EP2991870A1

EP2991870A1 - Method for optimising the energy consumption of a hybrid vehicle

Info

Publication number: EP2991870A1
Application number: EP14722262.4A
Authority: EP
Inventors: Maxime DEBERT
Original assignee: Renault SAS
Current assignee: Renault SAS
Priority date: 2013-05-03
Filing date: 2014-04-11
Publication date: 2016-03-09
Also published as: FR3005296A1; US20160167642A1; CN105246753A; FR3005296B1; WO2014177786A1

Abstract

A method for optimising the energy consumption of a hybrid vehicle on a journey on the basis of the energy management laws of said vehicle, the charge status of the traction batteries of same and the intended journey, characterised in that the distribution between the supply of combustion-induced torque and the supply of electrically-induced torque over the course of the journey is based on a prediction of the overall energy consumption over the journey, which is established on the basis of an estimation of the energy consumption and distribution between these two supplies, over different sections making up this journey.

Description

METHOD FOR OPTIMIZING THE ENERGY CONSUMPTION OF A

HYBRID VEHICLE

The present invention relates to the field of energy management of hybrid vehicles, having at least one source of thermal energy and a source of electrical energy.

More specifically, it relates to a method for optimizing the energy consumption of a hybrid vehicle on a route according to the energy management laws of the vehicle, the state of charge of its traction batteries, and the planned course.

This invention has a preferred application, but not exclusive, on plug-in hybrid vehicles, whose traction batteries can be recharged directly from a power socket on the power grid.

A common mode of management of energy on a rechargeable hybrid vehicle, is to prioritize first of all the electric discharge of the batteries, then to maintain the state of charge when the battery is lightly charged. This method is generally incompatible with a goal of reducing energy expenditure and protecting the environment. Depending on the distance and the profile of the planned route, it may indeed be more advantageous to ride in hybrid mode, even if it means reaching its destination with discharged batteries.

In order to be able to use the vehicle's energy (electrical and thermal) resources efficiently, the vehicle's energy management system must know the energy requirement of the vehicle and the amount of energy that can be recovered on the intended route. This need depends on a large number of parameters, such as the driving style, the environment (urban, highway, elevation profile), as well as various disturbances, specific to the vehicle (loading) or external (rain, wind, traffic density). , etc.). Publication US 2010/0305839 discloses an energy prediction method based on vehicle consumption patterns as a function of traffic conditions. This method does not include the particularities of each driver. It is therefore hardly compatible with an embedded energy management system.

The present invention aims to predict the energy category of the sections taken by a vehicle on a given route, in order to optimize the exploitation of its energy resources according to the particularities of the vehicle and the route.

For this purpose, it proposes that the distribution between the contribution of torque of thermal origin and the contribution of electrical torque on the route, is based on a prediction of the overall energy consumption of the route established according to a estimate of the consumption, and of the energy distribution between these two contributions, on various sections composing this course.

Preferably, the path is decomposed into sections in a database enriched with an estimate of the energy category of each section.

The present invention will be better understood on reading the following description of a nonlimiting embodiment thereof, with reference to the appended drawings, in which:

Figure 1 shows a network of curves representing fuel consumption according to the percentage of electrical energy used to make a kilometer, the average slope of the section, the energy category of the section

Figure 2 illustrates the classification of the road sections in the database used,

Figure 3 shows the results of a "logistic regression" on four energy segments categories, and

Figure 4 summarizes the optimization process. The invention proposes to exploit consumption curves of a hybrid vehicle, as a function of the percentage of electrical energy used. Figure 1 shows, as an example and comparison, a network of consumption curves for the same hybrid vehicle on a motorway cycle, a road cycle, an urban cycle, and bottling, to perform 1km with different average gradients.

The invention uses an on-board navigation system of a vehicle is always able to indicate the position and course of the vehicle. In addition, the system provides information on sections of the route, such as: average speed, number of lanes, lights, signs etc., allowing it to calculate the shortest, fastest route, and most importantly , the most interesting from the point of view of energy management. The proposed method is based on the operation of a specific cartographic database, by the navigation system.

This database is built from existing map data by listing a sufficient number of routes to establish a prediction model. A directory makes it possible to classify road sections, given by the mapping provider: in general, a section corresponds to a segment of road with identical characteristics. The sections can be from a few meters to several kilometers. They are classified according to the optimal distribution over the same distance found by an optimization algorithm using a calculation model, based on the fundamental principle of the dynamics, from information of course given by the vehicles, in particular the speeds and the slopes registered. This information also includes a network of curves such as those in Figure 1, showing the fuel consumption based on the percentage of electrical energy used to make a kilometer, on four different categories of sections. The sections are thus classified into energy categories, according to the shape of the consumption curve as a function of the percentage of electrical energy used. To characterize the pace we can for example rely on correlation functions.

The optimization of the energy consumption of a hybrid vehicle is carried out over an entire route, according to the energy management laws of this vehicle, the state of charge of its traction batteries and the planned route. For this purpose, an algorithm of the navigation system calculates an optimal distribution of energy between the thermal and electrical sources over the entire path planned to predict the energy requirements of the vehicle thereon. The prediction consists in estimating the energy category of the sections that the vehicle will use, thanks to the enriched database database. This demand classifies for example the planned route, in one of the four categories mentioned: traffic jam, urban, road and highway.

The database is advantageously built by recording on GPS vehicles their GPS position ("Global Position System"), and their speed. Each test run is then broken down into sections in the database, which is enriched with an estimate of the energy expenditure of the vehicle on each section. From the GPS coordinates and the sections borrowed, certain characteristics relating to these, are also noted. The optimization algorithm then knows how to find, on each of the sections, the optimal distribution of energy, minimizing the cost of displacement. As indicated above, the sections are ranked according to the shape of their curve closest to one of the pre-established categories, for example the four categories mentioned (traffic jam, urban, road and motorway).

The construction of the database is illustrated in a nonlimiting manner in FIG. 2 in the form of matrix dot clouds. In this example the sections are classified according to ten characteristic information:

• the type of section (among six categories),

• the maximum speed allowed on the section,

• the average speed observed updated with traffic info (default speed),

• the "attributes" of the road: presence of roundabout, bridge, urban, intersection, etc.,

• the "class" of the section (giving information on its maximum speed),

• a reference speed ("speed category"),

• the number of lanes (in the direction of traffic),

• traffic (presence or not of slowdowns, updated with traffic info)

• the presence of stop,

• the presence of traffic lights.

In this example, we see that all variables are relevant for sorting energy categories. For example a high regulatory speed shows a good correlation with the highway and road categories. It can be completed by the trips made by the customer if he wishes (recording of his trips).

The invention provides for implementing from these data, a statistical model for predicting the energy class of the course. This model is advantageously constructed using the "logistic regression" technique used in many fields, such as medicine and banking. However, other methods of classification / discrimination may be suitable (such as decision trees, neural networks, etc.), and applied to realize the invention.

The model of the logistic regression is for example of the form:

Pr (G = K - 1 \ X = x) _{rj aT}

l ^0g Pr (G = K \ X = x) ^{= k} -V ° ⁺ '^ - ^| ·

The model is specified in if - 1 logarithm function reflecting the constraint that the sum of the probabilities must be equal to 1. A simple calculation yields:

The estimation of the logistic regression model is done using the maximum likelihood method popularized by the statistician and biologist R. A. Fisher. Since Pr (G \ x) satisfies the distribution conditions, the log-likelihood function for N observations is written:

NOT

i = l

Once the optimization algorithm has identified the model parameters (equation 1) on the identification data, its validity must be verified on validation data. Figure 3 shows the results of the logistic regression for the four categories on the validation data. Continuous lines represent the iso-probabilities of belonging to a given class. The section category predictions obtained by this method are 97% reliable.

In summary, the distribution between the contribution of torque of thermal origin and the contribution of electrical torque on the route is based on a prediction of the overall energy consumption of the route established according to an estimate of the consumption and of the energy distribution between these two contributions on different sections composing this course. The implementation of the invention requires having a sufficient database to classify the sections and predict the category of a course. This database can be continuously enriched from data collected on vehicles in transit, so as to feed a reliable model of energy prediction. This model is preferably of the "classifier" type, for example a logistic regression model. It is preferably embedded on a vehicle in a navigation system, so as to communicate to the calculator performing the energy optimization, the probabilities of future energy demands. The database can also be updated from a driver training using the vehicle, in order to optimize the strategy for the latter.

As indicated in FIG. 4, the onboard GPS calculator, or a mobile communication tool of the "Smartphone" type, is capable of establishing a future of travel by breaking it down into borrowed sections in order to predict the energy consumption of the journey. The "category of route" information is then used within the vehicle calculator (HEVC) to determine the distribution between the electrical and thermal energy input on the path.

With other information, slope, length of the sections, the latter is able in application of the laws of energy management of the vehicle (LGE), the amount of electrical energy to be used on the sections to minimize the consumption of the vehicle and to optimize the energy stored in the vehicle's batteries. Preferably, the discharge curve of the battery on the path, minimizes the overall energy consumption of the vehicle.

The advantages of the invention are numerous. Among these, we can mention:

• the ability to adapt the energy prediction to the driver and his driving style, thanks to to the possibility of learning,

• reducing the consumption of plug-in hybrid vehicles, and

• the possibility of supplying electric power for regulated urban areas to "zero emission" vehicles.

Finally, it should be emphasized that the application of the invention is mainly automotive, but that the media used for its implementation are multiple ("smartphone", tablet, navigation computer landed, portable GPS, infrastructure calculator etc.. ).

Claims

A method for optimizing the energy consumption of a hybrid vehicle on a journey according to the energy management laws of the vehicle, the state of charge of its traction batteries and the planned path, characterized in that that the distribution between the contribution of torque of thermal origin and the contribution of electrical torque on the route, is based on a prediction of the overall energy consumption of the route established according to an estimate of the consumption and the energy distribution between these two contributions on different sections composing the planned route.

2. Optimization method according to claim 1, characterized in that the path is decomposed into sections in a database enriched with an estimate of the energy category of all sections.

3. Optimization method according to claim 2, characterized in that the sections are classified according to different criteria, to determine the optimal distribution of the energy requirement on each of them.

4. Optimization method according to claim 3, characterized in that the sections are classified according to the shape of their consumption curve as a function of the electrical energy used.

5. Optimization method according to claim 4, characterized in that the sections are classified into four categories classified as motorway, road, urban and traffic jam according to the shape of their consumption curve.

6. Optimization method according to one of the preceding claims, characterized in that the cumulation of the sections determines, using a statistical model, belonging course to an energy category to predict the energy needs of the vehicle on it.

7. Optimization method according to claim 6, characterized in that the category of the course is operated within the vehicle computer to determine the distribution between the supply of electrical energy and heat on the path.

8. Optimization method according to claim 5, 6 or 7, characterized in that the discharge curve of the battery on the path minimizes the overall energy consumption of the vehicle.

9. Optimization method according to one of the preceding claims, characterized in that the database is updated from a driver training.