Provision of balancing power by means of multiple pooled electric vehicles

Abstract

The invention relates to a method for providing a balancing power by means of a pool comprising multiple electric vehicles (EV), comprising the following steps: determining respective charging flexibilities of charging pairs ([EVSE / EV]-1 - [EVSE / EV]-n) of electric vehicles (EV) and charging points (EVSE) allocated to the pool; establishing a grid-frequency-dependent primary control power curve (PPRL) for each charging pair ([EVSE / EV]-1 to [EVSE / EV]-n) on the basis of its respective charging flexibility; transmitting the primary control power curves (PPRL) of the charging pairs ([EVSE / EV]-1 - [EVSE / EV]-n) to associated charging controllers which are designed to control the charging of the associated electric vehicle (EV) at least depending on a current grid frequency (f) according to its primary control power curve (PPRL); establishing, using the pooling instance (PI), an overall primary control power curve (PPRL |p) corresponding to an aggregation of primary control power curves (PPRL) of individual charge pairs ([EVSE / EV]-1 - [EVSE / EV]-n); and calculating a balancing power provided by the electric vehicles (EV) of the pool on the basis of the overall primary control power curve (PPRL |p).

Description

[0001] 23-3563 PIF 1 / 31 Providing Control Power Using Several Pooled Electric Vehicles The invention relates to a method for providing control power using a pool comprising several electric vehicles. The invention also relates to a computer program product comprising code which, when executed on a data processing device, performs the method. The invention further relates to a control system comprising several charging pairs consisting of individual electric vehicles and charging points, the charging processes of which can be controlled by a charging controller, and a central pooling instance linked to the respective charging controllers via data technology, wherein the control system is configured to perform the method. It is known from the prior art that primary control power (PCP) can be provided by individual charging pairs, each comprising a charging station and an electric vehicle coupled thereto.(also referred to internationally as "Frequency Containment Reserve", FCR). It is also known that candidates for providing primary control reserve are selected from a group (a "pool") of such charging pairs. The provision of primary control reserve PPRL(f) is controlled locally by a local measurement of the grid frequency f, e.g., according to the relationship PPRL(f) = P{PQ} · ∆f / 200 mHz, where P{PQ} corresponds to the maximum available (prequalified) power and ∆f corresponds to the deviation of the measured grid frequency f from a target grid frequency f0. ∆f can be limited by a maximum controllable frequency deviation, above which P{PQ} is not increased further. For example, the target grid frequency f0 can be 50 Hz or 60 Hz, depending on the specific circumstances.which target grid frequency f0 the electricity supply network receiving the primary control reserve uses. One problem with using the traction batteries of electric vehicles to provide primary control reserve is that, due to the minimum bid size of 5 MW for the control reserve, a large number of vehicles, charging (being charged and / or discharged) at a typical charging power P of approximately ± 10 kW, must be pooled to reach the minimum bid size. In addition, failed electric vehicles must be compensated for by other electric vehicles. This is typically done by aggregating individual vehicles into a pool by a pooling instance to access the primary control reserve of the pooled electric vehicles on an ancillary services market via market access, e.g., via an energy market aggregator.to market the energy. The energy market aggregator is obligated to monitor and provide proof of the primary control reserve provided to the grid operator. This means that the energy market aggregator must transmit the currently provided control reserve to the grid operator (e.g., to the grid operator's control center, for example, via the IEC 104 protocol) for monitoring purposes at specific intervals (latency), e.g., less than 10 seconds. If necessary, the performance data can be provided subsequently to verify the primary control reserve provided by the pool or the individual charging pairs. For conventional technical units that provide primary control reserve (e.g., generators at power plants or large battery storage systems),The primary control reserve of the plant is continuously measured and sent to the grid operator via a digital channel through the energy market aggregator – thus monitoring each plant individually. With many thousands of charging pairs aggregated into a pool (e.g., in the sense of a virtual power plant), this approach leads to very large data volumes and also to the problem that the measured values ​​from different charging pairs often arrive asynchronously. Summing the measured power of the individual charging pairs into a total power, which is then reported to the grid operator, means that the aforementioned latency criterion of less than 10 seconds for monitoring the total primary control reserve provided cannot be met. Furthermore, due to the asynchronous nature of the measured values, there is also the problem of correct aggregation, i.e., the problem of how to handle measured values ​​from different charging pairs that lead to different,but are measured at closely spaced times. US 10663932 B2 discloses methods, devices, and systems for charging and discharging an energy storage device connected to a power distribution system. In an exemplary embodiment, a controller monitors electrical properties of a power distribution system and provides an output to a bidirectional charger, causing the charger to charge or discharge an energy storage device (e.g., a battery in a plug-in hybrid electric vehicle, PHEV). The controller can contribute to stabilizing the power distribution system by increasing the charging rate when there is a current surplus in the power distribution system (e.g., when the frequency of an AC grid exceeds an average value) or by discharging current from the energy storage device to stabilize the grid.when there is a power shortage in the electricity distribution system (e.g., when the frequency of an AC grid is below an average value). WO 2020 / 120457 A2 discloses a device for charging and discharging a drive energy storage system of a hybrid or electric vehicle. The device comprises a frequency measurement module configured to measure the grid frequency of a power supply network; a control module configured to control, based on the measured grid frequency, charging the drive energy storage system from the power supply network or discharging the drive energy storage system into the power supply network to provide ancillary services; and a first communication module configured toto communicate as a master with the hybrid or electric vehicle to provide balancing power. The method describes the bundling of several vehicles into a vehicle pool and the activation of individual vehicles in the pool to provide balancing power. US 2008 / 0052145 A1 discloses systems and methods for a power aggregation system described. In one implementation, a service establishes individual internet connections to numerous electrical resources that are temporarily connected to the power grid, such as electric vehicles. The service optimizes power flows to meet the requirements of each resource and resource owner, while aggregating flows across numerous resources to meet the demands of the power grid. The service can utilize large quantities of electric vehicle batteries as new,Bringing a dynamically aggregated electricity resource online for the power grid. Owners of electric vehicles can participate in an electricity trading economy, regardless of where they connect to the power grid. DE 102009050042 A1 discloses a charging station for electric vehicles. Grid stabilization is achieved by having a grid frequency measuring device for recording a grid frequency and for detecting a deviation of the grid frequency from a target frequency, and by having a load control device in operative connection with the grid frequency device such that, in the event of a detected deviation of the grid frequency from the target frequency, the load control device regulates the electrical power supplied to an electric vehicle from the charging station. DE 102022127911 A1 discloses a method for providing primary control power for an energy market.with the following steps: Predicting a large number of mobility behaviors of a large number of electrically powered vehicles using an electronic computing device; Determining a potentially available primary control power depending on the prediction using the electronic computing device; Generating a control schedule for regulating the energy supply of a respective vehicle using the electronic computing device; When the primary control power is requested by the energy market,The process involves transmitting an individual control to the respective local control units of the individual motor vehicles, depending on the control schedule, by means of a communication device, and executing the individual control of each motor vehicle by means of the local control unit. A computer program product and a control system are also disclosed. DE 102017209801 A1 relates to a method for operating a large number of technical units as a network on an electrical distribution network and provides that a central control device receives flexibility data from each technical unit, by means of which the unit specifies a power range within which its electrical power may vary or is expected to vary. From the flexibility data of each unit, an individual control system is generated for each unit, depending on a predetermined optimization goal of the network.A non-binding incentive function is determined and provided to the respective unit. From each unit, a corresponding schedule is then received in response to the individual incentive function. This schedule describes the temporal progression of the power exchange planned according to a local optimization goal of the unit. A total schedule for the network is then generated from the schedules of the units. DE 102022129783 A1 discloses a system for relieving the load on a local power grid and a higher-level power grid, which are electrically coupled to each other at a grid feed-in point, comprising at least one charging unit electrically connectable to the local power grid for electrically charging an electrically operated vehicle, wherein the charging unit has an energy storage device, a computing unit with a processor, which is coupled to the charging unit via a signal connection.a data storage device and a receiving unit for receiving network data, which represents at least a demand for electrical power from the local power grid and a network frequency of the higher-level power grid, wherein the processing unit is configured to determine, based on the network data, an electrical power to be supplied to the higher-level power grid and / or the local power grid, and wherein the processing unit is configured to control the charging and / or discharging of the energy storage device in such a way that the determined electrical power is supplied to the higher-level power grid and / or the local power grid. It is an object of the present invention to at least partially overcome the disadvantages of the prior art in providing primary control power from a pool of electric vehicles and, in particular, to benefit a market participant in an energy market, especially a network operator and / or energy aggregator.to provide information about the total primary control power provided by a pool of electric vehicles with a particularly low latency. This task is solved according to the features of the independent claims. Preferred embodiments can be found in particular in the dependent claims. The task is solved by a method for providing control power for an energy market by means of a pool comprising several electric vehicles, with the steps: − Determining the respective charging flexibilities of the charging pairs of electric vehicles and charging points assigned to the pool; − Establishing a grid frequency-dependent primary control power curve for each charging pair based on its respective charging flexibility; − Transmitting the primary control power curves of the charging pairs to respective charging controllers, which are configured to control the charging of the respective associated electric vehicles depending on the current grid frequency according to its primary control power curve; − Establishing,through the pooling instance, a total PRL power curve, which corresponds to an aggregation of PRL power curves of individual charging pairs, − Calculation of the control power provided by the electric vehicles (EVs) of the pool based on the total PRL power curve (PPRL |p). 23-3563 PIF 6 / 31 This method has the advantage that the control power provided by the pool can be calculated with a latency of well under 10 seconds, e.g., almost in real time with approximately 1 second. This takes advantage of the fact that the individual PRL power curves or functions of the charging pairs assigned to or currently participating in the pool can be quickly aggregated into the total PRL power curve, and the total PRL power curve accurately reflects the provision of control power by the entire pool as a function of the grid frequency. This assumes that the grid frequency is the same for all electric vehicles connected to the pool.This is generally the case, for example, when the charging points are connected to the same grid. The provided control power can therefore be calculated using a single function, namely the overall PRL power curve, which is typically achieved in less than a second. Furthermore, the provided control power calculated via the overall PRL power curve corresponds to the actual provided control power with high accuracy and low tolerance. The overall PRL power curve can even be predicted with high accuracy for times in the immediate future. This is based on the fact that the charging pairs follow the control signals specified by their individual PRL power curves with a very high probability, and individual failures, with a typical large number of electric vehicles participating in the pool (typically more than a thousand or tens of thousands), are below the tolerance threshold. If vehicles fail and / or join,This can be registered promptly through continuous monitoring of the pool, and the overall PRL power curve can be recalculated. The sequence of the procedural steps is not limited to the sequence above. Rather, at least two of these procedural steps can also be carried out in a different sequence than specified or even simultaneously. It is a further development that the control power is a primary control power. However, it can also be a secondary control power, etc. 23-3563 PIF 7 / 31 The control power, in particular primary control power, can be allocated to the electricity supply network to which the charging points of the charging pairs are connected.or made available to an associated grid operator. The balancing power can be offered directly to the grid operator by the pooling entity or indirectly via another market participant, such as an energy market aggregator, acting as an intermediary. Generally, the pool's total flexibility can be offered in advance to a market participant, e.g., the grid operator or the energy market aggregator, for use as balancing power, particularly primary control reserve, for a specific time period. Upon acceptance of the offer in whole or in part, a corresponding overall primary control reserve power curve can be transmitted to the market participant in advance. The market participant can then call upon the balancing power as needed. Therefore, a training course can generally consist of two phases: first, the offer phase, and then the control phase.which reacts to a request for balancing power. The offer phase can and typically is carried out much earlier than the control phase. In particular, it is important that the balancing power only needs to be actually provided by the pooling instance if the grid operator requires it. To cover the required balancing power, a suitable overall balancing power curve can be established using the pooling instance, which may not include all charging pairs in the pool, but only a subset of them. This can be implemented, for example, by having a (demand) signal sent directly from the grid operator or indirectly via an intermediary such as an energy market aggregator to the pooling instance. The pooling instance, in turn, sends appropriate activation signals to selected, or possibly all, charging pairs to aggregate the desired overall balancing power curve.Charging pairs of the pool or their charging controllers are selected, whereupon these activated charging controllers modify the charging power or charging behavior of the associated charging pair based on the respective PRL power curves. The creation of the overall PRL power curve, which corresponds to an aggregation of the PRL power curves of individual charging pairs, can therefore also be described as "selective" or "demand-based" creation. If not all charging pairs are activated or selected to provide control power, the decision as to which charging pairs are activated and which are deactivated can be made based on one or more criteria, for example, the size of the available control power, the expected connection duration, the battery state of charge, the setting whether battery discharge is also permitted, contractually agreed priority rules, etc. If the demand for control power changes,Depending on the situation, certain charging pairs can be additionally activated or deactivated to provide control power. When charging pairs are deactivated, their charging controllers adjust the charging power again without considering the respective primary control power curves. The above procedure for providing control power, especially primary control power, is therefore only necessary if the control power is drawn from a previously created pool of charging pairs. Consequently, it is a further development that the pooling instance only activates as many charging pairs as are necessary to fulfill the requested control power. This is particularly advantageous if not all of the pool's offered flexibility or control power is accepted and / or actually used, but only a portion of it. The electric vehicle, for example, could be a plug-in hybrid vehicle (PHEV).or a fully electric vehicle, e.g., a battery electric vehicle (BEV). In this case, at least one traction battery of the electric vehicle can be used as the energy storage device to be charged. The electric vehicle can be, for example, a passenger car, truck, bus, motorcycle, etc. In this context, "charging" can refer to both charging and discharging. Determining charging flexibility is generally known and is based on the existence of a charging plan for the connected electric vehicle, which has been established taking into account charging constraints. The charging plan can, for example, include a desired target state of charge at a desired or estimated disconnection time. The charging plan depends on the charging constraints of the charging point (e.g., its charging power(s)), the charging constraints of the electric vehicle (e.g., its charging power(s), charging mode such as "instant charging" or "delayed charging"),etc.) and, if applicable, charging boundary conditions of the local power grid (e.g., phases with particularly inexpensive charging options) and the higher-level power grid (e.g., a maximum charging capacity available). The charging plan can be optimized for cost and / or environmental impact. The charging plan can be generated, for example, by the vehicle, the charging point, or another entity such as a home energy management system (HEMS). Once the charging plan is generated, it is managed by the energy supplier to which the charging point is connected. The charging power profile assigned to the charging plan is also referred to as the "baseline." If the target state of charge can only be achieved by setting the maximum charging power, there is no way to deviate from the charging plan; it is therefore not flexibly modifiable. In other words, this charging plan lacks flexibility. However, most charging plans offer some degree of flexibility.This means that the charging schedule can be temporarily deviated from, as long as the target state of charge or a predefined range around the target state of charge is still reached. For example, if a charging schedule specifies "instant charging" to a target state of charge of 80% SoC and the target state of charge is reached after two hours, but the electric vehicle is not scheduled to be disconnected until six hours later, the charging schedule results in a four-hour rest phase for the traction battery, during which it is maintained at the previously reached target state of charge of 80% SoC. The "baseline" then corresponds to the maximum charging power during instant charging and to zero during the subsequent rest phase. The rest phase can be used to temporarily charge and discharge the traction battery, as long as the target state of charge is at least approximately reached upon disconnection. Other charging scenarios include "late charging,"where a resting phase exists after connecting the electric vehicle, as the electric vehicle is charged at the latest possible time with the then maximum charging power. However, most charging plans in practice lie between the two extremes of "immediate charging" and "late charging" and usually show longer charging times with lower power than the maximum power, especially—as already indicated above—optimized for specific charging goals, e.g., particularly cost-effective charging and / or particularly environmentally friendly charging, possibly in conjunction with other consumers and, if present, intermediate storage and / or energy generation facilities of a property, etc. Charging flexibilities are also generally present here. 23-3563 PIF 10 / 31 These "charging flexibilities" are typically subject to certain boundary conditions, for example, the maximum permissible deviation from the baseline (e.g., of ± 10% SoC).a maximum number of discharge cycles within the coupling period, the decision as to whether discharging is possible or permitted, and ensuring that charge levels do not exceed or fall below the permitted limits, etc. In this case, the flexibility is used to provide primary control reserve. If the grid frequency is too low, the vehicle feeds energy back into the grid in addition to the baseline; if the grid frequency is too high, it feeds energy in, in addition to the baseline, each within the limits of the permitted flexibility. If the baseline is zero, this corresponds to net energy feed-in when the grid frequency is too low and net energy feed-in when the grid frequency is too high. If, however, the vehicle is currently being charged according to the baseline, it may be charged with a reduced amount of energy or at a slower rate compared to the baseline if the grid frequency is too low.In the event of an excessively high grid frequency, charging occurs with an increased amount of energy or at a faster rate. This utilizes the fact that the grid frequency increases when energy is fed into the grid and decreases when energy is drawn. Establishing a baseline in which the electric vehicle is charged for a longer period, particularly continuously, at a lower power than the maximum power is especially helpful for providing primary control reserve even for electric vehicles that do not allow (physical) discharge. It can be advantageous for the charging power according to the baseline to be absolutely greater than the absolute maximum permissible value of a discharge for primary control reserve, so that the total power actually flowing into the electric vehicle remains positive. The charging flexibilities of the respective charging pairs in the pool can be determined in a generally known manner.In particular, they must also be updated, e.g., taking into account already utilized charging flexibilities, changes to charging boundary conditions, etc. The charging flexibilities can, for example, be determined by the instance that also created the charging plan. The central pooling instance is connected to the charging pairs of the pool via data, either directly or indirectly via a HEMS or similar device. This allows the charging flexibilities to be transferred to the pooling instance. The grid frequency-dependent power curves for each of the charging pairs are generated by the pooling instance.Specifically, this is based on their respective charging flexibilities. The pooling instance can be, for example, an IT system of a vehicle manufacturer or fleet operator. A grid frequency-dependent PRL power curve for a charging pair corresponds to a relationship, representable as a curve, between the grid frequency and the control power to be provided by this charging pair. The individual PRL power curves are transmitted to the associated charging controllers of the respective charging pairs. The charging controllers are configured to control the charging (charging or discharging) of the electric vehicle according to its PRL power curve, depending on the current grid frequency. The charging controllers control the charging process autonomously, i.e., without further intervention from the pooling instance. Controlling the charging according to the PRL power curve can be implemented, for example, as follows:The charging control system superimposes the primary control power to be provided, derived from the PRL power curve and the grid frequency, onto the baseline and carries out the charging process of the traction battery(ies) accordingly. Furthermore, the pooling instance aggregates the overall PRL power curve from the (particularly selected) individual PRL power curves, e.g., by simple addition. The overall PRL power curve represents the primary control power to be provided by the entire vehicle pool as a function of the current grid frequency. Since it can be assumed that the charging pairs follow the control signals of their charging controllers, as specified by their individual PRL power curves, with a very high probability, and that individual failures, with a typical high number of electric vehicles participating in the pool, are below the tolerance threshold,The primary control power to be provided, as represented by the overall PRL power curve, can be equated with the primary control power actually provided with sufficiently high accuracy. Consequently, it is advantageously possible to approximate the control power actually provided by the pool with high accuracy simply by calculating the overall PRL power curve using the known current grid frequency. Since this is a simple calculation, the control power can be (re)calculated at very short intervals. 23-3563 PIF 12 / 31 One embodiment involves setting up the individual PRL power curves locally at the charging pairs (i.e., by the charging pairs themselves, the associated charging controllers, or local energy management systems that take the respective charging pair into account) based on their respective charging flexibilities.The data is reported or transmitted to the pooling instance, which then uses aggregation to construct the overall PRL power curve. This has the advantage that the pooling instance is relieved of the computational overhead of calculating the individual (local) PRL power curves. The individual PRL power curves can then also be transmitted locally (i.e., without involving the pooling instance) to the associated charging controllers if the charging controllers have not already generated them themselves. Constructing the overall PRL power curve by aggregation, in particular addition, from the individual PRL power curves is advantageously very easy to implement. In a further training module, the charging flexibilities of the charging pairs assigned to the pool can be reported or transmitted to the pooling instance, which then creates an overall charging flexibility and offers it to market participants for use as balancing power. This is one possible implementation.The charging flexibilities of the charging pairs assigned to the pool are reported to the pooling instance, which then generates the individual PRL power curves for each charging pair and transmits them to the corresponding charging controllers. This has the advantage that the local instances are relieved of the task of calculating the associated PRL power curves, and the pooling instance can also coordinate the individual PRL power curves if necessary. The pooling instance can then easily calculate or generate the overall PRL power curve from the individual PRL power curves. Here, too, an overall charging flexibility can be calculated from the individual charging flexibilities and offered to market participants for use as balancing power. It is a configuration in which the charging flexibilities of the charging pairs assigned to the pool are reported or transmitted to the pooling instance.The pooling instance then constructs the overall PRL power curve (e.g., first by calculating the overall charging flexibility from the individual charging flexibilities and then calculating the overall PRL power curve from that) and the pooling instance uses the overall PRL power curve to construct the individual PRL power curves for each charging pair by disaggregation and transmits them to the corresponding charging controllers. 23-3563 PIF 13 / 31 The transmission of the charging flexibilities of the individual charging pairs to the pooling instance can include transmitting not the fully determined charging flexibilities, but (only) the data required to issue the charging flexibilities, e.g., the respective charging requests, charging boundary conditions, and, if applicable, baselines, etc. The pooling instance can then determine the respective charging flexibilities as such. This "central" determination of the respective charging flexibilities and, if applicable, also baselines, achieves the advantage thatthat updating the calculation logic is simpler. In contrast, determining individual charging flexibilities "decentrally" is advantageous for reducing the computational effort in the pooling instance. Generally, the baselines of individual charging pairs can also be determined centrally or decentrally. In the former case, the baselines can then be transmitted to the respective charging controllers, for example, and in the latter case, to the pooling instance. The pooling instance can then, in a further development, determine an overall baseline for the pool, for example, through aggregation, and offer it on the market for purposes other than providing control power. One configuration involves the charging points being components of local power grids that are connected via an energy meter to a power grid as the receiver of the control power, in particular primary control power.are connected. A local power grid can, for example, be the power grid of a property that is connected to the electricity supply network via the energy meter. Additionally, there can be a separate energy meter for each charging point. A property can be, for example, a commercial or private property, e.g., a charging park, a parking lot with multiple charging stations, a residential building, especially a single-family home, etc. A property can have one or more charging points. The charging point can be a charging station, a wall-mounted charging station (so-called "wallbox"), an inductive charging pad, etc. It is a configuration in which the grid frequency is monitored, in particular measured, via the energy meters connected to the electricity supply networks.This allows the grid frequency for providing the control power to be determined advantageously in a particularly timely manner. The tracked grid frequency is then transmitted to at least 23-3563 PIF 14 / 31 a charging control system of the local power grid, by means of which the charging of at least one charging pair can be controlled. The charging control system then controls the charging of the at least one charging pair it can control using the respective PRL power curve, e.g. by superimposing it with the respective baseline. It is an embodiment in which a single PRL power curve for a charging pair is a particularly continuous function PPRL (f) of the requested control power PPRL with the properties − for f ≤ f, PRL,min : P PRL (f) = - P PRL,min with f PRL,min a predetermined lower frequency threshold below a target mains frequency f0 and P PRL,min> 0 represents the absolute maximum control power that can be discharged by the electric vehicle; − for fPRL,min < f < fPRL,max: KPRL(f) with fPRL,max a predefined upper frequency threshold above the target grid frequency f0 and KPRL(f) a curve segment that increases monotonically with higher grid frequencies f; − for f ≥ fPRL,max: PPRL (f) = PPRL,max with PPRL,max corresponding to the maximum control power that can be charged by the electric vehicle. Here, PPRL,min ≤ PPRL (f) ≤ PPRL,max. In particular, KPRL(f) follows continuously from the frequency thresholds fPRL,min and fPRL,max. Specifically, KPRL(f) is strictly monotonically increasing. In particular, KPRL(f0) = 0. In general, the deviations from the target network frequency f0, namely ∆fPRL,min = f0 - fPRL,min and ∆fPRL,max = fPRL,max - f0, can be equal or different in magnitude. PPRL,min and PPRL,max can also be equal or different in magnitude.The PRL power curves P(f) can be the same or different for at least two charging pairs. For example, fPRL,min, fPRL,max, fmin, PPRL,min and / or P can be different. PRL,max as well as the curve shape of K PRL (f) differ. It is a configuration that the curve segment K PRL (f) is at least sectionally li- near. This facilitates the aggregation of the individual PRL performance curves into the overall PRL performance curve. It is a further development that the curve section K PRL (f) corresponds to a linear function with a constant slope. It is a further development that the curve segment K PRL(f) is structured like a polygon. In particular, non-linear curves can also be approximated in this way while still using simple aggregation. If the curve segment KPRL(f) corresponds to a linear function, this can also be expressed in a further development as − - s · PPRL,min with s = ∆fPRL,min / fPRL,min for fPRL,min < f < f0; − 0 for f = f0; − + s · PPRL,max with s = ∆fPRL,max / fPRL,max for f0 < f < fPRL,max. It is an embodiment in which the charging control carries out a charging process of the electric vehicle of the charging pair based on a charging curve which corresponds to a superposition of an associated baseline with the PRL power curve. This is advantageously particularly easy to implement. Then, feeding back control power from the electric vehicle corresponds to a reduction of the baseline, while receiving control power into the electric vehicle corresponds to an increase of the baseline.The fact that the charging curve corresponds to a superposition of an associated baseline with the PRL power curve means, in particular, that at a specific time, the baseline value applicable at that time and the value calculated at that time from the grid frequency using the PRL power curve or function are aggregated, specifically added. If a baseline power PBas(t) is thus considered, the relationship Pges(f) = PBas + PPRL(f) can be assumed for the total charging power Pges at a specific time t. One implementation involves setting up the overall PRL power curve using the central pooling instance such that the overall PRL power curve flattens out as it approaches at least one of the frequency thresholds, relative to the corresponding predetermined minimum and / or maximum primary control power.This achieves the advantage of stabilizing the pool's control, as small jumps can occur in the gradient of the overall PRL power curve, and this design makes the overall PRL power curve "smoother." Constructing the overall PRL power curve in this way can be achieved, for example, by having the pooling instance select or selectively activate those individual PRL power curves for aggregating the overall PRL power curve when not all individual curves are needed to provide the requested control power. This ensures that the overall PRL power curve exhibits this desired behavior. 23-3563 PIF 16 / 31 It is a configuration such that when the network frequency f lies within a dead zone of f0 - fTmin ≤ f ≤ f0 + fTmax with fTmax > 0 around the target frequency f0, no primary control power is called upon from the pool.This can be particularly advantageous if other participants in the control loop are also active, in order to avoid oversteering. If a deadband is taken into account, the PRL performance curve can, for example, assume the value zero there or form a plateau with the value zero. In particular, the curve segment KPRL(f) can exhibit such a plateau in the deadband and otherwise, at least section by section, a strictly monotonically increasing curve. It is a configuration in which the overall PRL performance curve is recalculated at regular intervals. This advantageously keeps the difference between calculated and actually delivered total primary control power small.Reconstructing the overall primary control power curve (PCP) can, for example, involve deleting the PCP power curves of temporarily disconnected electric vehicles, adding PCP power curves of newly added charging pairs, and / or updating existing PCP power curves. This reconstruction or updating can be implemented quickly, e.g., approximately every second. An additional or alternative configuration is to reconstruct or update the overall PCP power curve triggered by an event. This also advantageously minimizes the difference between calculated and actual total primary control power. A further development is that an event could be the connection and / or disconnection of an electric vehicle to or from a charging point, local interventions by HEMS, a customer switching to instant charging, or the presence of competing interventions from other control options, e.g.,in accordance with Section 14a of the Energy Industry Act, etc. It is a configuration whereby the calculation of the balancing power based on the overall PRL power curve is carried out by the pooling instance, and the result of the calculation is reported or transmitted to at least one relevant market participant, e.g., for monitoring and / or reporting purposes. The relevant market participant comprises at least one market participant with whom it has been agreed to report the balancing power (see 23-3563 PIF 17 / 31). In particular, the pooling instance or its operator may be obligated to report the calculated balancing power to the relevant market participant. It is a configuration whereby the calculation of the balancing power based on the overall PRL power curve is carried out in such a way that the overall PRL power curve established by the pooling instance is reported or transmitted to at least one relevant market participant.This at least one corresponding market participant can then calculate the likely balancing power provided. If the overall balancing power curve is modified, it is transmitted, in particular, to the at least one corresponding market participant. The task can also be solved by a computer program product comprising code which, when executed on a data processing device, performs the procedure as described above. The computer program product can be designed analogously to the procedure, and vice versa, and offers the same advantages.The problem is further solved by a control system comprising several charging pairs consisting of individual electric vehicles and charging points, the charging processes of which can be controlled by a charging controller, and a central pooling instance linked to the respective charging controllers via data technology, wherein the control system is configured to carry out the method as described above. The control system can be designed analogously to the method and the computer program product, and vice versa, and has the same advantages. The properties, features, and advantages of this invention described above, as well as the manner in which they are achieved, become clearer and more readily understandable in connection with the following schematic description of an embodiment, which is explained in more detail in conjunction with the drawings. Fig.Figure 1 shows a property connected to an energy supply network, which has a charging point to which an electric vehicle is connected; Figure 2 shows a possible baseline and charging flexibility plotted against time; Figure 3 shows a sketch of a system with market participants, an external cooling instance, and several charging pairs; Figure 4 shows a possible PRL power curve plotted against a grid frequency for a charging pair; Figure 5 shows a possible overall PRL power curve plotted against a grid frequency; and Figure 6 shows a possible sequence of the procedure. Figure 1 shows a property LS connected to an energy supply network EVN, which has a charging point EVSE to which an electric vehicle EV is connected, here via a charging cable.The electrical energy or power exchanged between property LS and the EVN energy supply network can be sustainably monitored using an energy meter, e.g., a "smart meter" SM installed at the connection point and / or an optional energy meter EM belonging to property LS. The smart meter SM and the optional energy meter EM can also measure the current grid frequency f and transmit it via a data network value. Property LS, e.g., a single-family home, has electrical consumers (not shown) in addition to the charging point EVSE and may optionally also have electrical energy generation equipment (not shown) such as a photovoltaic system, a wind turbine, etc., as well as at least one stationary energy storage system (not shown). The charging point EVSE and the electric vehicle EV form a charging pair with associated charging boundary conditions.These charging boundary conditions can include, for example, maximum charging capacities of the charging point (EVSE) and the electric vehicle (EV), as well as the possibility of bidirectional charging, i.e., not only charging but also discharging. Together with a charging request, e.g., regarding a specified or predicted departure time, a target state of charge, or target optimization (e.g., for particularly cost-effective and / or environmentally friendly charging), etc., a future-oriented charging plan can be created with a baseline representing the required charging capacity. This charging plan can typically take into account variable electricity prices, the electricity mix, feed-in tariffs, and power limitations of the energy supply network (EVN).23-3563 PIF 19 / 31 In particular, if the property LS is equipped with an electrical power generation unit and / or a stationary energy storage system, the flow of electricity within the property LS, and, if an electric vehicle EV is connected, also to the vehicle, can optionally be controlled by an energy management system, e.g., a home energy management system (HEMS). The charging plan can then also take into account, for example, the electrical energy generated by the property and / or the possibility of intermediate storage in order to achieve optimized charging. The charging plan for the electric vehicle EV can be generated, for example, by the electric vehicle EV, the charging point EVSE, or the energy management system. A charging controller (not shown) that implements the charging plan by outputting corresponding control signals can be integrated into the electric vehicle EV, the charging point EVSE, or the energy management system.If the instance generating the charging plan does not include the charging controller, the charging plan can be appropriately transferred from this instance to the charging controller. Fig. 2 shows a possible baseline P as a plot of electrical power P in kW against time t. bas , P_bas, and a charging flexibility P used to provide control performance PRL , P_PRL. The baseline P bas Here, P has been assumed to be constant over the shown time period, using P as an example. bas> 0, meaning that the electric vehicle EV is charged with constant power according to the charging schedule over the specified period, e.g., from the energy supply network EVN, the local electrical energy generation facility, and / or from the stationary energy storage system. For the charging pair EV, EVSE assumed here, with bidirectional charging capability, it is assumed that charging flexibility can be provided over the specified period in the form of a charging cycle (P > P). bas ) or a discharge (P < P bas ) of the drive battery(ies) of the electric vehicle EV. Shown is the actual course of the control power P provided within the framework of charging flexibility. PRL While the charging flexibility is predictable, the control power P is determined PRLfrom the current grid voltage f or its deviation from the setpoint f0. This corresponds to an "a posteriori" view, such as that which can be generated, for example, by a reporting system based on measurement data (not substitute values). 23-3563 PIF 20 / 31 Fig. 3 shows a sketch of a system with market participants EMA, ENV, a pooling instance PI and several charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n, where n can typically be in the range of 1000...100,000, but is not limited to this range. The n charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n, which are data-linked to the pooling instance PI, transmit their respective baselines and flexibilities to the pooling instance PI. Alternatively, the charging requests and charging boundary conditions, etc., of the charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n are transmitted to the pooling instance PI, which then determines the respective baselines and flexibilities. The pooling instance PI can then, for example,The baselines are aggregated and offered to energy market participants for servicing, e.g., an energy market aggregator (EMA). The aggregated flexibilities are also offered to energy market participants, in this case as a contribution to primary control reserve. How much of the potentially available aggregated flexibility is then used as primary control reserve (P)? PRL (f) is provided, depends on the occurring grid frequency fluctuations and, for example, on a call-up by the relevant market participant, e.g., the energy market aggregator EMA. If the offered aggregated flexibility is accepted (in whole or in part) by a market participant, the associated primary control reserve P PRL(f) by controlling the charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n by means of a local measurement of the grid frequency f, provided locally on a pro rata basis. The pooling instance PI reports the aggregated primary control reserve, in particular to the receiving market participant, e.g., the operator of the energy supply network ENV and / or the energy market aggregator EMA. The latter may be the case, for example, if the energy market aggregator EMA is obligated to monitor and provide proof of the primary control reserve provided to the network operator ENV. This means that the energy market aggregator EMA must transmit the currently provided control reserve to the network operator ENV (e.g., to the control center of the network operator ENV, for example, via the IEC-104 protocol) for monitoring purposes at certain intervals, e.g., less than 10 seconds.To maintain this latency, the pooling instance PI generates individual grid frequency-dependent primary control reserve (PCR) power curves for all existing charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n that are authorized for use of the PRR, based on their respective charging flexibilities, and transmits these to the respective charging controllers. The charging controllers then control the charging processes of the respective charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n, also taking into account the respective PRR power curves, by means of a local measurement of the grid frequency f. For example, the primary control reserve provided via the PRR power curve can be superimposed on a baseline. The pooling instance PI aggregates a grid frequency-dependent overall PRL power curve from the individual PRL power curves of the charging pairs [EVSE / EV]-1, ...., [EVSE / EV]-n, e.g. by adding the individual PRL power curves.The pooling instance PI then reports the value of the total primary control power curve as a function of the current grid frequency, representing the total primary control power provided. To verify the total primary control power actually provided by the pool of charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n, the power data can be provided subsequently, if required. Fig. 4 shows a possible charging curve P plotted against an electrical power P in kW and a grid frequency f in Hz for a charging pair [EVSE / EV]-1, ..., [EVSE / EV]-n. ges (f), P_total, for a specific time, for which P ges (f) = P Bas + P PRL (f) with P Bas the value of the baseline at this time and P PRL (f) of the individual PRL power curve. The course of the charging curve P ges (f) corresponds to the course of the PRL performance curve, but is modified by the value P Bas shifted along the P-axis. Is P Bas = 0, P applies ges (f) = PPRL (f). The PRL performance curve P PRL (f) is continuous, where − for f ≤ fPRL,min: PPRL (f) = - PPRL,min with fPRL,min , f_PRL,min, a given lower frequency threshold below a target grid frequency f0 and PPRL,min , P_PRL,min, a maximum control power dischargeable by the electric vehicle EV, where PPRL,min is assumed to be greater than zero. PPRL (f) = - PPRL,min < 0 expresses that, from the perspective of the electric vehicle EV, primary power energy is discharged to raise the grid frequency f to too low. − for fPRL,min < f < fPRL,max: KPRL(f), K_PRL (f), with fPRL,max , f_PRL,max, a given upper frequency threshold above the target grid frequency f0 and K PRL(f) a linear curve segment following the frequency thresholds, 23-3563 PIF 22 / 31 − for f ≥ fPRL,max: PPRL (f) = PPRL,max with PPRL,max , P_PRL,max, a maximum control power chargeable by the electric vehicle EV, which is greater than zero. PPRL (f) = PPRL,max > 0 expresses that, from the perspective of the electric vehicle EV, primary power energy is charged to reduce the excessively high grid frequency f. Furthermore, K applies here. PRL (f0) = 0. In general, the deviations from the target network frequency f0 can be expressed as ∆f PRL,min = f0- f PRL,min and ∆f PRL,max = f PRL,max - be the same or different in amount. Also, P PRL,min and P PRL,max may be the same or different in magnitude. Since the curve segment K PRL (f) corresponds to a linear function with constant slope, this can also be expressed here as − - s · PPRL,min with s = ∆fPRL,min / fPRL,min for fPRL,min < f < f0; − 0 for f = f0; − + s · P PRL,maxwith s = ∆f PRL,max / f PRL,max for f0 < f < f PRL,max can be expressed. If the current value Pbas of the baseline is taken into account, the current value of the charging curve is Pges − P ges = P bas - P PRL,min for f ≤ f PRL,min ; − Ptotal = Pbas - s · PPRL,min with s = ∆fPRL,min / fPRL,min for fPRL,min < f < f0; − Ptotal = Pbas for f = f0; − P ges = P bas + s · P PRL,max with s = ∆f PRL,max / f PRL,max for f0 < f < f PRL,max - P ges = P bas + P PRL,min for f ≥ f PRL,min Possible values ​​could be, for example, f0 = 50 Hz, f PRL,min = 49.8 Hz, f PRL,max = 50.2 Hz and P PRL,min and P PRL,max = 2 kW can be assumed. Fig. 5 shows a possible plot of electrical power P in kW against a mains frequency f in Hz, which is obtained by adding all baselines and the individual PRL power curve P. PRL(f) of all participating charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n. Since the individual PRL power curves PPRL (f) can have different values ​​for fPRL,min, fPRL,max, PPRL,min and PPRL,max, the overall charging curve Pges |p (f) is then not constant, but usually, as shown here, section by section or polygonal linear. It flattens out at least approximately asymptotically towards the values ​​PPRL,min and PPRL,max 23-3563 PIF 23 / 31, which can be specifically achieved using the pooling instance PI by appropriately designing the individual PRL power curves PPRL (f). In particular, Pbas |p = Σ Pbas of the individual charging pairs, PPRL,min |p = Σ PPRL,min of the individual charging pairs, PPRL,max |p = Σ PPRL,max of the individual charging pairs, fPRL,min |p = min {fPRL,min} of the individual charging pairs, and fPRL,max |p = max {fPRL,max} of the individual charging pairs. Fig. 6 shows a possible sequence of the procedure, particularly based on the system shown in Fig. 3.In step S1, baselines and charging flexibilities are determined for the charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n assigned to the pool, consisting of n electric vehicles EV and charging points EVSE. This can be done by the charging pairs [EVSE / EV]-1, ..., [EVSE / EV]-n or their charging controllers, or alternatively, after the baselines and charging flexibilities have been transmitted by the pooling instance PI. In step S2, the pooling instance PI uses the charging flexibilities to generate respective grid frequency-dependent PRL power curves for each charging pair [EVSE / EV]-1, ..., [EVSE / EV]-n. In step S3, the pooling instance PI transmits the PRL power curves to the charging controllers for each charging pair [EVSE / EV]-1, ..., [EVSE / EV]-n. The charging controllers output control signals to the associated charging pair [EVSE / EV]-1, ...., [EVSE / EV]-n to control the charging process.In this process, the charging controllers can, in particular, superimpose or adjust their respective baselines with the control powers resulting from the grid frequency-dependent primary control power curves. The charging controllers can also contain parameters for the reserve for the local control of the primary control power. In step S4, the pooling instance PI generates an overall primary control power curve, which corresponds to an aggregation of the primary control power curves for the individual charging pairs [EVSE / EV]-1 to [EVSE / EV]-n. 23-3563 PIF 24 / 31 In step S5, the maximum primary control power resulting from the overall primary control power curve is released to an intermediary or marketer, e.g., an energy market aggregator EMA, for retrieval. This can occur even before step S1 (e.g.,(one day prior) an offer phase may have been conducted, during which the flexibility or available control reserve was offered to a marketer and at least partially accepted by the marketer. Therefore, a training process can generally consist of two phases: first, the offer phase, and then the control phase, which reacts to a call for control reserve. If the maximum primary control reserve is accepted by the marketer, in step S6 the charging pairs [EVSE / EV]-1 to [EVSE / EV]-n available for providing the primary control reserve are released by the pooling instance PI for local provision of the primary control reserve and then provide their individual primary control reserve. Upon release of the provision of the individual primary control reserve services, in step S7 the pooling instance PI calculates the primary control reserve provided by the pool from the overall primary control reserve power curve and informs the responsible instance, e.g.The marketer is notified, who then forwards the information to the grid operator. This process continues until the marketer no longer requires primary control reserve. In step S8, it is checked whether an update interval (e.g., one second) has been reached or whether an event relevant for an update has been registered. If this is not the case ("N"), the overall primary control reserve power curve is maintained; otherwise, it is updated ("J"), e.g., by branching to step S1, as shown. Of course, the present invention is not limited to the embodiment shown. In general, "one," "an," etc., can be understood as singular or plural, particularly in the sense of "at least one" or "one or more," etc., unless this is explicitly excluded, e.g., by the expression "exactly one," etc.23-3563 PIF 25 / 31 A numerical specification can also include exactly the specified number as well as a usual tolerance range, unless this is explicitly excluded.

[0002] 23-3563 PIF 26 / 31 Reference List EM Energy Meter EMA Energy Market Aggregator EV Electric Vehicle EVN Energy Supply Network EVSE Charging Point [EVSE / EV]-i i-th Charging Pair f Grid Frequency f0 Target Grid Frequency f_PRL,max Upper Frequency Threshold f_PRL,min Lower Frequency Threshold HEMS Home Energy Management System K_PRL Primary Control Reserve Curve Section LS Property P Electrical Power P_bas Baseline P_ges Single Charging Curve P_PRL Primary Control Reserve f_PRL,max Maximum Chargeable Control Reserve P_PRL,min Maximum Dischargeable Control Reserve PI Pooling Instance S1-S8 Procedure Steps SM Smart Meter t Time |p Pool-Related Quantity

Claims

23-3563 PIF 27 / 31 Claims 1. Method for providing control power by means of a pool comprising several electric vehicles (EVs), comprising the steps: − Determining the respective charging flexibilities of the charging pairs ([EVSE / EV]-1 - [EVSE / EV]-n) allocated to the pool from electric vehicles (EVs) and charging points (EVSEs); − Establishing a grid frequency-dependent PRL power curve (P PRL) for each charging pair ([EVSE / EV]-1 to [EVSE / EV]-n) based on its respective charging flexibility; − Transferring the PRL power curves (PPRL) of the charging pairs ([EVSE / EV]-1 - [EVSE / EV]-n) to associated charging controllers, which are configured to control the charging of the associated electric vehicle (EV) at least depending on a current grid frequency (f) according to its PRL power curve (PPRL); − Establishing, by the pooling instance (PI), an overall PRL power curve (PPRL |p), which corresponds to an aggregation of PRL power curves (PPRL) of individual charging pairs ([EVSE / EV]-1 - [EVSE / EV]-n); and − Calculating a control power provided by the electric vehicles (EVs) of the pool based on the overall PRL power curve (P PRL | p ).

2. Method according to claim 1, wherein the individual PRL performance curves (P PRL) are set up locally at the charging pairs ([EVSE / EV]-1 - [EVSE / EV]-n), are reported to the pooling instance (PI) and the pooling instance (PI) derives the total PRL power curve (P) from this. PRL | p ) by aggregation.

3. Method according to claim 1, wherein − the charging flexibilities of the charging pairs ([EVSE / EV]-1 to [EVSE / EV]-n) assigned to the pool are reported to the pooling instance (PI) and − the pooling instance (PI) derives the individual PRL power curves (PPRL) for each charging pair ([EVSE / EV]-1 - [EVSE / EV]-n) and transmits them to the associated charging controllers.

4. Method according to claim 1, wherein 23-3563 PIF 28 / 31 − the charging flexibilities of the charging pairs assigned to the pool ([EVSE / EV]-1 - [EVSE / EV]-n) are reported to the pooling instance (PI), − the pooling instance (PI) derives the overall PRL power curve (P) from this PRL | p) sets up, − the pooling instance (PI) sets up the individual PRL power curves (PPRL) for each charging pair ([EVSE / EV]-1 - [EVSE / EV]- n) from the total PRL power curve (PPRL |p) by disaggregation and transfers them to the associated charging controllers.

5. A method according to any of the preceding claims, wherein: − the charging points are components of local power grids connected via an energy meter (SM, EM) to a power supply network (EVN) as the receiver of the control power; − the grid frequency (f) is monitored via the energy meters (SM, EM); − the monitored grid frequency (f) is transmitted to at least one charging controller of the local power grid, by means of which charging of at least one charging pair ([EVSE / EV]-1 - [EVSE / EV]-n) can be controlled; and − the charging controller controls charging of the at least one charging pair ([EVSE / EV]-1 - [EVSE / EV]-n) that it can control by means of the respective PRL power curve (PPRL). 6.Method according to one of the preceding claims, wherein a single PRL performance curve (P. PRL ) for a charging pair ([EVSE / EV]-1 [EVSE / EV]-n) of a particularly continuous function of the requested control power with the properties − for f ≤ f PRL,min : : P PRL (f) = - P PRL,min with f PRL,min a predetermined lower frequency threshold below a target network frequency (f0) and P PRL,min a maximum control power that can be discharged by the electric vehicle (EV), − for fPRL,min < f < fPRL,max: KPRL(f) with fPRL,max a predetermined upper frequency threshold above the target grid frequency (f0) and KPRL(f) a curve segment that is at least partially linear and increases monotonically with higher grid frequencies (f), − for f ≥ f PRL,max : P PRL (f) = P PRL,max with P PRL,max a maximum control power that can be charged by the electric vehicle (EV) 23-3563 PIF 29 / 31.

7. Method according to any of the preceding claims, wherein the charging controller performs a charging process of the electric vehicle (EV) of the associated charging pair ([EVSE / EV]-1 - [EVSE / EV]-n) based on a charging curve (Pges) which corresponds to a superposition of an associated baseline (Pbas) with the PRL power curve (PPRL).

8. Method according to any of the preceding claims, wherein the overall PRL power curve (PPRL |p) is set up such that the overall PRL power curve (PPRL |p) flattens out as it approaches at least one of the frequency thresholds (fPRL,max, fPRL,max) relative to the associated predetermined minimum and / or maximum primary control power (PPRL,min, PPRL,max).

9. Method according to one of the preceding claims, wherein when the mains frequency (f) is within a deadband of f0-f Tmin ≤ f ≤ f0+ f Tmaxaround the target frequency (f0), from which no primary control power is called upon.

10. Method according to one of the preceding claims, wherein the total PRL power curve (P PRL | p ) is re-established at regular intervals.

11. Method according to one of the preceding claims, wherein the overall PRL performance curve (P PRL | p ) is re-established triggered by an event.

12. Method according to one of the preceding claims, wherein the calculation of the control power is based on the overall PRL power curve (P PRL | p ) is carried out by means of the pooling instance (PI) and the result of the calculation is reported or transmitted to at least one corresponding market participant (EMA, EVN).

13. Method according to one of claims 1 to 11, wherein the calculation of the control power is based on the total PRL power curve (P PRL | p) is carried out in such a way that the overall PRL performance curve (P) established using the pooling instance (PI) PRL | p ) to at least one relevant market participant (EMA, EVN) 23-3563 PIF 30 / 31 reported and at least one corresponding market participant (EMA, EVN) calculates the control power from it.

14. Computer program product comprising code which, when executed on a data processing device, performs the method according to one of the preceding claims.

15. Control system comprising several charging pairs ([EVSE / EV]-1 to [EVSE / EV]-n) of respective electric vehicles (EV) and charging points (EVSE), whose charging processes can be controlled by means of a charging controller, and a central pooling instance (PI) linked to the respective charging controllers via data technology, wherein the control system is configured to perform the method according to one of claims 1 to 13.

Images