a PERFORMANCE EVALUATION OF ROUNDABOUTS USING A MICROSCOPIC SIMULATION MODEL

Summary . Contemporary planning of the road network in cities assumes the use of transport models. The microscopic model allows provision of information and decision-making regarding, for example, emissions, traffic organisation, geometrics of the intersection and traffic control system. This article describes the development of comprehensive methodology applied for carrying out the research related to the design of geometrical parameters of the roundabout. The methodology involves developing instructions for preparing solutions for performance evaluations of roundabouts using a microscopic simulation model with the VISSIM software. The application of the developed methodology was prepared on the case study of the roundabout in Gdańsk. The results of the prepared case study were discussed.


INTRODUCTION
An intensive process of development of the new intelligent transportation systems (ITS) was observed. ITS connects many types of services and technologies with the goal of improving the effectiveness of transport systems [1,2,3,4]. Contemporary planning of the road network in cities assumes the use of transport models. Three types of transport model can be distinguished [5]:  macroscopic modelhighest level of aggregation, lowest level of detail  mesoscopic modelhigh level of aggregation, low level of detail  microscopic modellow level of aggregation, highest level of detail In this article, the authors examined the microscopic model, which describes the impact between individual vehicles in traffic [6]. The microscopic model allows provision of information and decision making regarding: emissions, traffic organisation, geometrics of the intersection, traffic control system and others [7]. Many simulation tools exist for developing traffic microscopic simulation models, for example, VISSIM, SimTraffic, AIMSUN, CUBE Dynasim, SUMO and others [8]. In this study, the VISSIM software was used for the case study problem investigated.
The microscopic simulation traffic model is a useful tool for analysis of the impact of planning new infrastructure. The author in [9] analysed the estimation of effects caused by the implementation of the Park&Ride system into a transport hub with the CUBE Dynasim model.
A simulation software was used to test a special type of traffic control. The dynamic management of traffic lights at pedestrian crossings with fuzzy logic controllers was tested with an external interface module in the VISSIM software for extensive simulative assessments [10].
Air quality is an extremely important factor in urban centres. The authors in [11] used a life cycle assessment model and prepared a microscopic simulation model with the VISSIM software for a comparative study of emissions by road maintenance works.
This article has two main aims. First, the authors described the development of methodology applied for carrying out the research related to the design of geometrical parameters of the roundabout. The methodology involves developing instructions for preparing solutions for performance evaluations of roundabouts using a microscopic simulation model with the VISSIM software. The second aim of the paper was connected with the application of the methodology to the case study of a roundabout in Gdańsk.
The article is organised as follows; section 2 provides information about the research problem and a description of the case study. The next section presents the development of the methodology applied to carry out the defined research problem. Section 4 contains the results of using a microscopic simulation model prepared through the developed methodology for the case study. Section 5 of the paper contains conclusions and propositions for future work.

CHARACTERISTICS OF THE RESEARCH PROBLEM
The research problem is related to the proposition for geometrical changes in the roundabout at Gdańsk. The changes are associated with intensive cargo traffic from the road inlet from the port of Gdańsk. Prognostic analysis of road traffic indicates an increase in traffic flow in subsequent years in the analysed inlet. The aim of solving the problem is to Performance evaluation of roundabouts using a microscopic simulation model 59.
compare traffic conditions depending on values of the traffic flow in the proposed variants of the design of the roundabout inlet for all prognostic years. The cargo traffic flow growth related to the development in the port was calculated in all prognostic years.
The need to carry out the analysis of such a problem in the above-mentioned scope is related to the transport service of the port in Gdańsk. Figure 1 presents the existing state of the analysed roundabout in Gdansk with the inlets descriptions. Solving the problem requires the development and validation of the following variants:  the existing state (mapping the process of handling traffic by crossing in the current traffic situation) and in all traffic flow prognostic years  proposals to change the geometry of the intersection on one of its inlets (E2) in all traffic flow prognostic years (variant 1)  proposals to change the geometry of an intersection on one of it's inlets (E2), while eliminating traffic on one of the relations that uses this intake (relation from inlet E2 to outlet S), because instead of a roundabout, this relation will be served by an overpass (variant 2) in all traffic flow prognostic years

METHODOLOGY OF THE RESEARCH PROBLEM
A comprehensive methodology for solving the indicated research problem was developed. The developed methodology allows for the assessment and provision of solution for performance evaluations of roundabouts using a microscopic simulation model with the VISSIM software. The selected traffic simulation software was developed by the PTV Group in Germany [12]. The VISSIM software is a comprehensive tool used in developing the microscopic simulation model, which also has the capability of rendering a 3D visualization [13].
The methodology described can be used to solve other research problems other than the described case study in the article. The implementation of the research based on the developed methodology attends to all the relevant aspects related to the development of the microscopic traffic simulation model. The developed method was divided into four stages:  defining and characterising the research problem  development and analysis of the research and measurements  microscopic simulation model development  analysis and evaluation of the simulation results Each stage is connected with the other. Solving the research problem with prepared methodology requires the usage of all stages. All of the stages were described using schemes as presented in Figures 3 to 6.
The first step of the methodology deals with defining and characterising the research problem. Figure 3 shows the scheme of the first step of the methodology. The five main categories related to the scope of the research problem were identified. The scheme considers all the important factors for solving the research problem. Defining the factors presented in Fig. 3 is extremely important because it directly influences the subsequent stages of the methodology. The methodology in that step showed the necessary measurements to be done.  Figure 4 shows the scheme of the second step of the methodology: development and analysis of the research and measurements. The second step describes the necessary analysis of the obtained data from the first step. The traffic and geometric data analyses are necessary to build a microscopic traffic model. The geometrical measurements are often conducted using geo-information software or with devices for non-contact measurement of geometric values. The traffic data were collected using different approaches, for example, video detection, manual counting. Performed analyses of the acquired traffic data require the use of standards described in the literature depending on the country, for example, HCM [14], HBS [15], Poland [16].
The methodology of the microscopic simulation model development was presented in Fig. 5. The diagram includes all components related to the development of the roundabout microscopic simulation model. The development of the model is possible only after the implementation of the previous steps. The development of the model is related to the geometrical structure of the intersection represented by, for example, the VISSIM software components, such as links, nodes, connectors, vehicle routes. The components are parametrised and define using analysed traffic data from the previous step. Analysis and evaluation of the simulation results scheme were presented in Fig. 6. The scheme was presented using a block diagram. The basis of the evaluation algorithm is to define the evaluation criteria and the selection of assessment measures. Defined in detail, the criteria grants understanding of the results in different technical variants and analyses the changes in the prognostic years. In order to check the correctness of the developed microscopic simulation model of the existing state, validation was performed. Calibration was performed, for example, by comparing the obtained values of an average wait time or travel time from the model and measured by video techniques [17]. Only the calibrated microscopic simulation model can be used to obtain and compare results in the technical variants and analyses the changes in the prognostic years. Presented in a comprehensive manner, the methodology shows the preparation of the complex microscopic simulation model. The developed methodology for solving the indicated research problem was used for the case study presented in section 2.

CASE STUDY RESULTS
The microscopic traffic simulation models were developed using data from road traffic measurements. Based on the obtained traffic flow data, the percentage shares of peak hoursmorning and afternoon in the daily traffic flow, as shown in Tab. 1 were determined.
For further analysis, the morning peak hour was chosen, because the intensity of traffic in this period is greater, which enables mapping of the worst traffic conditions. Data from traffic prognostic years including daily traffic volume were calculated for hourly traffic based on traffic percentage share from the morning peak hour in the daily traffic (8.06%). For each prognostic years horizon, traffic intensity and percentage shares of the routes and vehicle types structure for all relations at the round-about were determined. All elements in the microscopic simulation model were placed according to the method of developing models using the PTV VISSIM software.
On the basis of average loss of time at the roundabout and individual at the inlets for particular horizon in prognostic years, levels of service were determined on the basis of the "Method for calculating capacity of the roundabouts" published by the General Directorate of National Roads and Highways in 2004 [18]. This is the current valid method of assessing traffic conditions at roundabouts in Poland. It is related to the methodology of determining average values of time losses by the VISSIM software. For the existing state, the program did not generate traffic flow, which was set by users. This resulted from the limitation of the E2 inlet capacity and the formulating queues of vehicles that were not served at the simulation time.
The elimination of E2 -S relation at the roundabout in the second variant caused a radical improvement of traffic conditions. Undoubtedly, this direction should be explored in further analysis and design work. However, this solution involves incurring significant financial outlays.

CONCLUSIONS
The developed methodology adequately tackled the research problem presented in the article and may be applied to similar research problems. The development of the model using the VISSIM software, due to its wide possibilities, effectively determine the impact of changes made to the roundabout on traffic conditions. The following conclusions were distinguished for the analysed case:  Changing the geometry of the E2 inlet at the roundabout has a positive effect on the traffic conditions at this inlet because in individual prognostic horizons average values of time losses are less by about 40 [s/veh.]. However, this does not solve all communication problems occurring at this intersection, because in 2034, this facility will be working on IV level of service. This was observed in both the analysis of the existing state model and the models taking into account the change of the inlet geometry E2.  The improvement of traffic conditions after the addition of the lane at the E2 inlet will increase the average values of time losses for relations on other inlets. This phenomenon was caused by the increase in the traffic volume on the roundabout envelope, which reduces the average time interval between vehicles. The increase in traffic is the result of both the prognostic change in traffic potentials in the immediate vicinity of the intersection as well as the general pattern of socio-economic changes manifesting, inter alia, an increase in the motorisation index and an increase in the population's mobility.  Elimination of E2 -S relation at the roundabout resulted in a radical improvement of traffic conditions. Undoubtedly, this direction has potentials for future analytical and design work.