oppositions against the TGV in the Nord-Pas-de-Calais Region illustrates the great variety of situations. A typology based on the origin of the claims shows the great fragmentation of local system. But, one can show that all the oppositions refer to the " normal deviance " pattern.
The spatial relationship between motor vehicle accidents and residence location of the driver is formalized through the example of Brussels (Belgium). First, we describe the distance function between the two locations as well as its differentiation between crash severity levels. Second, we attempt to develop a simple probabilistic explanatory model in order to formalize and test the relationship. The results empirically show that most accidents occur in the close vicinity of the residence of the road user and that severe accidents tend to occur further away from the residence.