Electric vehicles platform for new charge infrastructure risk, tactical, and operational planning

Abstract

A method of monitoring and managing a network of charger modules for electric vehicles, EVs, is provided. The method comprises receiving charger module network data regarding the location of a plurality of charger modules in the network of charger modules, and receiving EV data from a plurality of EVs currently travelling in the region of the network of charger modules. The method further comprises generating a digital twin of the network of charger modules and EVs based on the charger module network data and the EV data, and monitoring, by the digital twin, the utilisation of the plurality of charger modules by the plurality of EVs. The method further comprises generating one or more predictions based on the monitoring of the utilisation of the plurality of charger modules by the plurality of EVs, and simulating, by the digital twin, the network of charger modules in one or more scenarios based on the generated one or more predictions. A report is then generated based on the simulations and outputted to a management entity.

Description

A35808 ELECTRIC VEHICLES PLATFORM FOR NEW CHARGE INFRASTRUCTURE RISK, TACTICAL, AND OPERATIONAL PLANNING TECHNICAL FIELD

[0001] Embodiments described herein relate generally to methods of management and planning for electric vehicle charging infrastructures. BACKGROUND

[0002] As the global electric vehicle market continues to grow, the associated charging infrastructure network for electric vehicles likewise needs to be developed and extended. At the same time, with increased energy usage constraints and increased congestion, such infrastructure must be efficient, flexible, and affordable. Therefore, when planning an extension or deployment of a new charging unit in a networked electricity system, there is the need to anticipate the utilisation of this new charging unit and its impact on the local power network, field engineers, and electric vehicle drivers. SUMMARY OF INVENTION

[0003] The present application relates to the field of electric vehicle infrastructureplanning, risk assessment, and electric vehicle charger allocation.

[0004] In accordance with a first aspect of the invention, there is provided a method of monitoring and managing a network of charger modules for electric vehicles, EVs, the method comprising: receiving charger module network data regarding the location of a plurality of charger modules in the network of charger modules; receiving EV data from a plurality of EVs currently travelling in the region of the network of charger modules; generating a digital twin of the network of charger modules and EVs based on the charger module network data and the EV data; monitoring, by the digital twin, the utilisation of the plurality of charger modules by the plurality of EVs; generating one or more predictions based on the monitoring of the utilisation of the plurality of charger modules by the plurality of EVs; simulating, by the digital twin, the network of charger modules in one or more scenarios based on the generated one or more predictions; and generating a report based on the simulations; outputting the generated report to a management entity.A35808

[0005] The present invention therefore provides a method of monitoring and managing a network of charger modules using a digital twin to provide more accurate predictions of charger utilisation and electric vehicle user behaviour. This can be adjusted dynamically and in real time based upon ongoing data collection, so that those predictions of electric vehicle user behaviours and charger module availability can be simulated and assessed for their accuracy. Recommendations regarding charger module deployments or redeployments may also be generated to ensure that the network of charger modules, and their utilisation by electric vehicles, may be optimised.

[0006] Any of the following may be applied to the above first aspect of the invention.

[0007] For each of the plurality of EVs, the EV data may include one or more of: information related to the type of EV; the current power consumption of the EV; the power consumption of the EV over a recent time period; and the current battery charge of the vehicle.

[0008] The EV data may be continuously received in real-time, and the digital twin may be continuously updated based on the received EV data.

[0009] The monitoring may include monitoring at least one of: which EVs of the plurality of EVs require charging at a charger module; which charger modules are currently experiencing a demand for charging that is above a demand threshold; which charger modules are currently experiencing a demand for charging that is below the demand threshold; and the remaining power capacity of at least one of the charging modules in the network of charging modules.

[0010] The method may further comprise: identifying, based on the monitoring, the EVs of the plurality of EVs that require charging at a charger module, and allocating each of the EVs of the plurality of EVs that require charging to a charger module; wherein the allocation of each EV to a charger module is optimised based on at least one of: the cost incurred by the EV travelling to the allocated charging module; the cost of running the allocated charging module and the power used charging the EV at that allocated charging module; the number of EVs currently assigned to the allocated charger module; the cost associated with any required increase in the power available at the allocated a charger module; and the cost of any required parking and charging fees at the allocated charger module.

[0011] The generated predictions may be predictions with respect to a future period of time, and the predictions may include at least one of: a prediction of the driving behaviour of at least one EV of the plurality of EVs over the period of time; a prediction of when at least one EV of the plurality of EVs will require charging at a charging module over the period of time; a prediction of the demand for charging at least one of the plurality of charger modules over the period of time; and the power availability of atA35808 least one charging module of the network of charging modules over the over the period of time.

[0012] The generated predictions may be based upon one or more of: the current battery levels of one or more of the plurality of EVs; the current rate of power consumption of one or more of the plurality of EVs; the model of one or more of the plurality of EVs; the location of one or more EVs; and the current route of one or more of the plurality of EVs.

[0013] Any of the generated predictions may be generated by one or more machine learning algorithms.

[0014] The method may further comprise: generating at least one recommendation for network optimisation based on the predictions, wherein the recommendation may include at least one of: a recommendation for the deployment of at least one new charger module within the network of charger modules; a recommendation for the redeployment of at least one charger module within the network of charger modules; wherein the generated report includes the at least one recommendation.

[0015] The simulating by the digital twin may be based on the recommendations.

[0016] The method may further comprise: generating a risk value based on the simulations, the risk value being associated with a risk of one or more charger modules experiencing a power crash, failure, or power shortage; wherein the generated report includes the at least one generated risk value.

[0017] The method may further comprise: generating an alert or notification to one or more EVs in the network of charger modules based on the risk value, wherein the alert or notification includes at least one of: a notification that one or more charger modules is not available for charging; a recommended route to an alternative charger module that is available for charging.

[0018] In accordance with a second aspect of the invention, there is provided a system comprising: one or more processors; a non-transitory memory; and one or more programs, wherein the one or more programs are stored in the non-transitory memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods of the first aspect of the invention discussed above.

[0019] In accordance with a third aspect of the invention, there is provided a non- transitory computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which, when executed by an electronic device with one or more processors, cause the electronic device to perform any of the methods of the first aspect of the invention discussed above.A35808

[0020] In the following, embodiments will be described with reference to the drawings in which:

[0021] Fig. 1 shows a block diagram of a computer system suitable for the operation of the method according to some embodiments.

[0022] Fig. 2 shows a flowchart of the function of the system of Figure 1 according to some embodiments.

[0023] Fig. 3 shows a flowchart of a method of allocating electric vehicles to charger modules according to some embodiments. DETAILED DESCRIPTION

[0024] The present application provides improved monitoring, planning, and risk assessment for electric vehicle charger networks, and an improved method for the optimised allocation of electric vehicles to charger stations (or modules). Aspects of the present invention include the generation and use of a digital twin of an electric vehicle charger network, which allows for improved monitoring of the network, more accurate predictions associated with the network, and more accurate simulations of the network based on those predictions.

[0025] Digital twins may be used to optimize the operation and maintenance of physical assets, systems, and manufacturing processes, helping organisations to make better- informed decisions, leading to improved outcomes in the physical world. Digital twins are provide an ability to collect and visualise near real-time data, enabling smart analytics and customised rules to effectively achieve objectives.

[0026] A digital twin may be defined as an evolving digital profile of the historical and current behaviour of a physical object or process that helps optimize performance. The digital twin may be based on massive, cumulative, near real-time, real-world data measurements across an array of different dimensions. At its core, a digital twin is a digital replica of a living or non-living physical entity. What distinguishes a digital twin from any other digital model is its connection to its physical twin. Such a digital representation provides both the elements and the dynamics of how a physical device or system operates and lives throughout its life cycle.

[0027] A digital twin must represent physical reality at a level of accuracy suited to its purpose. Digital twins integrate different technologies such as the Internet of Things, Artificial Intelligence, among others to create living digital simulation models that update and change as their physical counterpart change. A digital twin continuously learns and updates itself from multiple sources to represent its near real-time status, working condition or position. This learning system learns from itself, using sensor dataA35808 that conveys various aspects of its operating condition, as well as from human experts, such as engineers with deep and relevant industry domain knowledge. A digital twin may also learn from other similar machines, from other similar fleets of machines, and from the larger systems and environment of which it may be a part.

[0028] Digital twins may be developed for a range of different purposes, operate at different scales or adopt different approaches to modelling.

[0029] The present application employs a digital twin in a network of charging modules for electric vehicle in order to provide an accurate digital version of the physical network, allowing for more accurate monitoring of that network. In addition, the digital twin of the present application allows for the system to make more accurate predictions about the network, and to simulate that physical network under certain conditions and scenarios. This then allows for more efficient optimisation of the network, in terms of new charger module deployments or existing charger module redeployments, as well as a more optimised allocation of electric vehicle to those charger modules in the network.

[0030] An infrastructure for electric vehicle charging can be described as a network of charger modules 100.

[0031] Here, the network of charger modules 100 is located across a particular area (e.g. across a geographical region such as a city or rural area) and comprises a plurality of charger modules 110. Each charger module 110 may be configured to charge one or more electric vehicles (EVs) 105. A charger module 110 of the network of charger modules 100 may be located on private premises (i.e. a residential home or a business site), or may be located on public premises (e.g. an on-street charger).

[0032] As a result, a user of an EV 105 that is travelling along a route that traverses the area in which the network of charger modules 100 is located is able to charge their EV 105 if needed at one of the charging modules 110 in the network.

[0033] The system of the present application may be employed to manage the charging of multiple EVs, both dynamically and in real time, and over a future time period. The EVs in question may be driven by private users (i.e. members of the public), or the EVs may be driven for commercial purposes (e.g. where the EVs are electric vehicles being used by a business).

[0034] In the system of the present application, a digital twin (DT) 120 is generated based on data received from individual charger modules 110 within the network of charger modules 100, as well as from EVs travelling within the network of charger modules 100.

[0035] The system is then able to use the generated digital twin 120 to simulate the network and make predictions of the charging requirements for individual chargerA35808 modules 110 in the network over time. In addition, the system is able to use the generated digital twin 120 to simulate and make recommendations regarding where new charger modules 110 should be installed (e.g. at a location that has a high demand from users for charging their EVs). Furthermore, the system is able to use the generated digital twin 120 to simulate and make recommendations when existing charger modules 110 may be at an increased risk of experiencing a power failure or at an increased risk of experiencing a power shortage (e.g. due to high charger demand from users for charging their EVs).

[0036] In addition, the system is able to use the generated digital twin 120 to simulate and provide recommendations for users of EVs 105 in the network, such as when the user should next seek to charge their EV, which nearby charging module 110 in the network would be the most optimal charger module 110 for charging the EV 105, and how long charging that EV 105 would take at that recommended charging module 110.

[0037] Figure 1 is a block diagram of the system of the present application. It is notedthat not all of the units of the system shown in Figure 1 are necessary in order to implement the invention as described herein, and a skilled reader would understand that certain modules may be considered optional.

[0038] As shown in Figure 1, the system of the present application may comprise a plurality of components. For example, the system may comprise a Digital Twin (DT) module 220, an EV Charge Location Engine (EVCLE) 225, an EV Charging Prediction (EVCP) module 230, and a Connected Vehicle Smart Box (CVSB) module 235. In some embodiments, the system may comprise a dedicated Smart EV to Charger Allocation Engine (SEVCAE) 240 for determining an optimal allocation of EVs to charger modules 110 within the network of charger modules 100.

[0039] Here, the CVSB module 235 receives data in real time from a plurality of EVs 105currently within the network of charging modules 100. This data may be referred to as “EV data”. The EV data received from each EV 105 may provide the system with information relating to the route being taken by the EV and the charging requirements of that EV 105.

[0040] For example, the EV data received by the CVSB module 235 may include information about the type of electric vehicle 105 (i.e. the model of the vehicle), the current power consumption of the vehicle, the recent power consumption of the vehicle (e.g. power consumption over a given time period or over the course of one or more recent journeys), and / or the current battery charge level of the vehicle.

[0041] The EV data received by the CVSB module 235 may also include informationregarding the current route of the EV 105. For example, if the EV is travelling to a particular destination and is currently in an autonomous driving mode, or if the user isA35808 using a navigational aid to reach a particular destination (e.g. by employing GPS navigation).

[0042] The data received by the CVSB module 235 may also include information regarding the user’s parking usage or habits. For instance, such information may include how frequently the user parks (which may be indicated by the average length of the user’s recent journeys). Such information may also include any parking locations that a user uses more frequently (e.g. at a particular business location, residential location, or any location preferred by a user of an EV 105).

[0043] The EV data received by the CVSB module 235 may be collected by one or more sensors of the EV 105, a smart box of the EV 105, or may be collected by a user device in the EV (e.g. a driver’s mobile phone using a mobile app).

[0044] The system then uses the data received by the CVSB module 235 to generate a digital twin (DT) 120 of the network of charging modules 100. Alternatively, if a DT 120 has previously been generated, the data may be used to update the DT 120. Here, the DT 120 is generated, updated, and stored on the Digital Twin (DT) module 220.

[0045] The DT module 220 may also receive real time data from the system regardingcurrent levels of power availability and current levels of power consumption for each of the charger modules 110 in the network of the charger modules 100. The DT module 220 may then use that real time data in the generation of the DT 120 or, if the DT 120 has previously been generated, the updating of the DT 120.

[0046] The system then uses the DT 120 as a virtual representation of the physical network of charger modules 100, allowing the system to monitor the utilisation of charge modules 110 in the network of charger modules 100 in real time.

[0047] For example, the DT 120 allows the system to monitor which charger modules 110 are currently under higher demand for EV charging, and which charger modules 110 are currently under-utilised. For example, this may include monitoring which charger modules 110 in the network of charger modules 100 are currently experiencing a demand for charging that is above (i.e. under a higher demand) or below (i.e. under utilised) a certain demand threshold. Such a demand threshold may take any appropriate form (e.g. the number of EVs 105 currently charging from a given charging module 110, and / or the number of EVs currently allocated or travelling to given charging module 110).

[0048] This also allows the system to monitor the remaining power capacity of the charger modules 110 in the network of charger modules 100, as well as monitor the current demand for EV charging based on the data from the EVs 105 currently in the network 100 (i.e. that data received by the CVSB module 235 and provided to the DT module 220).A35808

[0049] The system may then use the EV Charging Prediction (EVCP) module 230 to generate predictions based on the real time monitoring of the network of charger modules 100. Such predictions may include predictions of user driving behaviour, charging demand for certain charging modules 110, and power availability both at certain charging modules 110 and across the network 100 as a whole.

[0050] The predictions generated by the EV Charging Prediction (EVCP) module 230 may also cover a certain time period. For example, the EV Charging Prediction (EVCP) module 230 may make predictions for the next given period of time (e.g. for the next 6 hours), or may make predictions based on the monitoring data from the DT 120 for a particular time of day or a particular day of the week.

[0051] For example, the EVCP module 230 may predict the expected demand for EV charging at a particular charger module 110 or group of charger modules 110 over a coming time period (e.g. over the next 3, 6, or 12 hours).

[0052] For example, the EVCP module 230 may predict, based on the monitoring by the DT 120, that a number of EVs 105 currently travelling in the network 100 will soon require charging. This may be due to the monitoring data from the DT 120 indicating that the battery levels of those EVs 105 is currently below a certain threshold, or that the particular model of an EV has a higher than average rate of power consumption.

[0053] As a further example, the EVCP module 230 may predict, based on the monitoring by the DT 120, the behaviour of a number of EV users currently travelling in the network that have a historical preference for parking near to certain charger modules 110 within the network 100. As a result, the EVCP module 230 may predict that those charger modules 110 are likely to experience a higher demand for charging in the short-term.

[0054] The EVCP module 230 may also use the monitoring data from the DT 120 to make predictions regarding power availability in parts of the network of charger modules 100. For example, the data provided to the DT 120 may include an indication that the power supplied to a certain area of the network 100 is temporarily reduced at a certain time. As a result, the EVCP module 230 may predict that the power availability (i.e. the power available to charge EVs 105 in the network) in that area will be temporarily reduced, as the power supply in that area is temporarily reduced whilst demand for charging from EVs 105 remains at a certain level.

[0055] In some embodiments, the EV Charging Prediction (EVCP) module 230 may generate a prediction of upcoming charger (i.e. power) demand for a plurality of EVs 105 as a function of one or more of the current battery levels of each of those EVs 105, the rate of power consumption of each of those EVs (which may be related toA35808 the model of the EV), the location of each of those EVs 105, the current route or journey of each of those EVs.

[0056] In some embodiments, the EV Charging Prediction (EVCP) module 230 may generate a prediction of the remaining charging (or power) capacity at a site (or in an area of the network 100) as a function of power consumption (i.e. how much power is being consumed by one or more charger modules 110), charge activities scheduled (i.e. where one or more charger modules 110 are expected to be in demand for charging soon), and actual charger demand (i.e. where one or more charger modules 110 are currently in demand for charging).

[0057] In some embodiments, the EVCP module 230 may employ a machine learningmodule 230a to make any of the predictions regarding the utilisation of the network of charger modules 100 discussed above.

[0058] In some embodiments, the predictions generated by the EVCP module 230 maybe passed to the EV Charge Location Engine (EVCLE) 225. Here, the EVCLE 225 may take the predictions of user driving behaviour, charging demand, and power availability to generate recommendations for new charger module deployments.

[0059] For example, if the monitoring of the DT 120 indicates that an area within the network of charger modules 100 frequently experiences a demand that exceeds a certain threshold, the EVCLE 225 may generate a recommendation that a new charger module 110 should be deployed in that area.

[0060] In addition, the EVCLE 225 may assess the predictions received from the EVCP module 230 and generate a recommendation for a redistribution of charge modules 110 within an area of the network of charger modules 100.

[0061] For example, if an area within the network 100 is consistently under-utilised (e.g. if certain charger modules 110 are not used to charge EVs 105 frequently), then the EVCLE 225 may generate a recommendation that one or more charger modules 110 in that area be removed, or be redeployed to another area where the demand for charging EVs 105 is higher.

[0062] The system is also then able to use the DT 120 to carry out simulations of the network of charger modules 100. For instance, the DT 120 may be used to simulate the expected demand for EV charging at a particular charger module 110 or group of charger modules 110 over a chosen period using the predictions generated by the EVCP module 230.

[0063] For example, where the EVCP module 230 predicts that a number of EVs 105 currently travelling in the network 100 will soon require charging, the DT 120 may then simulate the increased short-term demand on the charging modules 110 in that particular area of the network 100 (or, as may be needed, across the network 100).A35808

[0064] As a further example, where the EVCP module 230 predicts that certain charger modules 110 are likely to experience a higher demand for charging in the short-term due to EV users currently travelling in the network having a historical preference for parking near to those certain charger modules 110, the DT 120 may then simulate whether those certain charger modules 110 will be able to meet that higher predicted demand.

[0065] Therefore, this allows to system to more accurately assess in advance whether the network of charger modules 100 will be able to meet the predicted demands from EVs 105 currently in the network.

[0066] It is highlighted that the DT 120 may simulate the charging demand over any period of time where adequate data can be collected and predictions can be made. For example, where the CVSB module 235 receives data allowing the EVCP module 230 to generate a prediction regarding the EV charger demand over a 4 hour period on a Wednesday, the DT 120 may then simulate that demand over that 4 hour period.

[0067] In some embodiments, the simulations of the network of charger modules 100 carried out by the DT 120 may also utilise the recommendations generated by the EVCLE 225.

[0068] This simulating by the DT 120 therefore allows the system to utilise this virtual representation of the network of charger modules 100, so that the network 100 can be more accurately modelled in view of the data collected and predictions generated regarding EV driver behaviour, network power availability, and charger demand.

[0069] That is to say, using the simulations from the DT 120 allows the system to more accurately model EV driver behaviour, charger power availability, and how a level of charger demand might affect the network 100.

[0070] This also allows the system to model how the installation of new charger modules 110 might affect the ability of the network 100 to meet charger demand in certain areas or at peak times.

[0071] Similarly, this also allows the system to model how the relocation of existing charger modules 110 might affect the ability of the network 100 to meet charger demand in certain areas or at peak times.

[0072] Furthermore, the DT module 220 may use the DT 120 to also determine a risk of one or more charger modules 110 of the network of charger modules 100 experiencing a power crash or failure, or a power shortage as a consequence of over- demand.

[0073] For example, a simulation by the DT 220 may suggest that a charger module orcharger modules 110 in the network 100 is / are likely to experience a power crash or failure, or a power shortage, in the near future. This may be ascertained by simulatingA35808 the levels of power availability in the charger module(s) and concluding that there is a risk of the charger module(s) 110 failing or experiencing a shortage of power.

[0074] The risk determined by the DT 120 may take the form of a risk value, where that risk value is determined to exceed a particular threshold.

[0075] If the DT module 220 does determine that there is a risk, the DT module 220may then output a notification to a management entity, such as a management system or management staff, indicating that a risk that one or more charger modules may experience a power failure, or power shortage, has increased. This notification may take the form of an alert or a report. The notification may include a recommendation to replace the one or more charger modules 110 in the network of charger modules 100 that are determined as being at a higher risk of power crash / failure, or to install one or more additional charger modules 110 to address the higher risk of a power shortage.

[0076] Similarly, the DT module 220 may receive the recommendations generated bythe EVCLE 225 regarding new charger module installations or deployments, and may output a report to the management entity (e.g. a management system or management staff) detailing those recommendations. Alternatively, the EVCLE 225 may output that report directly to a management system or to management staff.

[0077] The DT module 220 may also proactively notify users of EVs 105 in the network 100 that are scheduled to charge at the one or more charger modules 110 that are determined as being at a risk of power crash / failure, or a power shortage. Here, the DT module 220 may notify the users of those EVs 105 that the charger module 110 is not available for charging, and recommend a route to one or more alternative charger modules 110 that are available for charging (e.g. where those one or more alternative charger modules 110 have sufficient power capacity and are not at a risk of power crash / failure).

[0078] The alternative charger module(s) 110 recommended to the users of those EVs 105 may be selected by determining an optimised route for those EVs (e.g. a route that minimises travel time, charging time, and / or optimises the overall power availability in the network 100).

[0079] In view of the above, the system is then able to provide network planning and recommendations for new charger installations within a hybrid charging infrastructure (either existing or planned), combining private chargers either installed at service provider sites or at individual locations (for example, in a private home) with public chargers (for example, on street chargers), to cope with a scope of EV users / drivers.

[0080] This approach optimises the overall power balance, cost, EV driver / user experience, and automatically verifies multiple charger technologies, multiple drivingA35808 practices, and consequently EV battery power requirements over extended periods of time (e.g. over several days / nights).

[0081] Figure 2 shows a flowchart of the function of the system of Figure 1. As such, the functionality discussed above with regard to Figure 1 may be applied to the steps of the flowchart shown in Figure 2.

[0082] Some of the steps shown in Figure 2 are optional steps (that is to say, not essential to work the invention of the present application). Such optional steps are indicated as such with a dotted line.

[0083] At Step 1000, charger module network data is received, where the chargermodule network data includes information regarding a network of charger modules 100, the network comprising a plurality of charger modules 110 distribution over a physical area (e.g. a geographical region such as a city or rural area). The information in the charger module network data may include real time data regarding current levels of power availability and current levels of power consumption for each of the charger modules 110 in the network of the charger modules 100.

[0084] In addition, at Step 1000 EV data is received in real time from a plurality ofEVs 105 currently within the network of charging modules 100. The EV data may include information relating to the current or planned route being taken by each of the plurality of EVs 105, the charging requirements of those EVs 105, the type of EVs 105 (i.e. the models of one or more of those vehicles), the current power consumption of those EVs, the recent power consumption of those EVs, and / or the current battery charge level of those EVs. The EV data may also include information regarding the parking usage and habits of the users of each of those EVs, such as how frequently each user parks and / or any parking locations that that user of the EV 105 in question uses more frequently.

[0085] At Step 1100, a digital twin (DT) 120 of the network of charging modules 100 is generated based on the EV data and the charger module network data, forming a virtual representation of the physical network of charger modules 100. Alternatively, if a DT 120 has previously been generated, the EV data and the charger module network data may be used to update the DT 120.

[0086] At Step 1200, the utilisation of charge modules 110 in the network of charger modules 100 is monitored in real time.

[0087] For example, which charger modules 110 are currently under higher demand for EV charging, and which charger modules 110 are currently under-utilised, are monitored. The monitoring may also include the remaining power capacity of the charger modules 110 in the network of charger modules 100, as well as the current demand for EV charging based on the received EV data.A35808

[0088] At Step 1300, predictions are generated based on the monitoring of the network of charger modules 100. The predictions may include predictions of EV user driving behaviour, EV user parking behaviour, charging demand for one or more chosen charging modules 110, and power availability at chosen charging modules 110 in a chosen area and / or across the network 100 as a whole. The predictions generated may also cover a chosen time period.

[0089] The predictions may be generated using a machine learning algorithm. Here, the machine learning algorithm may be any suitable type or method of machine learning algorithm. In some embodiments, the machine learning algorithm may be trained on test data supplied to the system, or may be trained on any other suitable training data.

[0090] In some embodiments, Step 1300 may proceed to Step 1400. If Step 1400 is not included, the process proceeds directly to Step 1500.

[0091] At Step 1400, recommendations for charger modules 110 deployments may be generated based on the predictions. Here, the predictions of user driving behaviour, charging demand, and power availability are assessed to generate the recommendations for charger module deployments.

[0092] The recommendations for charger module 110 deployments may include recommendations for new charger module 110 deployment, or may include recommendations for the redeployment or redistribution of existing charger modules 110. The recommendations are generated in order to optimise the network of charger modules 110 based on the predicted user driving and parking behaviour, charging demand, and power availability in the network 100.

[0093] At Step 1500, the DT 120 of the network 100 simulates the network 100 based on the predicted user driving behaviour, charging demand, and power availability. Here, the virtual representation of the physical network of charger modules 100 generated by the DT 120 is used to simulate scenarios based on the generated predictions of Step 1300.

[0094] In some embodiments, the DT 120 of the network 100 simulates the network 100 based on the generated recommendations for charger module 110 deployments. The simulation may include a simulation based on the recommendations for new charger module 110 deployment, or the recommendations for the redeployment or redistribution of existing charger modules 110.

[0095] In some embodiments, Step 1500 may proceed to Step 1600. If Step 1600 isnot included, the process proceeds directly to Step 1700.

[0096] At Step 1600, a risk may be determined based on the simulation run by the DT 120. Here, the risk is a risk of one or more charger modules 110 of the network of charger modules 100 experiencing a power crash or failure, or a power shortage as aA35808 consequence of an excessive power charging demand from the EVs 105 in the network 100.

[0097] At Step 1700, a report is generated based on the simulations of Step 1500 and output to a management entity. Here, the management entity may be a management system or management staff. The output may take the form of a notification, an alert, or a report. For example, the generated report may include one or more predictions generated in Step 1300, and the one or more results of the associated simulation(s) of Step 1500.

[0098] In some embodiments, the output to the management entity may also include the generated recommendations for charger module 110 deployments and / or redeployments generated in Step 1400.

[0099] In some embodiments, the output to the management entity may also include the risk determined in Step 1600. Here, the output may indicate that a risk of one or more charger modules experiencing a power failure, or power shortage, has increased. The notification may include a recommendation to replace the one or more charger modules 110 in the network of charger modules 100 that are determined as being at a higher risk of power crash / failure, or a recommendation to install one or more additional charger modules 110 to address the higher risk of a power shortage in a particular area.

[0100] At Step 1800, an alert or notification may be output to one or more EVs 105 in the network 100 based on the determined risk from Step 1600. Here, the alert or notification may be output to one or more EVs 105 that are scheduled to charge at the one or more charger modules 110 that are determined as being at a higher risk of power crash / failure, or a power shortage. The alert or notification may notify the users of those EVs 105 that the one or more charger modules 110 is / are not available for charging, and recommend a route to one or more alternative charger modules 110 that are available for charging (e.g. where those one or more alternative charger modules 110 have sufficient power capacity and are not at a risk of power crash / failure).

[0101] The alternative charger module(s) 110 recommended to the users of thoseEVs 105 may be selected by determining an optimised route for those EVs (e.g. a route that minimises travel time, charging time, and / or optimises the overall power availability in the network 100).

[0102] Figure 3 shows a flowchart of a method used to allocate an EV 105 in the network 100 to a charger module 110 of the network of charger modules 100 according to some embodiments. This method may be implemented with, or separately to, any of the above subject matter and embodiments discussed with reference to either of Figures 1 or 2.A35808

[0103] Here, the system may include a Smart EV to Charger Allocation Engine (SEVCAE) 240, configured to allocate a charging module 110 to each EV 105 currently within in the network of charging modules 100.

[0104] The allocation of a charging module 110 to an EV 105 is designed to be optimised for the network of charger modules 100.

[0105] For example, the allocation of an EV 105 to a charging module 110 may seek to ensure that certain aims are achieved. For example, the allocation may seek to ensure that one or more of the following objectives are achieved: - The cost incurred by the reassignment or redirection of the EV to the allocated charging module 110 is minimised. - The cost of running the charger module and the power usage charging the EV at that allocated charging module 110 is minimised (e.g. in scenarios where the running costs of charger modules 110 varies across the network 100). - The number of EVs 105 assigned to charger modules 110 is maximised (e.g. maximum utilisation of the charger modules 110 based on the current demand for charging by EVs 105 in the network of charger modules 100). - The cost associated with increasing the power available at a charger module in order to meet a higher demand for charging is minimised. - The cost of parking and charging fees for charger modules 110 that are located in a public location (e.g. on street public charging) is minimised.

[0106] Furthermore, the allocation of EVs 105 may be subject to certain constraints defined by the current configuration and characteristics of the network of charging modules 100. For example, such constraints may include: - The availability of the parking needed for the EV to be allocated to a given charger module 110. This may be a function of the type or model of EV and the parking capacity at each charging location. - Each EV 105 is not assigned to more than one charger module 110. - The number of EVs 105 currently assigned to the allocated charging module(s) at a given charging location is less that the number of charger module outlets available at the charging location. That is to say, there needs to be enough charging outlets available at the charging location in order to charge the assigned EV. - The current level of power consumption for the charger module(s) 110 at a given site. This may be a function of the numberA35808 of EVs, and the type or model of those EVs, already assigned to the charger module(s) at a given charging location and the power capacity of the charger module(s) at that charging location. - The power requirements of the battery of the EV being assigned. That is to say, the EV should be allocated a charger module 110 at a charging location that meets the power needs of the EV’s battery. - For charging locations that are at a public site (e.g. public on-street charging), the availability of outlets or plugs for public charging. That is to say, there needs to be sufficient outlets or plugs at the public charging location to charge the EVs assigned to that location. - That the charging technology required by the EV (which may be based on the type or model of the EV) being compatible with the charger technology of the assigned charger module 110.

[0107] The SEVCAE 240 is then configured to calculate the most optimal allocation of EVs 105 to charger modules 110, taking into account the above objectives and constraints.

[0108] In more detail, the SEVCAE 240 in some embodiments may represent the allocation of the EVs 105 using the following decision variables Yvb and Xcb: - Xcb represents the number of charger modules 110 of type “c” assigned to a charging location “b” (e.g. a site or building). - Yvb represents the allocation of an EV “v” to the most suitable charging location “b” where a charger module is deployed. Here, this allocation must be valid taking into account constraints such as those described above.

[0109] The decision variable Yvb may be a binary decision variable. That is to say, Yvb may take the value “1” if the vehicle “v” is assigned to charging location “b”, and otherwise take the value “0” if the vehicle “v” is not assigned to charging location “b”.

[0110] The decision variables may be summarised as shown in Table 1 below: Decision Variables Name Description Units Xcb Number of charger modules of type “c” assigned to charging location “b” Integer Yvb 1 if electric vehicle “v” is assigned to charging location “b” Binary Table 1

[0111] In some embodiments, a modifier hypv may be employed. Here, the modifier hypvallows the model employed by the SEVCAE 240 to prioritise home parkers by takingA35808 a value greater than 1. This then ensures that users of EVs 105 that have a charger module 110 at their home are consistently allocated to that home charger module 110.

[0112] In some embodiments, if the model employed by the SEVCAE 240 is configured to include no distinction between home charge modules 110 for an EV and other charge modules 110 (such as charge modules 110 located at a business site or public street), then the modifier hypv is omitted or takes the value “1” (i.e. it does not serve as a modifying variable).

[0113] The model employed by the SEVCAE 240 then calculates a solution for each EV 105 to ascertain whether it can be allocated to a charging module 110 in the network of charging modules 100. If no solution can be found within the constraints of the model (i.e. constraints such as those discussed above), then the SEVCAE 240 cannot allocate that EV 105 a charging module 110 at that time.

[0114] The model may employ multiple variables or attributes associated with eachEV 105, each charger module 110, and each charging location, as well as other variables associated with travel distance and type of charge technology “t”.

[0115] Examples of these variables and attributes may be summarised as shown inTables 2, 3, 4, 5, and 6 below: Sets Name Description Units V Set of electric vehicles “v” - B Set of charging locations “b” - C Set of types of charger modules “c” - Table 2 Attributes of Each Electric Vehicle “v” Name Description Units tv Type of electric vehicle - pnv Power requirements kW mvAverage milage driven per day - bcv Battery capacity £ / kWh pkv Parking requirements £ / kW hypvHome parking modifier (hypv> 1 if vehicle has a home charger module) - Table 3A35808 Attributes of Each Charging Location “b” Name Description Units tb Type of charging location - pcb Power capacity at charging location kW pkb Parking capacity at charging location - epb Electricity price at charging location £ / kWh cirb Power cap increase cost rate £ / kW pubb 1 if charging location is public, 0 otherwise - ncb Number of charging outlets at charging location (0 if private) - Table 4 Attributes of Each Charger Module Type “c” Name Description Units pnc Power needed kW spc Plugs for simultaneous charging (i.e. charging multiple EVs at once) - icc Installation cost £ mcc Technology cost £ Table 5 Other Attributes Name Description Units dvb Matrix of current distance of electric vehicle “v” from charging location “b” km ctvt Matrix matching electric vehicle “v” to charger technology “t” 1 if match ctct Matrix matching charge module type “c” to charger technology “t” 1 if match Table 6

[0116] In view of the above example variables and attributes, the constraints discussed above may be summarised as shown in Table 7 below:

[0117] Formalisation of Constraints 1) For all charging locations, the total allocated electric≤ ^^^^^^ ∀^^ ∈ ^^vehicles must be equal to or less than the parking ^^ ∈ ^^ capacity of the charging location ∑^^2) Each electric vehicle not assigned to more than one ^^^^ ≤ 1 ∀^^ ∈ ^^^^ ∈ ^^ charger moduleA35808 3) For all charging locations, the number of available^^ ∈ ^^ than the power cap at that charging location ∀^^ ∈ ^^,∑ ^^6) The number of electric vehicles is not more than th ^≤ ^e ^^^ ^^^^^where ^^ ∈ ^^ number of outlets available for public charging pubb = 1 7) The type of charging technology of the charger If Yvb = 1, then t and c exist and module is compatible with the charging technology ctvt = 1 and ctct = 1 and Xcb ≥ 1 required by the electric vehicle Table 7

[0118] Here, constraint number 3) states that for all EVs compatible with a particular charger type (ctvt) there must be enough available outlets (or plugs) of a particular charger type (ctct). The number of outlets (or plugs) is obtained using the number of charger modules 110 (Xcb) and the number of outlets each charger module has (spc).

[0119] Further, it is highlighted that constraint number 4) can be relaxed into a heuristic of optimisation driving the increase of power limit if that brings advantages to the overall objective function. The power supply limit may be ignored if the planning scenario calls for increasing that power supply limit.

[0120] In addition, constraint number 6) states that if the charging location is one or more public (e.g. on-street) charger modules (i.e. pubb = 1), then the number of EVs allocated may not be higher than the number of available outlets (or plugs) of those one or more charger modules at that charging location (ncb).

[0121] In view of the above variables and attributes, and in view of the above constraints, the SEVCAE 240 is then configured to determine the most optimal allocation of EVs 105 to charger modules 110.

[0122] This method of optimisation may be based upon searching for a solution that minimises the objective cost of allocating the EVs 105 to charger modules 110. This search may be implemented with integer programming and heuristic search.A35808

[0123] In this regard, in some embodiments this optimisation method may seek to achieve the following formalised objectives shown in Table 8: Formalisation of Objectives ^^^^^^1) The total cost of travelling to charging location 1= ∑ ∑ ^^^^^^ . ^^^^^^ . ^^^^ . ^^^^^^ . ℎ^^^^^^^^ ∈ ^^ ^^ ∈ ^^ (i.e. the distance cost) ^^^^^^2 = ∑ ∑ ^^^^^^ . (^^^^^^ + 3. ^^^^^^). (1 − ^^^^^^^^)2) The total cost of all deployed charger modules ^^ ∈ ^^ ^^ ∈ ^^ ^^^^^^3) Maximising the number of electric vehicles 3= −1 ∗ ∑ ∑ ^^^^^^^^ ∈ ^^ ^^ ∈ ^^ allocated to a charger module ^^^^^^ = ∑ ∑ (^^ . ^^^^ − ^^^^ ) . ( )4) The total cost of increasing the power limit 4^^^^ ^^ ^^ ^^^^^^^^ . 1 − ^^^^^^^^^^ ∈ ^^ ^^ ∈ ^^ when needed and allowed. 5) The cost of parking fees and charging fees in the 1 ^^^^^^5 = ∑ ∑ (^^^^^^^^ + ^^^^^^^^ . . ^^^^∈ ^^ ^^ ∈ ^^ ^^^) . ^^^^^^case of electric vehicles being allocated to public ^^ ^^^charger modules (e.g. on street public charging) Table 8

[0124] Here, the first objective Obj1 may be to minimise the driving distance between the current location of the EVs and their allocated charging location. Since the effect of moving an EV to a new charging location will mean a certain travel distance for the user driving it, the cost of this commute can be accounted for. This may be captured in the model employed by the SEVCAE 240 by multiplying Obj1 by a weighting factor W1.

[0125] The second objective Obj2 may be to minimise the cost of charger modules 110, including their installation and maintenance cost. The variable Xcb is the integer decision variable indicating how many charger modules “c” are assigned to a charging location “b”. The variable icc is the one-time installation cost and this is added to the maintenance cost mcc. The maintenance cost mcc may be aggregated over a period of time, for instance over 3 years. The formula of Obj2 may include the variable pubb, since the model may assume that the installation and maintenance costs can be ignored if the location is considered a public charging location, such as on street charging (here, there may instead be a utilisation cost).

[0126] The third objective Obj3 may be to maximise the number of vehicles assigned.

[0127] The fourth objective Obj4 may be to minimise the cost of increasing the capacity of the power supply from the grid if needed due to increased allocation of EVs to a given charging location (where the charger modules 110 at that charging location are private or non-public charger modules 110). Here, this objective may be optimised where theA35808 power capacity of the charging location can be raised, dependent upon constraint number 4) discussed above with regard to Table 7.

[0128] The fifth objective Obj5 may be to minimise the cost of parking fees and charging fees. Parking fees may only be present for public charging locations and may be set to zero for private charging locations (for instance, business sites or private residences).

[0129] The above objectives may then be balanced based upon the real time data recorded by the system. For example, the SEVCAE 240 may assign a ranking for each objective based on the current real-time requirements of the network 100, and / or assign a weighting to each objective, so that objectives with a higher priority are met over objectives of a lower priority.

[0130] Although not essential in the allocation of charger modules 110 to the EVs 105, in some embodiments the SEVCAE 240 may also receive information from the DT module 220. The SEVCAE 240 may then employ that information in generating the optimum allocation of charger modules 110 to each of the EVs 105 (for instance when assigning a ranking for each objective and / or when assigning a weighting to each objective).

[0131] In some embodiments, the allocation of an EV to a charger module 110 may be to meet an immediate and / or near future demand for charging that vehicle. In some embodiments, the allocation of an EV to a charger module 110 may for part of a long-term charging schedule for that EV. For example, where the driving activity of an EV is repetitive, that EV may be allocated to a particular charger module 110 at a particular time for 3 hours of charging every 2 days. Such a long-term charging schedule may be dynamically updated as new allocations are made, or as new EV or charger module data becomes available.

[0132] With reference to Figure 3, at Step 2000 the system (e.g. the SEVCAE 240) identifies all EVs 105 within the network of charger modules 100 that require charging in near term.

[0133] For example, the system may use EV data received from the EVs (e.g. the data received from EVs by the CVSB module 235) to identify those EVs 105 within the network 100 that will require charging within the next hour (or any suitable selected period of time). Alternatively, the user of an EV may manually send a request to the system (e.g. over a wireless data network) for their EV to be allocated to a charger module 110 for charging, at which point the system identifies that EV as requiring charging.

[0134] In some embodiments, an EV may be determined to be in need of charging at a charging station if the battery level of that EV is below a predetermined threshold. Alternatively, in some embodiments an EV may be identified as to be in need of chargingA35808 at a charging station if the battery level of that EV is projected to fall below that predetermined threshold within a given period of time (in view of that EVs current rate of power consumption).

[0135] At Step 2100, the system (e.g. the SEVCAE 240) identifies the objectives for calculating an optimum allocation of the EVs identified as being in need of charging to appropriate charging modules 110. These objectives may correspond to the objectives discussed herein, although other objectives many be identified depending on the system and the specific configuration of the network of charger modules 100.

[0136] At Step 2200, the system (e.g. the SEVCAE 240) assigns a ranking for each objective based on the current real-time requirements of the network 100 (e.g. based on the data received from EVs by the CVSB module 235). In addition, or separately, the system (e.g. the SEVCAE 240) may assign a weighting to each objective, so that objectives with a higher priority are met over objectives of a lower priority.

[0137] That is to say, by ranking or weighting (or by both), the system is able to prioritise solutions that optimise more important objectives over those objectives of a lower importance.

[0138] At Step 2300, the system (e.g. the SEVCAE 240) identifies the constraints in calculating the optimum allocation of the EVs identified as being in need of charging to appropriate charging modules 110. These constraints may correspond to the constraints discussed herein, although other constraints many be identified depending on the system and the specific configuration of the network of charger modules 100.

[0139] At Step 2400, the system (e.g. the SEVCAE 240) calculates a solution to each objective for each EV identified as being in need of charging at a charging module 110. Each solution to each objective for each EV must also conform to the identified constraints.

[0140] At Step 2500, the system (e.g. the SEVCAE 240) determines, for each EV 105 that is identified as being in need of charging, a charger module 110 which that EV can be allocated to for charging. This determination is based on the calculated solutions of each objective for that EV, as well as the weighting or ranking applied to each objective for that EV.

[0141] At Step 2600, the system (e.g. the SEVCAE 240) sends a notification to each EV 105 identified as being in need of charging at a charger module 110, the notification including the charger module 110 which that EV has been allocated to. In some embodiments, the notification may additionally include a recommended route for the EV to travel to the charger module 110 it has been allocated to.

[0142] In some embodiments, the notification may take the form of a schedule of charging for that EV, covering a predefined period of time (for example a schedule forA35808 periodically charging at one or more allocated charger modules 110 over a period of time).

[0143] Any of the above discussed methods may be performed using a computer system or similar computational resource, or system comprising one or more processors and a non-transitory memory storing one or more programs configured to execute the method. Likewise, a non-transitory computer readable storage medium may store one or more programs that comprise instructions that, when executed, carry out the methods described herein.

[0144] Whilst certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the application. Indeed, the novel devices, and methods described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the devices, methods and products described herein may be made without departing from the scope of the present application. The word "comprising" can mean "including" or "consisting of" and therefore does not exclude the presence of elements or steps other than those listed in any claim or the specification as a whole. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope of the application.

Claims

A35808 CLAIMS 1. A method of monitoring and managing a network of charger modules for electric vehicles, EVs, the method comprising: receiving charger module network data regarding the location of a plurality of charger modules in the network of charger modules; receiving EV data from a plurality of EVs currently travelling in the region of the network of charger modules; generating a digital twin of the network of charger modules and EVs based on the charger module network data and the EV data; monitoring, by the digital twin, the utilisation of the plurality of charger modules by the plurality of EVs; generating one or more predictions based on the monitoring of the utilisation of the plurality of charger modules by the plurality of EVs; simulating, by the digital twin, the network of charger modules in one or more scenarios based on the generated one or more predictions; generating a report based on the simulations; and outputting the generated report to a management entity.

2. The method according to claim 1, wherein, for each of the plurality of EVs, the EV data includes one or more of: information related to the type of EV; the current power consumption of the EV; the power consumption of the EV over a recent time period; and the current battery charge of the vehicle.

3. The method according to claim 1 or 2, wherein the EV data is continuously received in real-time, and the digital twin is continuously updated based on the received EV data.

4. The method according to any preceding claim, wherein the monitoring includes monitoring at least one of: which EVs of the plurality of EVs require charging at a charger module; which charger modules are currently experiencing a demand for charging that is above a demand threshold; which charger modules are currently experiencing a demand for charging that is below the demand threshold; andA35808 the remaining power capacity of at least one of the charging modules in the network of charging modules.

5. The method according to claim 4, wherein the method further comprises: identifying, based on the monitoring, the EVs of the plurality of EVs that require charging at a charger module, and allocating each of the EVs of the plurality of EVs that require charging to a charger module; wherein the allocation of each EV to a charger module is optimised based on at least one of: the cost incurred by the EV travelling to the allocated charging module; the cost of running the allocated charging module and the power used charging the EV at that allocated charging module; the number of EVs currently assigned to the allocated charger module; the cost associated with any required increase in the power available at the allocated a charger module; and the cost of any required parking and charging fees at the allocated charger module.

6. The method according to any preceding claim, wherein the generated predictions are predictions with respect to a future period of time, and wherein the predictions include at least one of: a prediction of the driving behaviour of at least one EV of the plurality of EVs over the period of time; a prediction of when at least one EV of the plurality of EVs will require charging at a charging module over the period of time; a prediction of the demand for charging at least one of the plurality of charger modules over the period of time; and the power availability of at least one charging module of the network of charging modules over the over the period of time.

7. The method according to claim 6, wherein the generated predictions are based upon one or more of: the current battery levels of one or more of the plurality of EVs; the current rate of power consumption of one or more of the plurality of EVs; the model of one or more of the plurality of EVs;A35808 the location of one or more EVs; and the current route of one or more of the plurality of EVs.

8. The method according to any preceding claim, wherein the generated predictions are generated by one or more machine learning algorithms.

9. The method according to any preceding claim, wherein the method further comprises: generating at least one recommendation for network optimisation based on the predictions, wherein the recommendation includes at least one of: a recommendation for the deployment of at least one new charger module within the network of charger modules; a recommendation for the redeployment of at least one charger module within the network of charger modules; wherein the generated report includes the at least one recommendation.

10. The method according to claim 9, wherein the simulating by the digital twin is based on the recommendations.

11. The method according to any preceding claim, wherein the method further comprises: generating a risk value based on the simulations, the risk value being associated with a risk of one or more charger modules experiencing a power crash, failure, or power shortage; wherein the generated report includes the at least one generated risk value.

12. The method according to claim 10, wherein the method further comprises: generating an alert or notification to one or more EVs in the network of charger modules based on the risk value, wherein the alert or notification includes at least one of: a notification that one or more charger modules is not available for charging; a recommended route to an alternative charger module that is available for charging.A35808 13. A system comprising: one or more processors; a non-transitory memory; and one or more programs, wherein the one or more programs are stored in the non-transitory memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods of claims 1 to 12.

14. A non-transitory computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which, when executed by an electronic device with one or more processors, cause the electronic device to perform any of the methods of claims 1 to 12.

Images