WO2013009178A2

WO2013009178A2 - Method and device for determining the charging behaviour of electric vehicles and a charging system incorporating such a method

Info

Publication number: WO2013009178A2
Application number: PCT/NL2012/050497
Authority: WO
Inventors: Egbert Wouter Joghum Robers; Crijn Bouman; Ali UGUR; Johannes Gerardus KAPTEIN
Original assignee: Abb B.V.
Priority date: 2011-07-11
Filing date: 2012-07-11
Publication date: 2013-01-17
Also published as: WO2013009178A3; NL2007081C2

Abstract

The present invention relates to a method for determining the charging behaviour, comprising obtaining the parameters from an electric vehicle, estimating the charging characteristics based on at least the parameters, measuring the actual charging characteristics of the electric vehicle and using the actual charging characteristics to improve the estimation of the charging characteristics.

Description

Method and device for determining the charging behaviour of electric vehicles and a charging system incorporating such a method

The present invention relates to a method for determining the charging behaviour for electric vehicles and a charging system incorporating such a method.

Due to technological, economic and environmental reasons the electric vehicle has become a reality. The ever increasing energy demand and the decreasing of the energy resources have lead to a dramatic increase of the oil prices. Besides this the high carbon emission of the internal combustion engines and the environmental problems caused by it, has made the use of electric vehicles necessary.

For electric vehicles to become widespread electric vehicle chargers have to be placed at various locations. Several aspects with respect to availability and ease of use will gain more importance as the electric vehicle infrastructure market matures. In urban areas or along highways chargers are more frequently visited than in the countryside and to create maximum availability it is therefore an advantage if they have multiple ports for charging several vehicles simultaneously. A multi-port charger is a cost-efficient alternative compared to installing two separate chargers. For multi-port chargers it is important to distribute the power capacity efficiently among the different ports. Another aspect concerning availability of chargers is that the vehicle driver wants to know the progress of charging. Vehicle drivers are especially interested to know when they have obtained sufficient energy in their battery to continue their journey. To be able to do this the charging characteristics of the electric vehicles has to be estimated accurately. A third - slightly different - aspect which will gain importance when many chargers are installed is the power which has to be delivered by the grid. Because of the concerns over strain on the grid and, more important, transformers and other parts of the local distribution system, the grid operators wants to know where and when the electric vehicles will be plugged in and how much their energy demand will be. To produce the aforementioned information for the electric vehicle user, charger or any other party the charging characteristics of the electric vehicle's has to be determined. But this is not a simple task because the charging of an electric vehicle battery is a highly non-linear process which depends on many factors, especially in the case of so- called DC fast charging. These factors can be both internal vehicle factors as external factors. The dependency of these factors may even mean that charging behaviour of the same vehicle is different depending on the actual situation or state the vehicle is in at that particular moment. When multiple vehicles of different make are being considered the situation becomes increasingly complex.

The European Patent application EP 2 219 278 discloses a charger with communication means for obtaining charge characteristics from vehicles connected to the charger via power exchange ports. The charger however, comprises no power converter.

US Patent US 2010/010704 discloses a storage unit with historic charge data. However, the charge data stored are for a specific vehicle, and can be retrieved only based on the ID of that specific vehicle. This system therefore has the disadvantage that for each specific vehicle, historic data of said vehicle is required in the database. That means, the system cannot determine charge parameters for an yet unknown vehicle.

It's a goal of the present invention to provide a method and system for determining the charging characteristics of electric vehicles, which overcomes the above mentioned disadvantages, or at least to provide a useful alternative.

The invention thereto proposes a method and a device according to claims 1 and 15. The present invention is an electric vehicle battery charger which is configured for estimating the charging characteristics of an electric vehicle at least based on the parameters it receives from the electric vehicle. One could think of parameters such as battery voltage, battery capacity and state-of-charge. The charger uses recorded charge sessions to improve the ability of estimation. The recorded charge sessions may be from the same vehicle, but also from other vehicles, so that also yet unknown vehicles can be charged optimally, based on known vehicles with corresponding characteristics. The estimation can be done each time an electric vehicle interacts with the charger. This could be for example the case when the electric vehicle connects to the charger to receive power or when the electric vehicle is on the way and communicates with the charger. The method could furthermore use external parameters as input for the estimation, for instance parameters given by a user or parameters received from a data network. The method may comprise reserving a power budget for each of the electric vehicles connected to the charger based on the estimated charging characteristics and charging the vehicles according to this power budget. A power budget could consist of a maximum current, a maximum voltage, a maximum power rating or combination of these items. A power budget could also involve a time parameter: for instance the maximum current which can be drawn over time, the maximum power which can be delivered over time or a maximum voltage over time. The method may further comprise calculating the charge time and the amount of energy provided at a certain moment based on the estimated charging characteristics and providing the calculated charge time and amount of energy to the electric vehicle or its user. The method may further comprise aggregating the estimated charging characteristics of a plurality of electric vehicles and providing the aggregated estimated charging characteristics to the at least one third party. A third party is for example a grid provider or a power plant.

The charger configured for determining the charging behaviour, comprises means for obtaining the parameters from an electric vehicle, means for obtaining characteristics of the actual charging of the electric vehicle, means for estimating the charging

characteristics based on at least the obtained parameters and using the characteristics of the actual charging to improve the estimation of the charging characteristics, at least one first power exchange port, for exchanging power with a power supply and at least one second power exchange for exchanging electric power with a battery of a vehicle to be charged. A means for estimating the charging characteristics could be a processor, controller, server or any other computing device.

The present invention provides the advantage that the charging characteristics can be estimated no matter what the physical conditions, vehicle type, battery chemistry, charge protocol or cell design is. The invention has the ability to generate a estimation of the charging characteristics of an electric vehicle when only very limited information on the vehicle is available. The invention has the ability to generate a estimation of the charging characteristics of an electric vehicle when this behaviour is highly non-linear and dependent on various factors including external factors. The invention furthermore has the ability to improve itself each time it charges an electric vehicle.

From the prior art it is known that charge stations have the difficulty to estimate the charge time correctly. In most cases there is no estimation mechanism but only a fixed end-time setting or a pre-programmed expectancy which usually is a very poor representation of the actual charging characteristics. In most cases the charging characteristics is taken as a linear process which is then extrapolated over time. This gives a very poor estimation of charge processes, especially in the case of so-called "fast charge" processes as these are highly non-linear. As an example of the non- linearity of fast charging one could consider that during a fast charge the differences in charge power can sometimes be more than 80% between the first phase of charging and the last phase of charging. By using the present invention we are able to estimate the charging characteristics and charge time correctly and provide it to the electric vehicle user or utilize it for other purposes.

It is also possible that the estimation is done during charging, providing an estimation for the remaining part of the charge session. An on-the-fly estimation is done to estimate charging characteristics in the future time t+T, based on the current state.

Charge stations are distributed locations over country, therefore the grid provider has a big advantage in knowing the power demand of the charge stations in the coming hours so it could take measures to prevent imbalance or power outages. The estimated charging characteristics of all the vehicles which are connected or have the intention to connect to the chargers are aggregated and delivered to the grid provider or any other third party.

In one embodiment of the invention, the estimation could be based on more than just the basic parameters at the start of a charge session. One example is using the outdoor temperature in the neighbourhood of the charger. Research has shown that the charging characteristics of an electric vehicle is dependent of the temperature of the battery. In some cases however it is not possible for the charger to obtain the battery temperature via the communication channel with the car. In that case one could utilize a

measurement of the outdoor temperature close to the charger as an indicator with a correlation to the battery temperature. The outdoor temperature could for example be obtained via a sensor or via a computer network link.

There are two kinds of charging characteristics defined in the method. The estimated charging characteristics which is a forecast for the actual charging characteristics, and the actual charging characteristics which refers to the charge session over time which really takes place.

The database of recorded charging characteristics can be kept locally in the charger while being updated on a regular basis via a web link to a larger database. This has the advantage that if the weblink is down the system still works correctly. The local database in the charger can be updated with different algorithms for different locations. It could be that at cold locations for instance it is better to have a different kind of charging characteristics in the database than for relatively warm locations.

The electric vehicle may comprise any electric powered automobile, scooter, airplane, truck, motorcycle, spacecraft, locomotive, boat, bicycle, plug in hybrid, fork lift, submersible vessel and agricultural machine. In a preferred embodiment a charger is a multiport DC fast charger. The benefit of such multiport DC charger is that the power converter to convert the AC power from the grid to the DC power used to charge battery can be shared by the two outputs, thus providing a cost advantage over two separate chargers, as the power converter is the most expensive component of a DC charger. In such charger usually the power rating of the first power exchange port, which connects the charger to the AC power supply, is lower than the sum of the power ratings of the second power exchange ports (outputs), which connect the chargers to the vehicle. The challenge in designing such multiport DC charger is the distribution of power over the two outputs in such way that both vehicles are charged simultaneously while the power supply or power converter are not overloaded. The charger therefore needs some kind of a decision making unit which has to decide over time which amount of power capacity is allocated to which output.

Dividing the power over the outputs of a multiport charger is however in many cases not as simple as it seems. This is caused by the fact that some fast charge standards available in the market today operate according to a so-called master-slave concept, where during a substantial part of a charge session the vehicle acts as the master which controls the charger as a slave. A good example of such charging standard is the CHAdeMO standard which is applied in several electric vehicles today. The above means that there are limited possibilities for the charger to make decisions on the power level at a certain output (secondary power exchange port) of the charger. Fast charging processes are usually highly non-linear processes which are dependent on multiple factors such as the battery temperature, external temperature and several start conditions. Considering the above, it becomes clear that it is very complex from a charger point-of-view to estimate the charging behaviour, especially when the vehicle acts as a master controller. This will be explained in the detailed description in more detail.

The invention is not limited to a specific physical arrangement as there are many possibilities for arranging the system. The invention will now be explained into more detail with reference to the following figures. Herein:

- Figure 1 shows a typical power curve of charging an electric vehicle;

- Figure 2 shows a graph of three relevant charging curves set against time;

- Figure 3 shows a charging system that is capable of estimating the charging characteristics of an electric vehicle based on past experiences;

- Figure 4 shows a plurality of chargers which can communicate with each other;

Figure 5 shows an embodiment wherein the method for determining the charging characteristics is implemented in the controller of multiport charging station;

- Figure 6 shows an implementation of the method for estimating the charging characteristics;

- Figure 7 shows another implementation of the method for estimating the

charging characteristics;

- Figure 8 shows a different setting of the present invention;

- Figure 9 shows another example which displays the typical charging

characteristics;

- Figure 10 shows a specific embodiment of a charger which is equipped with a user interface which can inform the user on the progress in charging;

- Figure 11a shows a flow diagram of an embodiment wherein the vehicle user can give the preferences for charging; Figure 1 lb shows a charge curve wherein the user preferences are taken into consideration;

Figure 1 illustrates a typical power curve of charging an electric vehicle based on the agreement between the charger and the electric vehicle on the maximum power of 50 kW. In many cases such agreement is made between the electric vehicle and the charger before the actual charging starts. When charging starts the control of the session is carried over to the vehicle and the charger acts as a slave. The value of 50 kW is in this case the maximum power which can be provided by the multiport charger as a whole: the sum of the power delivered by the multitude of outputs cannot exceed this value.

One can see that after the first part the charge power is significantly less than the agreed maximum of 50 kW. Our research has shown that the shape and form of the charge curve are related or correlated to the start parameters of the charge session. Such parameters are for example the State of charge (SOC) at the start, the battery voltage and more advanced parameters such as battery temperature, impedance of the battery or voltage rise after applying a charge current. If, while charging the first vehicle, a second vehicle would arrive after 15 minutes which plugs in to another output of the multiport charger for example, it would not make sense to stick to the agreement of keeping 50kW reserved for the first vehicle. We can see from the graph that there is already approximately 20kW which is not utilized by vehicle 1 and this amount is becoming even more at later moments during the session. Based on this knowledge of the "future situation" the charger could "give away" a power budget of at least 20 kW to the second vehicle from the start. It is however necessary for the charger to know that the power demand from the first vehicle will not suddenly increase after it has given away 20 kW to the second vehicle. The charger would need to have some kind of estimation of the power drawn by vehicle 1 in the remainder of the charge session to make sure that the total power limit of the multiport charger is not exceeded.

The invention as described in this document could be utilized to create such estimation. In this example the charger is equipped with an internal computer containing an extensive list of the results of many charge sessions. Based on all the parameters present at the start of charging vehicle 1 the charger selects from the list of result the 3 most relevant charge curves. The selection could be based on a comparison of the parameters present at the start of charging with the parameters in the list. These curves could be voltage, current or power curves against time or a combination. These curves could be represented as follows in figure 2.

The graph displays 3 power curves and two boundary lines. The charge curves (thin lines) represent the charging power over time of 3 earlier charge sessions with more or less identical start parameters. It can be seen that there is some difference in charging power over time for the 3 curves. The solid boundary line represents the highest forecasted charge power over time while the dashed line represents the lowest forecasted charge power over time. The boundary lines can be used to create a estimation of power drawn by vehicle 1 after charging has started. The lowest forecasted charge power boundary will produce a best case estimation of the amount of power drawn, the highest boundary line over time produces a worst case estimation. There is a high likelihood that all charge sessions will have a power pattern which lies within the envelope between the highest and the lowest boundary. To keep the system safe the software on the internal computer is programmed such that it will take the highest forecasted power boundary as the estimation of the power which will be required by vehicle 1. The charger can now utilize the estimation to release a power budget at the moment a second vehicle 2 arrives. When this vehicle plugs in the computer could look at the highest forecasted power line and take for example the highest value expected to be required by vehicle 1 in the near future and release the rest of the power for vehicle 2. Vehicle 2 will now be charged according to the budget. The advantage is that the charger now has the ability to safely give away a power budget based on the estimation, while in a situation without this estimation method it could not safely give away any power budget until charging of the first vehicle is completed.

Because the estimation is based on a list of previous results there is a high likelihood that this estimation covers the charging characteristics, but there is of course not a 100% certainty that this is the case. It could be that vehicle 1 charges with slightly different behaviour, while the start conditions of charging are comparable to the 3 cases in the graph. This could cause a problem if at some point in time after the budgets are given away the power request is higher than the maximum of the forecasted power. In this case the power converter or the power supply in the charger could be overloaded. The computer could be equipped with a software implemented method to detect this and safely terminate the charge session at port 2 in the case that continuation would cause unsafe situations. When this has happened the computer in the charge station could add the resulting fast charge curve at port 1 to the list of charge sessions. Now the behaviour of this new car is also in the list and in future moments the charger can anticipate on this new car and new behaviour at the beginning of the charge sessions. Such system could be seen as a self-learning system which improves in performance over time. All above described calculations performed in the computer could also be performed in a networked system, cloud computing system or any other configuration of controllers and computers.

In a more advanced example the above multi-port charge station could be equipped with a computer or other form of intelligence which implements a stop-start sequence. This stop-start sequence is basically a sequence where charging is stopped and restarted. The advantage of this method in combination with certain charge protocol standards like CHAdeMO is that when charging is restarted the negotiation phase of charging is done again and new parameters could be negotiated with the vehicle. The estimation method as described in the above example could be used in combination with such stop-start sequence to redistribute the power budget on both outputs of a multiport charger somewhere in the charge session.

In more detail, based on the power estimation of the charge session of port 1 and the power estimation of port 2 the computer can calculate whether it makes sense to redistribute the power more drastically by stopping one of the charge sessions, or both and change the power limitation on both outputs. Figure 3 Shows a schematic view of the present invention. It depicts a charging system that is capable of estimating the charging characteristics of an electric vehicle based on past experiences. The charger stores recorded charge sessions in a database. When a vehicle connects to the charge station the electric vehicle parameters are provided. Non- limiting examples of the parameters are the Battery State of Charge and the battery size. Furthermore the battery voltage can be measured. The charge controller will look up recorded charge sessions in the database with similar parameters and uses these records to estimate how the charge session will progress. The estimated behaviour can be used to calculate the charge time and the power delivered to the vehicle battery. The estimated behaviour could also be used to reserve a power budget for the vehicle and communicate a power limit to the vehicle which is not to be exceeded during charging.

A slightly different embodiment (illustrated in figure 4)than the above multiport example, is the following example of multiple chargers which are connected to one AC power supply. In this embodiment a plurality of chargers is connected to, for example, one single distribution board supplying them with AC power from the grid. The plurality of chargers may consist of AC chargers, DC fast chargers or a combination of both. In some cases it may be advantageous to connect the plurality of chargers to a distribution board in such way that the power rating of the sum of the input connections of each charger (first power exchange ports) is lower than the sum of the output connections (second power exchange ports). The advantage lies in the fact that is some cases it is very expensive to upgrade a utility supply connection or distribution board. As a simple illustration we could mention a situation where two 50kW DC chargers, both with output ratings of 50 kW DC, and input ratings of 55 kVA AC, are connected to an AC supply with a total rating of 80 kVA.

In the above situations we have a technical challenge which is very similar to the multiport station: we have to prevent the supply connection from being overloaded while at the same time offering the capability to charge electric vehicles at all outputs simultaneously . The method according to this invention could be utilized to prevent overloading the supply in a very similar manner as the previous example. Figure 4 shows that a plurality of chargers could for example be equipped with communication means to communicate to each other. This could be a wired or wireless connection, or the chargers could even be part of a larger data network. Each charger could utilize a estimation mechanism according to the invention and based on the power estimations on each port negotiate with the other chargers how much power it is allowed to draw in the near future. The method used to decide on the power budgets for each individual charger could be very similar to the method used in the previous example involving the multiport charger.

In a more advanced example, the charging system can be equipped with a method to adjust the estimation based on parameters which are present during charging. A more detailed example could be as follows: Our research has shown that the internal resistance of a battery is one of the important indicators of the expected charging characteristics of a vehicle. The internal resistance however can in most cases not be detected before charging has commenced. The system could be programmed to act in the following matter: When the vehicle arrives, the start parameters are obtained via a data connection with the vehicle. Based on these values a estimation of charging characteristics according to the method is made and a power budget is calculated. The power or current limits are communicated to the vehicle and charging will start according to this budget. During the first phase of charging detailed parameters of the actual charging characteristics are monitored. More specifically the voltage rise over the battery over time and the current ramp-up over time are monitored. From these two values a good understanding of the internal resistance of the battery can be obtained. A way to determine the internal resistance or impedance during the current ramp-up is to divide the voltage rise (delta V) by the current rise (delta I). For example, when a vehicle starts charging the current ramp-up can be 20 Ampere per second to 120 A in 6 seconds. During this period the measured voltage rise can be 12 Volt or 2 Volt per second. The internal resistance can easily be calculated by dividing 12 Volt by 120 Ampere or 2 Volt by 20 Ampere, resulting in a calculated internal resistance (or impedance) of 0.100 Ohm. After calculating these values the database can be searched for charge sessions with a internal resistance which is close to the internal resistance observed. Based on the outcome the expected charging characteristics could be estimated more accurately and based on that a decision could be made on how to proceed. In some cases this could mean that based on this new estimation there is the need to change or optimize the charge process and calculating a new power budget. As an example the charge process could be stopped and restarted with the purpose to communicate a new power limit to the vehicle and charge the vehicle according to this new limit. Another example could be to adjust a estimation of the charge time or the amount of energy charge at a certain moment to obtain a higher accuracy

Figure 5 shows an embodiment wherein the method for determining the charging characteristics is implemented in the controller of multiport charging station. Two vehicles are connected to the outputs of the charging station. The controller

communicates with vehicles and receives the parameters from the vehicles. Based on stored charge profiles from previous sessions and the received parameters from the vehicles the controller estimates the charging characteristics for the two vehicles. Based on the estimated charging characteristics the charging strategy is determined for the vehicles. The controller controls the power converter according to the commands received from the vehicles and the vehicles are charged. The actual charging

characteristics of the vehicles is recorded by the controller and saved in its database. Figure 6 shows an implementation of the method for estimating the charging

characteristics. Charge summaries of previous charge sessions are stored in a database. The summary describes the complete charge session in a plurality of parameters. When a new vehicle arrives at the charger, several parameters from the vehicle are provided to the charger. These parameters are used to filter charging characteristics from a plurality of charge summaries which are stored in the database. A non-limiting example of the filter is shown in a flow diagram.

Figure 7 shows another implementation of the method for estimating the charging characteristics. An adaptive system, in this case a neural network, is trained with real data from the electric vehicle. The neural network is fed with the parameters, for example the battery size, voltage and the requested current by the electric vehicle. From these parameters a first estimation is made for the charge parameters. For training, the estimation is stored in the memory. During charging, the estimation is compared with the actual charging characteristics. The result of this is an error, which is fed back into the neural network to train it. The training can be done electric every time a vehicle is charged, or only at certain times or electric even only when the error is larger than a certain threshold.

The adaptive system can be any type of white box, grey box or black box system.

The estimation can be (for example) a table of values for current for electric every minute after the first estimation. The input nodes get parameters, and the output nodes give the charging characteristics as points in time. The error is fed back to the network to change the functions in the nodes in all the layers.

Figure 8 show a different setting of the present invention. The estimated charging behaviour for a plurality of charging sites can be accumulated at a server, and provided to a power plant, grid operator or any other third party. The cumulative demand forecast can be made for particular area or a certain subset of chargers. Based on the cumulative demand forecast the third party could act by buying or producing the demanded energy. Another possibility is that the third party provides feedback to the charging sites in order to reschedule their demand. This can be relevant when the charging estimations are done for vehicles on the road going to the charging site.

An example of an external parameter which could be used in a system according to the present invention is the charge time duration preference which is entered by a user or obtained via a data link to a computer system. In this case the system could take the preferred charge time or the amount of energy which is added to select a certain charging characteristics from the database. If we take as an example a CHAdeMO charger we know that in this case during the initialization phase the charger has to tell the vehicle the maximum current which the vehicle can request from the charger.

However the maximum power which is communicated also influences the amount of power delivered over time. As an example we could take figure 6 which displays the typical charging characteristics. Figure 9 shows two situations: one where the vehicle is given a 120 Amp limit and one where the vehicle is given an 80 Amp limit during the initialization phase. The graph clearly shows that if a vehicle is given a 120 Amp limit it will add a lot of energy to the battery in the first minutes of charging. When the car is given a 80 Amp limit, initially less energy will be added in the first minutes but in the last period of charging more energy is added per time unit. The system described in this invention could be utilized to estimate this behaviour and use it to decide on the power limit given to the vehicle. Limiting the power of a charger can be beneficial for instance to avoid peak-loads on the local electricity grid. As a practical example, a user could enter a desired amount of energy which he likes to be added in a certain amount of time (for instance 10 kWh within 30 minutes) and based on that input the system could access a database with charge profile estimations to select the lowest power limit at which the user preference is still met. In this way the charge profile estimation method actively helps to avoid peak loads on the electricity grid.

In one specific embodiment a charger is equipped with a user interface which can inform the user on the progress in charging (figure 10). Such interface could give the user a estimation of the end-time of charging or the amount of energy expected to be added at a certain moment in the nearby future, for example such user interface could tell the user that 8 kWh will be added in the first 5 minutes and another 6 kWh in the second 5 minutes. The user interface can be part of a charger or any other application in an electric vehicle or a mobile phone. To give a reliable estimation, the method according to this invention is utilized to estimate the amount of energy added to the battery at several moments during the charge session. The following text will describe the above in more detail.

In this example a user arrives at the charge station to get his vehicle charged. In this specific case it is a DC fast charge station equipped with a charging connection according to the international CHAdeMO standard which allows communication with the vehicle via a CAN bus interface and furthermore the ability to supply DC power to the battery inside the vehicle via separate power pins. When the driver connects the vehicle to the charger and presses the "start" button, the vehicle and charger start exchanging information via the CAN bus. The CHAdeMO communication protocol is used as a data format for the data exchange. The vehicle communicates parameters on the battery to the charger. One could think of parameters such as battery voltage, battery capacity and state-of-charge. The charger receives these parameters from the vehicle. Secondly the charger receives other parameters which are relevant to the charging process, such as the maximum power limit of charging and the outside temperature. This information can be received via a network link connecting the charger to a data system, for instance a smart grid system or a charging infrastructure management system. Under the CHAdeMO protocol standard the settings of a charge session are negotiated in the initialization phase of charging. Once the charger and the vehicle agree on a certain charging strategy (comparable to a handshake) the control is handed over to the vehicle. When the control is handed over to the vehicle, the execution phase of charging will start. During the execution phase the vehicle (master) is in control and will demand a certain current from the charger (slave). These so-called current commands have to be followed by the charger within a certain timeframe. If the charger does not deliver the requested current to the vehicle in time, the vehicle will terminate the charge session. This basically means that the charger has limited control of the charge session and also on the duration of a charge session. However, in this example the charger is equipped with an internal computer containing an extensive list of the results of many charge sessions. Based on all the parameters present at the start of charging the charger selects from the list of result the 3 most relevant charge curves. These curves could be represented as follows in figure 2. The graph displays 3 curves and two boundary lines. The charge curves (thin lines) represent the charging power over time of earlier charge sessions with more or less identical start parameters. It can be seen that there is some difference in charging power over time for the 3 curves. The solid boundary line represents the highest forecasted charge power over time while the dashed line represents the lowest forecasted charge power over time. The boundary lines can be used to create a estimation of the amount of energy added during a charge session at each point in time. For instance the user interface could tell the user that it expects 10 kWh to be added in the first 10 minutes and 5 kWh in the second 10 minutes of charging. Integrating the amount of power over time of the lowest forecasted charge power boundary will produce a conservative estimation of the amount of energy added, integrating the highest boundary over time will create an optimistic estimation of the energy added to the battery. The system may choose to present the conservative scenario to the user to avoid disappointments with the users. Many advanced user interfaces can be imagined for instance containing a estimation of the driving range which is added to the vehicle in the next 10, 20 or 30 minutes.

The method could also be used to display situation specific messages, for instance commercial messages, to a user of the system. These messages could for example be displayed on the user interface of the charging station, via a mobile phone or another internet based medium. As the method according to the present invention offers a estimation of the duration and energy delivery this knowledge could be used to select a specific commercial message which is appropriate for that charge session. For example, when the duration of the charge session is estimated to be 10 minutes, the system could display an advertisement for a cup of coffee. In the case that a 30 minute session is estimated it could display an advertisement for a lunch in a nearby restaurant. Such functionality could work as follows: When a user connects to the charge session a estimation is made. Based on the estimation the system selects a certain message from a resource such as a database containing multiple messages. The selection could be based on the estimated duration of the session. Local advertisers could fill the database with commercial messages and for each message indicate which charge time is appropriate for that message. The charging system now displays the selected message to the user.

This example is clearly not limited to commercial messages only; any message or type of information could be selected and displayed to the user based on the estimation according to the method.

Figure 11a shows an flow diagram of an embodiment wherein the vehicle user can give the preferences to the electric vehicle. This could be for example the amount of energy delivered to the electric vehicle within a time limit. The charger is implemented in such way that it has to give a power limit to the vehicle before charging starts. The charger estimates a charging characteristics with the lowest possible power limit within the preferences given by the user, such as a time limit and the amount of energy to be delivered. Figure 1 lb shows a way of doing this by, estimating charge profiles by different values of power limit (Plimitl and Plimit2) and selecting the lowest power limit which satisfies the user preferences (Plimit2 in this case). The vehicle charges then the vehicle according to this power limit.

Some vehicles types have an own system which can determine their charge time accurately. In an embodiment the invention can determine the type of the vehicle and depending on the vehicle may decide to use the charge time estimation delivered by the vehicle. It's also possible that the charger compares the charge time estimation delivered by the vehicle with the data stored in its database, and depending on the accuracy utilize it.

Claims

1. Method for providing an estimation of the charging characteristics such as power drawing, energy consumption and duration of charging a battery of at least one electric vehicle, comprising:

Obtaining at least one parameter from the at least one electric vehicle;

- Estimating charging characteristics by calculation means based on the at least one parameter;

- Providing the estimated charging characteristics to at least one from the group of the charger, a controller inside the charger, the vehicle, a user of the vehicle, a webserver, a grid provider, grid operator or a power plant;

Comparing the at least one parameter with pre-stored parameters of various electric vehicles parameters in a database;

Selecting charging characteristics from a vehicle in the database of which the stored parameters are best matching the at least one parameter;

- Providing the selected characteristics as an estimation of the charging

characteristics.

2. Method according to claim 1, comprising:

- Updating the calculation means based on characteristics of the actual charging of the vehicle and the at least one parameter.

3. Method according to claim 2, comprising

- Updating the database by recording characteristics of the actual charging of the vehicle and the at least one parameter.

4. Method according to any of the preceding claims, comprising obtaining the at least one parameter before the vehicle is coupled to a charging station.

5. Method according to claim 1, comprising:

- Determining a power budget for at least one electric vehicle at the charger, based on the estimated charging characteristics;

- Charging the electric vehicle according to the power budget.

6. Method according to claim 5, wherein the budget comprises a maximum current, voltage, power, or charging time or other parameter based on the budget to the vehicle.

7. Method according to any of the preceding claims, wherein providing the estimated charge characteristics to the vehicle or its user, comprises displaying and/or

communicating the characteristics.

8. Method according to any of the preceding claims, comprising:

- Selecting information or a specific message based on the estimation of the

charging characteristics from a resource;

- Providing the information or a specific message to the vehicle or its user.

9. Method according to any of the preceding claims, comprising:

- Estimating the charging characteristics based on at least one additional parameter obtained from another source than the electric vehicle.

10. Method according to claim 9, wherein the at least one additional parameter is based on estimated charging characteristics of another vehicle to be charged at the same charger.

11. Method according to claim 10, comprising:

- Re-estimating the charging characteristics of the vehicle when the other vehicle connects to the charger;

- Updating the charge settings between the charger and the first vehicle.

12. Method according any of the preceding claims, comprising:

- Re-estimating the charging characteristics in response to a change of a parameter related to said battery or a change of a parameter related to the energy source or a change of any other external parameter;

- Updating the charge settings between the charger and a vehicle coupled to a port.

13. Method according to any of the preceding claims, compri - Obtaining user preferences from a user via a interface of the charger or vehicle;

- Determining a power budget that satisfies the user preferences within the

boundaries of the estimation;

- Setting the charger according to the power budget;

- Charging the electric vehicle with the charger according to the power budget.

14. Method according to claim 1 wherein estimating the charge characteristics based on at least the obtained parameters comprises:

a) Providing the initial parameters to a neural network;

b) Computing a first estimation of the charging characteristics based on the initial parameters;

c) Compare the estimated charging characteristics with actual charge

characteristics and calculate an error, defined by the difference between the estimated and the actual charging characteristics;

d) Providing the resulting error to the neural network for training.

15. Charger, configured for use in a method according to one of the preceding claims, comprising:

means for obtaining parameters from an electric vehicle;

means for obtaining the characteristics of the actual charging of the electric vehicle;

means for estimating the charging characteristics based on at least the obtained parameters and using the characteristics of the actual charging to improve the estimation of the charging characteristics;

at least one first power exchange port, for exchanging power with a power supply;

at least one second power exchange port, for exchanging electric power with a battery of a vehicle to be charged;

at least one power converter for converting the power between the at least one first power exchange port and the at least one second power exchange port.

16. Charger for an electric vehicle according to claim 15, wherein:

- the first power exchange port has a power rating which is lower than the sum of the power ratings of the second power exchange ports; - the method according to claim lis utilized to influence the power flow from the first power exchange port to the at least two second power exchange ports.

17. Charging system for an electric vehicle according to claim 15 or 16, comprising a user interface wherein the method according to claim 1 is utilized to display information about the progress of charging.

18. Charging system for an electric vehicle according to any of claims 15-17, comprising a user interface wherein the method according to claim 1 is utilized to display a message or specific information which is selected based on the estimation.

19. Charging system according to any of claims 15-18, comprising communication means for exchanging data with a neural network.

20. Charging system according to any of claims 15-19, for use in a method according to claim 2, comprising the database or means for connecting to the database.