Non-compliance risk analysis

context

Since the entry into force of the Euro 6 standard in 2015, the certification of thermal vehicles has been carried out in real traffic conditions, on driving cycles called RDE cycles (from the English acronym “Real Driving Emissions”). The great diversity of these cycles (due to the weather, topography, topography, road network and local legislation or even driving style) represents a real challenge for car manufacturers who must control the risk of non-compliance of their vehicles in the multitude of cases that may be encountered.

Taillights of cars in an overnight traffic jam on a highway with blurred lights in the distance.

compliance factor

The evaluation of a vehicle on an RDE cycle is carried out by calculating a compliance factor, which is based in particular on the mass of pollutants emitted per kilometer traveled. To meet the standard, this compliance factor must be less than 1. It is crucial for a manufacturer to assess what is the risk (i.e. the probability) of his vehicle obtaining a compliance factor exceeding the standard, on an RDE cycle drawn at random.

approach

The method developed makes it possible to generate the probability distributions of the conformity factors of each of the pollutants. The results are presented in the form of a dashboard allowing to dynamically filter on the different driving styles, cycle phases (start, city, road, highway), weather conditions, mass embedded in the vehicle, etc. and thus to evaluate the performance of the vehicle in various situations.

This dashboard also makes it possible to sort the cycles according to their severity and pollutant emission level, to ergonomically identify the most demanding cycles from the database of thousands of cycles and to display their characteristics in detail: GPS track, speed profile, altitude profile, altitude profile, temporal evolution of emissions during the cycle, etc. The tool offers all the turnkey information manufacturers need to drive these cycles on site with their vehicle.

innovation

ONE STATISTICAL METHOD TO CONTROL THE RISK

OSE has developed a robust analysis of the risk of non-compliance, based on representative RDE cycles and a rigorous control of emission model confidence intervals.

Christophe LECLERCQ

General manager

methodology

The method for analysing the risk of non-compliance with pollutant emissions standards developed by OSE is based on 3 pillars:

GENERATING RDE CYCLES

Thanks to the OseRoad tool, hundreds of thousands of RDE cycles are generated to cover geographic, topographic, meteorological, legislative and driving styles diversity.
The database is then reduced to a few tens of thousands of cycles via discriminating physical variables and advanced clustering techniques, in order to make it usable for numerical simulations in a reasonable time.

EMISSIONS MODELING

OSE designs pollutant emission models — or uses those from manufacturers — calibrated via bench measurements or PEMS (Portable Emissions Measurement System).
The objective: to simulate vehicle emissions over all the cycles in the database.
A dedicated uncertainty model makes it possible to control measurement errors and those of the emissions model.

Development of a selection and reconstruction algorithm

The results are post-processed to calculate the probability distribution of the compliance factor and the relevant characteristics to select and filter the cycles.

This stage provides global risk indicators while allowing access to the smallest details of the most demanding cycles.

Non-compliance risk analysis

Or how to master the complexity of evaluating vehicles in real traffic conditions.

Risk analysis: From the global indicator to the finest granulometry

From the compliance factor to the finest criteria: a complete vision of the risk. Do you want to go further?

ONE STATISTICAL METHOD TO CONTROL THE RISK

OSE has developed a robust analysis of the risk of non-compliance, based on representative RDE cycles and a rigorous control of emission model confidence intervals.

GENERATING RDE CYCLES

EMISSIONS MODELING

Development of a selection and reconstruction algorithm

a single conversation can spark Innovation