Methods and systems for dynamical power sharing

Abstract

The present disclosure relates to an autonomous decision-making method for dynamical power sharing for a cluster of one or more controllable power units, comprising: obtaining for each controllable power unit a local power setpoint using a global reference power of the cluster, and a global power consumption or production of the cluster; continuously calculating and updating the local power setpoint for each controllable power unit using a feedback loop using the global reference power of the cluster, the global power consumption or production of the cluster, a local priority factor for each controllable power unit, and a local compensation factor for each controllable power unit. The present disclosure further relates to a system configured for dynamical power sharing for a cluster of one or more controllable power units. The present disclosure enhances scalability, mitigates single points of failure, and reduces data communication needs, making it suitable for applications that require secure, efficient, and decentralized power management, such as electric vehicle charging.

Description

[0001] P7345PC00

[0002] 1

[0003] Methods and systems for dynamical power sharing

[0004] The present disclosure relates to methods and systems for dynamic and autonomous power sharing in networks. The present disclosure aims at enhancing efficiency and stability in decentralized energy systems. An embodiment of the disclosure concerns electrical vehicle charging wherein power demand fluctuations and peak loads necessitate a reliable and flexible distribution method.

[0005] Background

[0006] The increasing reliance on renewable energy sources and distributed energy systems and units has placed new demands on power distribution networks. Modern energy systems often consist of clusters of controllable electrical units, such as energy generation units, energy storage systems, and various load devices. Efficiently managing and sharing power among these controllable units is important for maintaining grid stability and ensuring optimal energy use both for renewable and nonrenewable energy sources, especially in the context of decentralized and fluctuating energy supplies.

[0007] Traditionally, power sharing within such systems has been managed using centralized control architectures, where a single control unit oversees the distribution of power to and from individual units. While effective in small-scale applications, centralized control systems face several limitations. As the number of connected units increases, the centralized system encounters computational burdens and communication traffic, making it difficult to scale efficiently. Additionally, centralized systems are prone to single points of failure — if the central control unit fails, the entire system can be disrupted, leading to grid instability or power outages.

[0008] Another drawback of centralized systems is the heavy reliance on continuous data communication between the control unit and the connected units. This extensive data exchange increases the risk of cyberattacks and security breaches, and it makes the system vulnerable to external threats. Moreover, in scenarios where communication networks are unstable or unavailable, such as in remote or industrial locations, the centralized architecture may be impractical or unreliable. This creates a need for more resilient and secure methods of power management that can adapt to changing conditions without constant data flow.P7345PC00

[0009] 2

[0010] Hence, there is a need for new control architectures for power sharing that meet the growing demand for efficient power allocation and address the challenges related to scalability and vulnerability in existing and centralized power sharing architectures.

[0011] Summary

[0012] It is an objective of the present disclosure to provide an improved method for autonomous decision-making and dynamic power sharing among clusters of controllable electrical units. The solution aims to address the limitations of centralized control systems by enhancing scalability, reducing the risk of single-point failures, and minimizing the need for continuous data communication while ensuring efficient and secure power distribution.

[0013] The present disclosure concerns methods and systems for dynamic power-sharing to coordinate the distribution and management of controllable power units. Instead of relying on a centralized management system, a distributed management system is introduced, where each power unit autonomously determines its power setpoint based on global inputs such as the cluster reference power (Pref) and the measured cluster pOWer (Pmeas).

[0014] Specifically, the present disclosure relates to an autonomous decision-making method for dynamical power sharing for a cluster of one or more controllable power units, comprising: obtaining for each controllable power unit a local power setpoint using a global reference power of the cluster, and a global power consumption or production of the cluster; continuously calculating and updating the local power setpoint for each controllable power unit using a feedback loop using the global reference power of the cluster, the global power consumption or production of the cluster, a local priority factor for each controllable power unit, and a local compensation factor for each controllable power unit. The controllable power units may be distributed controllable power units.

[0015] One advantage of this distributed architecture is its ability to enhance scalability, robustness, and cybersecurity while reducing both the need for extensive data exchange and the vulnerabilities of centralized systems. The distributed architecture reduces both data traffic and the computational burden, making the system more efficient. Possible benefits can be summarized in the form of three main categories as follows.P7345PC00

[0016] 3

[0017] Roll-out of power infrastructure. The presently disclosed method facilitates an efficient deployment of various types of controllable power units. It enables installation in locations with limited electricity network capacity, as the method enables coordinated power management. The ability to deploy power infrastructure quickly can be important for meeting growing demands in various sectors, and the presently disclosed method is expected to accelerate the adoption of such technologies, contributing to improved energy efficiency and reduced carbon emissions across multiple industries.

[0018] Unlocking the flexibility potential of power systems. The operational processes of controllable power units often offer time flexibility, which can be harnessed with the right coordination tools. The presently disclosed method provides an effective mechanism to manage power distribution while still meeting the operational goals of individual users. Benefits to customers and society include: (1) Cost-optimized operation; power utilization can be shifted to times of lower electricity prices, reducing overall costs. (2) Green operation; utilizing power during periods of higher renewable energy availability reduces carbon emissions. (3) Reliable grid operation; coordinating large-scale power units can help maintain the stability of the electricity grid, which holds economic and societal value.

[0019] Reduced data traffic and enhanced cybersecurity. The presently disclosed method decreases the need for extensive data communication when coordinating the operation of a large number of power units. High data traffic can lead to issues such as network clogging, slower response times, higher infrastructure costs, and increased energy consumption. By reducing the data exchange requirements, the system enhances cybersecurity through a smaller attack surface, easier detection of anomalies, faster incident response, and mitigation of threats like Distributed Denial of Service (DDoS) attacks. Data privacy is also improved, as individual operational preferences and measurements do not need to be communicated, further securing the system.

[0020] Brief description of drawings

[0021] Various embodiments are described hereinafter with reference to the drawings. The drawings are examples of embodiments and are intended to illustrate some of the features of the presently disclosed method and system for dynamic and autonomousP7345PC00

[0022] 4

[0023] power sharing in networks, and are not limiting to the presently disclosed method and system.

[0024] FIG. 1 shows a system comprising a plurality of controllable power units connected to a power system through a point of common coupling (PCC), wherein a cluster reference power and the measurement outcome of a power consumption or production are communicated to the controllable power units.

[0025] FIG. 2 shows the control architecture comprising input and outputs, as well as the feedback loop, for a single controllable power unit connected to a power system.

[0026] FIG. 3 shows a flowchart illustrating one possible implementation of the presently disclosed method of calculating and updating the local power setpoint for each of the controllable power units.

[0027] FIG. 4 shows a simulation of power profiles for a system comprising two controllable power units, wherein the local priority factors are communicated between the controllable power units.

[0028] FIG. 5 shows a simulation of power profiles for a system comprising two controllable power units, wherein the local priority factors are communicated between the controllable power units, and wherein the compensation factor is reduced compared to FIG. 4.

[0029] FIG. 6 shows a simulation of power profiles for a system comprising two controllable power units, wherein the local priority factors are communicated between the controllable power units, and wherein the power of the first controllable power unit is suddenly reduced to 0.

[0030] FIG. 7 shows a simulation of power profiles for a system comprising two controllable power units, wherein the local priority factors are not communicated between the controllable power units.

[0031] FIG. 8 shows a simulation of power profiles for a system comprising two controllable power units, wherein the local priority factors are not communicated between the controllable power units, and wherein the power setpoint for each unit is updated less frequently compared to FIG. 7.

[0032] FIG. 9 shows a simulation of power profiles for a system comprising two controllable power units, wherein the local priority factors are not communicated between the controllable power units, and wherein the local priority factor of the first controllable power unit is reduced compared to FIG. 7.P7345PC00

[0033] 5

[0034] FIG. 10 shows a simulation of power profiles for a system comprising three controllable power units, wherein the local priority factors are communicated between he controllable power units, and wherein unit 1 is configured for power consumption only, whereas units 2 and 3 are configured for both power consumption and power production.

[0035] Detailed description

[0036] Further details:

[0037] Embodiments of the presently disclosed method and system are associated with various advantages and / or technical effects.

[0038] The present disclosure concerns an autonomous decision-making method for dynamical power sharing for a cluster of one or more controllable power units, comprising: obtaining for each controllable power unit a local power setpoint using a global reference power of the cluster, and a global power consumption or production of the cluster; continuously calculating and updating the power setpoint for each controllable power unit using a feedback loop using the global reference power of the cluster, the global power consumption or production of the cluster, a local priority factor for each controllable power unit, and a local compensation factor for each controllable power unit. The global reference power may be a measured global reference power. In case a measured value is not available, for example, due to communication failure or outage, it may also be possible to use an estimated global reference power. The same applies for other measured values.

[0039] In the present disclosure, a “cluster” refers to a grouping or aggregation of power units arranged in a collaborative configuration to share resources, perform collective functions, or optimize performance within a defined system or network. The cluster may comprise power units that are physically proximate or virtually connected and are capable of interacting or coordinating operations to achieve specific objectives, such as power distribution, load balancing, or redundancy. Certain terms and variables relating to the cluster herein are referred to as being “global”, such as “global cluster power setpoint” or “global cluster power consumption”, to emphasize that these quantities are measured or defined centrally and without reference to the individual and controllable power units.P7345PC00

[0040] 6

[0041] In the present disclosure, the rate at which the power setpoint for each controllable power unit is calculated and updated may be any predefined rate, but preferably higher than any intrinsic and limiting time scale of the application at hand. Hence, “continuously” should be interpreted broadly and comprises, for example, updating rates ranging from several times per second, such as at a rate of 100Hz, to once every few minutes. The specific update rate may be chosen based on a number of factors such as type of application, number and type of controllable power units associated with the cluster, and network setup and requirements.

[0042] In the present disclosure, the term “power” is used to denote electrical power, and refers to the rate at which electrical energy is transferred, consumed, or generated within a system. In DC systems, power is measured in watts (W). In AC systems, we may distinguish between apparent power, measured in volt-amperes (VA); active power, measured in watts (W); and reactive power, measured in volt-amperes reactive (VAR). In multi-phase systems, electrical quantities such as power, voltage, and current can be defined for each individual phase separately. This allows for phasespecific analysis and control, which is particularly relevant in systems where phases may experience different loading conditions or imbalances. As would be realized by a person skilled in the art, the term “power” as used in the present disclosure is not strictly limited to power as such, since it is also possible to measure and calculate based on terms that are directly related to the power, including current and voltage.

[0043] The local compensation factor and the local priority factor associated with a controllable power unit may both comprise multiple values. For example, a bidirectional controllable power unit configured for both power production and power consumption may have one local compensation factor and / or local priority factor associated with production and another local compensation factor and / or local priority factor associated with consumption. This may facilitate achieving optimized resource allocation, enhanced operational efficiency, and / or flexibility.

[0044] In one embodiment of the present disclosure, at least two of the controllable power units are connected through a point of common coupling. The point of common coupling serves as a centralized node where power flows from the connected controllable power units are aggregated and managed. This arrangement facilitates coordination of power consumption or production across the cluster, allowing forP7345PC00

[0045] 7

[0046] precise regulation of global power metrics such as total consumption, production, or power balance.

[0047] The point of common coupling provides a shared interface that enables the system to monitor and manage the global cluster power consumption or production in real-time. This structure simplifies the measurement of aggregated power metrics and ensures that individual power units operate cohesively within the cluster. The shared coupling point also enhances the efficiency of power sharing by enabling rapid feedback and adjustment of local power setpoints based on dynamic changes in the overall cluster requirements. For example, if one unit experiences a sudden increase in load, the system can redistribute power among the units connected to the coupling point to maintain balance and prevent instability. The implementation of a point of common coupling may vary depending on the specific application and system design. For instance, it can be realized through a physical connection such as a shared bus in an industrial power system or through virtual aggregation in a software-defined grid for distributed energy resources.

[0048] In an embodiment, one or more compensation factors are determined for the controllable power units of a cluster in order to enable dynamic balancing and reallocation of available cluster power among the controllable power units. The compensation factors may be used to adjust the local power setpoints of the controllable power units such that the aggregate power consumption and / or power production of the cluster converges towards a desired cluster reference power while accounting for deviations between allocated power shares and actual measured power.

[0049] A purpose of the compensation factors is to provide a mechanism for continuous reallocation of the available cluster power among the controllable power units. This may be particularly advantageous in systems in which the controllable power units reach consensus on how to share the available cluster power with limited communication between the controllable power units and / or without a centralized management system. In such cases, the compensation factors provide a corrective term that allows the local control actions of each controllable power unit to collectively achieve a consistent global behavior, for example by compensating for temporary mismatches, local constraints, changing availability, or changes in the cluster composition.P7345PC00

[0050] 8

[0051] In practical operation, the compensation factors may effectively cause one or more controllable power units to “make space” for other controllable power units by reducing their own local power setpoints when their measured power consumption or production exceeds a power share to which they are presently entitled (for example as determined by a base power allocation derived from a local priority factor). Conversely, the compensation factors may cause one or more controllable power units to increase their local power setpoints when their measured power consumption or production is below their allocated share and where additional cluster power is available to be utilized.

[0052] The compensation factors may be updated continuously, for example in a feedback loop based on measured local and / or global power values, and / or may be updated in response to triggering events such as addition or removal of controllable power units, changes in availability or constraints of controllable power units, or changes in the cluster reference power.

[0053] In one embodiment of the present disclosure, the compensation factors are defined to subtract or add a share of the measured power consumption or production above or below a base power, wherein the base power is defined as the cluster reference power weighted by the associated local priority factor. This approach ensures that each controllable power unit contributes to or draws from the cluster power based on its designated share, allowing for precise adjustments to the power distribution within the cluster. In this embodiment, the compensation factors are defined to add or subtract a share of the measured power consumption if the controllable power unit is configured to produce or consume energy, respectively.

[0054] By employing compensation factors tied to the measured power consumption or production relative to a base power, the system dynamically balances power flows in response to real-time conditions. The base power, weighted by local priority factors, ensures that individual power units are assigned consensus-based power levels. This provides a robust mechanism for maintaining stability and optimizing resource utilization across the cluster, as deviations from the base power can be immediately corrected by redistributing the excess or deficit power among the units.P7345PC00

[0055] 9

[0056] In one embodiment of the present disclosure, the compensation factors are proportional such that they subtract or add a proportional share of the measured power consumption or production above or below the base power for each controllable power unit. This proportionality ensures that adjustments to power distribution are consistently aligned with the relative priority of each power unit, thereby maintaining a balanced, self-regulating, and equitable power-sharing arrangement within the cluster.

[0057] By defining the compensation factors as proportional, the system dynamically scales power adjustments in response to changes in the measured power consumption or production of individual units. For example, if a unit's consumption exceeds its allocated base power, a proportionate reduction is applied to ensure that the cluster relaxes to its defined power constraints. Similarly, for units producing excess power, proportional compensation prevents overloading the cluster while ensuring optimal utilization of available resources. This approach minimizes the risk of instability and ensures that the system can respond seamlessly to fluctuations in power demand or generation.

[0058] In one embodiment of the present disclosure, the compensation factors are defined to subtract or add a share of the previous power setpoint of the associated controllable power unit. This approach ensures that adjustments to the power setpoints are informed by the unit’s recent operational history, enabling smooth transitions and enhanced stability in the power distribution process across the cluster.

[0059] By utilizing the previous power setpoint as a reference, the system introduces a feedback mechanism that reduces abrupt changes in power allocation. This can help prevent sudden fluctuations that might destabilize the cluster or place undue stress on individual power units. For instance, if a unit's power requirements change due to a shift in load or generation capacity, the system can incrementally adjust its power setpoint based on a proportion of the prior setpoint, providing a more gradual and controlled adaptation to new conditions. By anchoring changes in power setpoints to previous values, the system can better accommodate the dynamic nature of power demands while minimizing the risk of overshooting or undershooting the required adjustments.P7345PC00

[0060] 10

[0061] In one embodiment of the present disclosure the compensation factors are triggerbased such that they after defined events subtract or add a fixed or proportional share of the global power consumption or production above or below the base power for one or more of the controllable power units. This configuration enables the system to respond dynamically to specific conditions or events by implementing targeted adjustments in power distribution, enhancing both adaptability and operational control within the cluster.

[0062] Trigger-based compensation factors allow the system to initiate adjustments only when predefined conditions are met, such as a threshold being reached, a time-based event occurring, or an external signal being received. For example, the compensation factors may be triggered to redistribute power following a sudden spike in demand, a drop in generation capacity, or the detection of an imbalance in the cluster’s power metrics. By tying the activation of compensation factors to these events, the system can minimize unnecessary adjustments and focus on addressing significant changes that require immediate intervention. This embodiment integrates event-driven adjustments with the system’s broader dynamical power-sharing framework.

[0063] In one embodiment of the present disclosure, the method further comprises communication of the local priority factors or their sum between the controllable power units. This feature enables the exchange of priority information within the cluster, ensuring that each power unit can adjust its behavior in a coordinated manner based on the collective priorities of the system.

[0064] By facilitating the communication of local priority factors, the method ensures that each unit has access to up-to-date information regarding the allocation of the cluster’s reference power. This allows the system to dynamically redistribute power consumption or production based on the relative importance or role of each unit. For instance, a unit with a higher priority factor may consume or produce more power during times of scarcity, while lower-priority units are proportionally constrained.

[0065] Sharing the sum of priority factors across the cluster can instead support a sparser and more cyber-resilient communication protocol without compromising functionality. This variation provides a global reference for maintaining balance, as the total allocated power can be continuously aligned with the cluster’s operational objectives.P7345PC00

[0066] 11

[0067] This communication framework enhances the scalability and adaptability of the system by enabling efficient collaboration among power units.

[0068] In one embodiment of the present disclosure, the method comprises dynamically updating at least one of the local priority factors. This feature allows the system to adapt to changing conditions within the cluster by modifying the priority levels of individual controllable power units in real-time, ensuring that power-sharing decisions remain optimal and / or aligned with current operational needs.

[0069] The ability to dynamically update local priority factors provides significant flexibility in managing the cluster. For example, priority levels can be adjusted based on factors such as shifts in power demand. A controllable power unit supplying critical infrastructure might be assigned a higher priority during a power shortfall, while less important units could be temporarily deprioritized.

[0070] In one embodiment of the present disclosure, the method comprises using a sum of all local priority factors, each defining the allocated share of the cluster reference power for the individual electrical power units, to respect a cluster power consumption or production limit. This ensures that the total power utilized or generated by the cluster remains within predefined thresholds, maintaining the balance and stability of the overall system.

[0071] By using the sum of the local priority factors, the system enables a coordinated approach to power sharing that respects the cluster's operational constraints. Each controllable power unit’s share of the cluster reference power is determined in proportion to its priority factor, and the cumulative sum ensures that the combined allocation does not exceed the global power limit. Using the sum of the local priority factors serves to define a common normalization convention across the controllable power units. For instance, during peak demand, the system can dynamically adjust power distribution to prevent overconsumption while still adhering to the relative priorities of individual units. Similarly, in production scenarios, this embodiment ensures that power generation stays within safe or optimal levels by scaling contributions from the units based on their priority.P7345PC00

[0072] 12

[0073] In one embodiment of the present disclosure, the method comprises all the local priority factors are fixed to be the value reciprocal to the number of controllable power units for each of the controllable power units. This configuration ensures an equal distribution of the cluster reference power among all units, providing a straightforward and balanced approach to power sharing within the system.

[0074] By assigning equal priority factors, the system simplifies the allocation process and associated communication, as each controllable power unit receives an identical share of the cluster reference power. This approach may be effective in scenarios where all units have similar capacities, roles, or operational requirements. For example, in a grid comprising multiple identical energy storage units, equal priority factors ensure that each unit contributes equally to power supply or storage without introducing additional complexity. This uniform distribution also minimizes the need for dynamic adjustments, allowing the system to maintain consistent performance under stable operating conditions.

[0075] In one embodiment of the present disclosure, at least one local priority factor is calculated using technical specifications of an incoming unit upon connecting to a controllable power unit, or using electrically linked units, or using a set of sub-units, or using one or more user requests, or using power ratings of units, or a combination thereof. This approach allows the system to dynamically assign local priority factors based on the unique characteristics or requirements of the connected units.

[0076] By calculating local priority factors based on technical specifications, the system can account for parameters such as the unit's power capacity, efficiency, or operational role. For example, a unit with a higher capacity or important function may be assigned a greater priority factor, ensuring it receives an appropriately larger share of the cluster reference power. Similarly, local priority factors can be adjusted to reflect relationships between units, such as electrically linked devices that must operate in coordination, or sub-units that contribute to a larger functional entity. User-defined requests or constraints, such as prioritizing specific loads or optimizing energy costs, can also be incorporated into the priority calculations, enabling further customization of the system's power-sharing strategy.P7345PC00

[0077] 13

[0078] In one embodiment of the present disclosure, the method comprises a subsequent step after the power setpoint for each of the controllable power units have been calculated, the additional step involving forcing the power setpoint to stay below a maximum power setpoint and / or above a minimum power setpoint. This ensures that the power delivered or consumed by each unit remains within predefined limits, providing a safeguard against overloading or underutilizing the units within the cluster. The step of forcing the power setpoint does not necessarily have to be done in one step. It can also be done in incremental steps or using a rate limiter to avoid too big sudden changes.

[0079] By enforcing maximum and minimum power setpoints, the system ensures that each controllable power unit operates within safe and efficient ranges. For example, a maximum power setpoint can prevent a unit from exceeding its capacity, thereby protecting it from damage or inefficiency due to overloading. Conversely, a minimum power setpoint can ensure that certain units maintain a baseline level of operation, such as guaranteeing minimum power delivery for certain loads or ensuring sufficient charging levels in energy storage systems. This step adds an additional layer of control, maintaining system stability even in scenarios with rapidly changing power demands or generation levels.

[0080] In one embodiment of the present disclosure, the controllable power units are configured to consume or produce at least a predetermined fraction of the available cluster power. This ensures that each unit contributes to or draws from the cluster's power resources in a manner that aligns with the system's operational objectives and avoids underutilization or overloading of individual units.

[0081] By setting a predetermined fraction of the available cluster power for each controllable power unit, the system maintains a balanced distribution of power while ensuring efficient utilization of resources. For instance, in scenarios where a unit is important for supporting base loads or grid stability, it can be configured to consume or produce a minimum fraction of the cluster power to ensure continuous operation.

[0082] In one embodiment of the present disclosure, wherein the global cluster reference power is dynamically updated in response to conditions affecting delivery or extraction of power, for instance controlled by availability of local renewable energy sources, peak shaving, participation in aggregated portfolios, electricity prices, services related toP7345PC00

[0083] 14

[0084] provisions to grid operations, or combinations thereof. This ensures that the system remains responsive to changing external and internal conditions, allowing for optimized power-sharing strategies that align with operational requirements and / or market dynamics.

[0085] By dynamically updating the global cluster reference power, the system can adapt to fluctuations in power generation or consumption within the cluster. For example, in a renewable energy grid, changes in solar or wind availability can prompt adjustments to the reference power, ensuring that the cluster maintains stability while maximizing the use of available resources. Similarly, the system can respond to external signals such as electricity market prices, shifting power production or consumption to align with cost optimization strategies. In peak shaving scenarios, the cluster reference power can be reduced during periods of high demand to alleviate stress on the larger grid, contributing to overall system efficiency.

[0086] This dynamic adjustment capability is particularly useful in applications where external conditions significantly impact power availability or requirements, such as smart grids, industrial microgrids, or virtual power plants.

[0087] In one embodiment of the present disclosure, the method further comprises broadcasting the global cluster reference power and the global cluster power consumption or production to the individual controllable electrical power units using a communication system. This facilitates real-time coordination within the cluster, ensuring that each power unit has access to the latest global power metrics necessary for dynamically adjusting its operation.

[0088] By broadcasting the global cluster reference power and power consumption or production data, the system can establish a unified communication framework that allows each controllable power unit to operate in harmony with the cluster's objectives. For instance, when the global reference power changes due to variations in demand or generation, the updated value is immediately communicated to all units, enabling them to adjust their local power setpoints accordingly. Similarly, broadcasting the global power consumption or production data provides each unit with an aggregated view of the cluster's status, ensuring balanced and efficient power distribution.P7345PC00

[0089] 15

[0090] In one embodiment of the present disclosure, the local priority factors are broadcasted using a communication system. This enables the seamless dissemination of the relative importance or operational roles of each controllable power unit across the cluster, ensuring coordinated decision-making and efficient power-sharing behavior.

[0091] By broadcasting the local priority factors, the system allows each power unit to adjust its operation in alignment with the priorities of the entire cluster. For example, a unit with a lower local priority factor can reduce its power consumption or production when higher-priority units require additional resources. This mechanism ensures that the overall power distribution reflects the operational goals and constraints of the cluster while dynamically adapting to changes in demand or generation. The broadcasted priority factors also allow new units to integrate into the system with minimal disruption, as they can immediately align their behavior with the cluster's established priorities. The communication system, which may be used to communicate local priority factors, can be an analogue, digital, or any other technically feasible communication system.

[0092] In one embodiment of the present disclosure the communication system is Wi-Fi, Ethernet, TCP / IP, Modbus, MQTT, OPC UA, Profibus, Profinet, CAN bus, BACnet, Zigbee, LoRaWAN, HART, RS-232, RS-485, DeviceNet, EtherCAT, AS-lnterface, IO-Link, Bluetooth, Z-Wave, DNP3, KNX, EnOcean, Fieldbus, LonWorks, Serial Modem, ISM band communication, M-Bus, Power Line Communication (PLC), Thread, 6L0WPAN, Sigfox, Cellular (3G, 4G, 5G), SCADA, or combinations thereof. This selection allows for a wide range of communication protocols to be employed, ensuring the adaptability of the system to various infrastructures and operational requirements.

[0093] The use of diverse communication systems enables the method to be implemented across different technological environments. For example, in a localized industrial setting, wired protocols such as Ethernet or Modbus may be preferred for their reliability and high data rates. In contrast, wireless protocols like Wi-Fi, Zigbee, or Cellular communication are advantageous in scenarios where physical connectivity is impractical or when the system is distributed over a large geographic area, such as in renewable energy grids or remote microgrids.

[0094] In one embodiment of the present disclosure, the global power consumption or production is obtained as the sum of the power consumption or production from each ofP7345PC00

[0095] 16

[0096] the controllable power units and wherein the power consumption or production of each controllable power units or their sum is communicated to all the controllable power units. This configuration provides a comprehensive and transparent overview of the cluster’s overall power metrics, enabling precise and coordinated decision-making across all units.

[0097] By aggregating the power consumption or production data from each controllable unit, the system calculates the global power metrics required to maintain balance and optimize resource allocation within the cluster. Communicating this aggregated information, or the individual measurements, ensures that each unit has access to realtime data on the cluster’s operational status. For example, if the global power consumption approaches the cluster’s reference power limit, individual units can immediately adjust their power setpoints to prevent overloading. Conversely, in production scenarios, the shared data allows units to collaboratively optimize output to meet demand efficiently.

[0098] In one embodiment of the present disclosure, the method is configured for systems comprising a combination of at least one controllable power unit and at least one uncontrollable power unit, wherein the method is applied to a subset of the controllable power units. This configuration allows the system to operate effectively in mixed environments where not all units have the capability to dynamically adjust their power consumption or production.

[0099] By focusing the method on a subset of controllable power units, the system ensures that dynamic power sharing and optimization can still be achieved, even when some units are fixed in their behavior, or are controllable but not part of the set of units to which the control method applies. For example, in a hybrid energy system comprising renewable generators, energy storage units, and fixed-output power systems, the method can actively manage the controllable units — such as batteries or adjustable loads — to balance the overall power flows within the cluster. The uncontrollable units or the units which are not part of the set to which the control method is applied, while not directly influenced by the method, are accounted for in the global power metrics, ensuring that their fixed contributions or demands are incorporated into the system's power-sharing decisions.P7345PC00

[0100] 17

[0101] In one embodiment of the present disclosure, the method is configured for dynamical current sharing instead of dynamical power sharing. This configuration enables the system to focus on managing the distribution of electrical current within the cluster, ensuring balanced operation among the controllable power units based on currentspecific parameters.

[0102] By implementing dynamical current sharing, the system can adapt to scenarios where precise control over current distribution is beneficial. For example, in applications involving batteries, fuel cells, or other energy storage devices, maintaining balanced current flows can be important to avoid overloading individual units or causing imbalances that could impact the overall stability of the system. Dynamical current sharing also supports applications where the electrical characteristics of the cluster, such as resistance or voltage variations, require current-specific adjustments to optimize performance.

[0103] In one embodiment of the present disclosure, the controllable power units are charging stations configured for charging electrical vehicles. This configuration enables the system to dynamically manage power distribution among multiple charging stations, optimizing the use of available cluster power while addressing the specific demands of electric vehicle charging.

[0104] By implementing the method in the context of charging stations, the system ensures that each station receives an appropriate allocation of power based on factors such as demand, vehicle charging priority, or grid constraints. For instance, charging stations connected to vehicles with high-priority charging requests or low battery levels can be allocated more power, while those connected to vehicles near full charge can be deprioritized. This dynamic allocation helps to maximize charging efficiency across the cluster while avoiding overloading the grid or individual power units.

[0105] This application is beneficial for electric vehicle charging networks, including public charging hubs, fleet depots, or residential charging systems, where power demand fluctuates significantly and efficient and effective distribution can be important. The presently disclosed approach optimizes energy use, reduces waiting times for users, and prevents bottlenecks in the charging process, making it effective for public charging hubs, fleet operations, and smart city infrastructures. Furthermore, itP7345PC00

[0106] 18

[0107] enhances the resilience of the network, enabling it to handle fluctuations in grid supply or unexpected surges in demand without compromising stability or service quality.

[0108] In one embodiment of the present disclosure, the method further comprises using the local priority factors to set an urgency of a power consumption or production request from the one or more charging stations. This allows the system to dynamically allocate power among charging stations based on the specific urgency or priority of the vehicles being charged, optimizing the charging process for both efficiency and user satisfaction.

[0109] By incorporating urgency into the allocation process, the system ensures that electrical vehicles with higher charging priority, such as those with low battery levels or timesensitive requirements, receive a greater share of the cluster power. For example, a local priority factor can be dynamically adjusted based on factors such as the state of charge of the connected vehicle, user-defined charging preferences, or external inputs like fleet scheduling needs. This targeted power distribution prevents delays for high-priority vehicles while maintaining sufficient power for lower-priority stations, ensuring balanced and efficient operation of the charging network.

[0110] The presently disclosed system and method is not limited to charging stations for electrical vehicles. Rather, the disclosed system and method may be employed in a broad range of technical fields in which multiple controllable units share a common resource and are coordinated to collectively satisfy one or more global constraints and / or a global reference value (for example a reference power, reference current, reference flow, reference throughput, or other aggregated setpoint). The controllable units may comprise consuming units, producing units, storage units, or hybrid units, and may be connected behind a common point of coupling or may be distributed and coordinated as a portfolio.

[0111] By way of non-limiting example, the disclosed system and method may be applied in one or more of the following application areas:

[0112] The disclosed system and method may be applied to power plants or generation parks comprising multiple generation subsystems and / or multiple individual generating units, such as wind power plants comprising multiple wind turbines (optionally of differentP7345PC00

[0113] 19

[0114] types and rated powers), photovoltaic parks comprising multiple PV strings, inverters, or subsystems, or any combination thereof. In such systems, the disclosed control method enables the individual generating units to coordinate their respective output powers in order to collectively fulfill a reference power for the overall plant or park, including allocation of the plant reference among the individual units.

[0115] The same principles may also be applied to other groups of electric power producing units that are coordinated to share a total setpoint, including, without limitation, fuel cells, hydropower units, diesel or gas generators, biogas generators, generators powered by synthetic fuels, steam turbines, and combinations thereof.

[0116] The disclosed system and method may be applied to clusters of controllable electrical loads that share a common electrical infrastructure and / or a common grid connection point, including but not limited to electric vehicle charging clusters comprising multiple chargers at a common location. More generally, the disclosed system and method may be used for coordinating any plurality of controllable loads to share a total available power budget or a total setpoint, for example electric vehicles, electrolysers, heat pumps, electric heating elements, water pumps, industrial drives, or other flexible loads.

[0117] The disclosed system and method may be applied to energy storage systems comprising multiple storage subsystems and / or multiple storage units that collectively absorb or inject power according to a common reference. Such storage units may include, without limitation, electro-chemical storage (e.g., batteries), electrical storage (e.g., capacitors), mechanical storage (e.g., pumped hydro storage, compressed air energy storage), or other storage technologies. For example, a large-scale battery system may comprise multiple battery modules, racks, strings, or battery containers that are coordinated to allocate an aggregated charging / discharging setpoint among the individual storage units.

[0118] The disclosed system and method may be applied to microgrids and hybrid energy systems comprising combinations of local generation (e.g., photovoltaic generation, wind generation, diesel gensets), energy storage, and controllable loads within a defined area. Such microgrids may operate grid-connected and / or in islanded mode for a limited or unlimited period of time. The disclosed system and method may provideP7345PC00

[0119] 20

[0120] coordinated allocation of available power and may facilitate compliance with local operational constraints and system limits.

[0121] The disclosed system and method may be applied to building management systems and facility management systems that comprise multiple controllable subsystems and actuators. Non-limiting examples include heating systems, cooling systems, ventilation systems, air handling systems, lighting systems, and integrated energy systems including EV charging and / or storage.

[0122] The disclosed system and method may also be applied to distributed systems comprising multiple controllable units that are not connected behind a single point of common coupling, but rather are distributed across a larger geographical area with separate connection points. Non-limiting examples include so-called “virtual power plants” comprising portfolios of wind turbines, photovoltaic systems, batteries, and / or controllable loads. In such distributed configurations, the controllable units may coordinate to satisfy an aggregated reference value for the portfolio, for example where global consumption and / or production is obtained by summing measured values across multiple sites.

[0123] The disclosed system and method may be applied in electrical distribution networks, for example distribution grid feeders, substations, and transformer-supplied networks in which a plurality of customer installations are connected to shared grid components such as cables and transformers. In such systems, the disclosed system and method may coordinate controllable units (e.g., EV chargers, heat pumps, electric heating, photovoltaic inverters, batteries, and other flexible devices) to share available capacity and to maintain system limits, for example to avoid overload of transformers and cables.

[0124] The disclosed system and method may be applied in demand response and flexibility service systems used by utilities and / or aggregators. Such systems may coordinate large numbers of flexible devices to reduce, shift, or modulate electricity consumption and / or production, for example in response to grid conditions, congestion, or market price signals, thereby contributing to grid stability and cost optimization.P7345PC00

[0125] 21

[0126] The disclosed system and method may be applied to industrial load management systems and industrial process coordination systems, for example in factories comprising flexible and energy-intensive processes such as furnaces, pumps, compressors, and production lines. In such systems, the disclosed system and method may coordinate a plurality of controllable units to reduce or shift power consumption during peak demand periods and / or high price periods while maintaining production targets and / or process constraints.

[0127] The disclosed system and method may further be applied to real-time flow control systems in which a plurality of actuators are controlled in a feedback loop based on local and global measurements. Non-limiting examples include water distribution systems and other fluid networks in which pumps and / or valves are adjusted to maintain system limits and to meet demand. In such applications, the global resource may comprise a flow rate, pressure, or other aggregate system parameter, and the disclosed coordination principles may be used to allocate available capacity among multiple controllable units while adapting to disturbances and changing demands.

[0128] Accordingly, the above application areas are provided as non-limiting examples, and the disclosed system and method may be employed in any system in which coordinated allocation of a shared resource among multiple controllable units is desired.

[0129] In one embodiment, the compensation factors are determined such that the controllable power units adjust their respective local power setpoints to accommodate a change in composition of the cluster, including addition or removal of one or more controllable power units, while maintaining the cluster reference power. In this embodiment, the compensation factors provide a mechanism that allows the cluster to remain stable and to continue fulfilling the cluster reference power even when the set of controllable power units changes over time. Such changes in composition may occur, for example, when a controllable power unit is newly connected to the cluster, becomes available for control, is disconnected from the cluster, becomes unavailable, or is otherwise removed from coordinated operation. When such a change occurs, the compensation factors are determined such that the remaining controllable power units autonomously adjust their respective local power setpoints to re-allocate the available cluster power. ‘Change in composition’ does not have to be construed literally. For example, a charging station for a vehicle will typically be part of the system all the time,P7345PC00

[0130] 22

[0131] but when a vehicle connects to the charging station and starts using it, it may in practice be seen as a new entity.

[0132] In one embodiment, the compensation factors are determined such that the controllable power units adjust their respective local power setpoints to accommodate changes in local power consumption or production driven by changes in local operating conditions. The local constrains may be, for example, that a vehicle starts charging or is fully charged, but it could also mean, in a different system, in which there are photovoltaic units, for example, that a units stops producing power since there is no solar irradiation.

[0133] Accordingly, when a new controllable power unit is added, the compensation factors may cause one or more existing controllable power units to reduce their local power setpoints in order to create available capacity so that the added controllable power unit can be assigned a non-zero local power setpoint without violating the cluster reference power or other cluster constraints. Conversely, when a controllable power unit is removed or becomes idle, the compensation factors may cause one or more remaining controllable power units to increase their local power setpoints in order to utilize the available cluster power and thereby maintain the cluster reference power.

[0134] One embodiment of the presently disclosed method further comprises determining, for each controllable power unit, a balancing term configured to allocate unused cluster power that is not utilized or not produced by the controllable power units.

[0135] In this embodiment, a balancing term (which may also be referred to as an unused-power allocation term) is determined for each controllable power unit in order to enable utilization of cluster power that would otherwise remain unused. Such unused cluster power may arise, for example, when one or more controllable power units are idle, temporarily unavailable, constrained by local limits, or otherwise operate below an allocated base power and therefore do not consume or produce their designated share of power. As a consequence, the aggregated measured power consumption or production of the cluster may fall below what could be achieved under the current cluster reference power, even though additional cluster power is available.P7345PC00

[0136] 23

[0137] The balancing term enables one or more controllable power units to exploit this unused cluster power by increasing their local power setpoints, thereby improving overall utilization of available capacity and facilitating convergence of the cluster power towards the cluster reference power.

[0138] One embodiment of the presently disclosed method further comprises determining, for each controllable power unit, a preservation term configured to target the controllable power unit to maintain a local power setpoint at least at a current power level of the controllable power unit. The preservation term may be used to bias the control of the controllable power unit such that, when possible, the controllable power unit maintains its operation at or above a current power level, for example a presently measured power consumption or power production level, or a power level corresponding to a current or most recent local power setpoint.

[0139] Accordingly, the preservation term provides a mechanism for retaining stable operation and continuity at the level of each controllable power unit while the cluster as a whole continues to adapt dynamically to changes in reference power, available capacity, and cluster composition.

[0140] In one embodiment of the presently disclosed method, the local power setpoint is determined as Psetv= Av+ Bv+ Cv+ Dv., where Avis the base power, Bvis the balancing term, Cyis the preservation term and Dyis the compensation factor.

[0141] In this embodiment, the local power setpoint Psetyfor each controllable power unit is constructed as a combination of multiple contributions, each contribution serving a distinct control purpose. The base power AYrepresents the primary allocated share of the cluster reference power for the controllable power unit, for example determined based on a local priority factor or other allocation policy.

[0142] The balancing term Byenables redistribution of unused cluster power, such that available cluster power that is not utilized or not produced by one or more controllable power units can be allocated to other controllable power units capable of increasing their power consumption or production.P7345PC00

[0143] 24

[0144] The preservation term Cyprovides a power-preservation objective, for example biasing the controllable power unit to maintain a local power setpoint at least at a current power level of the controllable power unit.

[0145] The compensation factor Z)ymay provide a corrective contribution that supports dynamic re-allocation of the available cluster power among the controllable power units, for example to ensure that deviations between allocated power shares and measured power values are compensated and that the cluster converges toward a consensus allocation even with limited communication.

[0146] In one embodiment, the method operates in a multi-phase system. In this embodiment, the step of continuously calculating and updating the local power setpoint comprises calculating and updating the local power setpoint for each of a number of phases of the multi-phase system. Similarly, the global reference power of the cluster may comprise a reference power for each phase, and the global power consumption may comprise a power consumption for each phase.

[0147] In some embodiments, after determining a respective local power setpoint for each phase, a subsequent step may be performed in which one or more actual power setpoints to be applied to the controllable power unit are derived based on the phasespecific local power setpoints. Such a subsequent step may take into account physical or operational characteristics of the controllable power unit. For example, if a controllable power unit is configured to consume or produce substantially equal power on all phases (for example, a symmetric or balanced unit), the actual local power setpoint applied to the controllable power unit may be determined based on a function of the phase-specific setpoints, such as a minimum value, a maximum value, an average value, or another suitable aggregation of the phase-specific setpoints. In this manner, the method may be applied both to controllable power units that can be controlled independently per phase and to controllable power units that operate symmetrically across phases.

[0148] In further embodiments, one or more cluster-level quantities used by the method may be estimated rather than directly measured or explicitly communicated. For example, the global power consumption or production of the cluster, and / or the global reference power of the cluster (including per-phase values in a multi-phase system), may beP7345PC00

[0149] 25

[0150] estimated based on locally available measurements, historical values, inferred information, or indirect observations. Such estimation may be advantageous in systems with limited measurement infrastructure, limited communication bandwidth, or decentralized operation.

[0151] Similarly, other values used in the determination of local power setpoints may be estimated, assumed, or approximated. For example, priority factors associated with other controllable power units in the cluster, and / or an aggregated quantity such as a sum of local priority factors of the cluster, may be estimated or assumed rather than explicitly known. Likewise, power measurements of other controllable power units may be estimated based on partial information, historical behavior, or aggregate measurements, and a local priority factor associated with a controllable power unit may itself be estimated or adapted over time based on observed operating conditions or performance.

[0152] By allowing one or more quantities to be estimated rather than explicitly measured or communicated, the disclosed method enables scalable and robust operation in clusters with limited communication, partial observability, or fully decentralized control. Each controllable power unit may thus determine and update its local power setpoint based on locally available information and estimated global quantities, while the cluster as a whole converges toward a coordinated allocation that satisfies the cluster reference power and other system constraints.

[0153] In one embodiment of the present disclosure, the controllable power units are residential living units configured for consuming, producing, or delivering electrical power. This configuration allows the system to dynamically manage power flows within residential clusters, optimizing energy consumption, production, and distribution based on the specific needs of individual homes or units.

[0154] By applying the presently disclosed method to residential living units as controllable power units, the method enables the integration of various energy assets such as rooftop solar panels, battery storage systems, and home appliances into a cohesive power-sharing network. For instance, surplus energy from one residential unit can be shared with others in the cluster or delivered back to the grid, while during highP7345PC00

[0155] 26

[0156] demand periods, the system can allocate available power to homes with certain needs or prioritize energy-saving measures.

[0157] In one embodiment of the present disclosure, the controllable power units are wind turbines, solar panels, batteries, fuel cells, organic matter fuel burning cells, hydropower units, electrolysis cells, or combinations thereof, configured for producing, delivering, or storing electrical power. This configuration enables the method to integrate and manage a diverse range of energy generation and storage technologies, facilitating efficient and dynamic power-sharing across a heterogeneous cluster.

[0158] By incorporating these various types of controllable power units, the method allows for the effective utilization of renewable energy sources, energy storage devices, and other generation technologies. For example, during periods of high wind or solar availability, wind turbines and solar panels can contribute excess power to the cluster, while batteries can store surplus energy for later use or provide backup during periods of low generation. Similarly, hydropower units and organic matter fuel cells can provide stable power output to complement the variable nature of renewable sources, while electrolysis cells can utilize excess energy to produce hydrogen for long-term storage or other industrial applications.

[0159] The present disclosure also relates to a system comprising the processing circuitry configured to perform the method of the present disclosure. This provides a system equipped to execute the steps of the method, enabling real-time coordination and optimization of power flows within the cluster. The processing circuitry is designed to handle the computational demands of the method, including calculating local power setpoints, updating local priority factors, managing compensation factors, and processing real-time power metrics from the controllable power units. For example, the circuitry can continuously monitor the global cluster power consumption or production, dynamically adjust the reference power, and implement feedback loops to ensure that power-sharing decisions are precise and responsive to changing conditions.

[0160] The present disclosure also relates to a system configured for dynamical power sharing for a cluster of one or more controllable power units, the system comprising: local measuring units configured for measuring a local power consumption or production for each of the controllable power units; local processing circuitry in each of the respectiveP7345PC00

[0161] 27

[0162] controllable power units configured for calculating a local power setpoint for each of the controllable power units, wherein the measuring units and the local processing circuitry are configured for obtaining and continuously updating local power setpoints for each of the respective controllable power units to achieve dynamical power sharing.

[0163] In one embodiment of the present disclosure, the system further comprises a measuring unit configured for measuring a global cluster power consumption or production, or measuring units for each of the controllable power units configured for measuring the global cluster power consumption or production as the sum of the power consumption or production of the individual power units. This configuration allows the system to accurately monitor the overall power metrics within the cluster, providing the foundational data beneficial for effective power management and distribution.

[0164] The measuring unit or units ensure that the system has a real-time view of the cluster’s global power consumption or production, enabling precise and dynamic coordination among the controllable power units. For example, in a cluster comprising renewable energy sources and energy storage devices, the system can use the measured global power metrics to balance supply and demand, adjust power setpoints for individual units, and maintain overall stability. By incorporating measurements from individual power units, the system can account for localized variations in consumption or production, ensuring that the aggregated data reflects the cluster’s true operating conditions.

[0165] This capability is particularly beneficial in complex energy systems, such as microgrids, industrial power setups, or distributed renewable energy networks, where maintaining a comprehensive understanding of global power metrics can be important for stability and efficiency.

[0166] In one embodiment of the present disclosure, the system further comprises global processing circuitry configured for receiving and broadcasting the cluster power consumption or production and a cluster reference power. This configuration enables centralized coordination and communication of power metrics within the cluster, ensuring that all controllable power units operate in alignment with the system’s objectives.P7345PC00

[0167] 28

[0168] The global processing circuitry collects data on the cluster's power consumption or production from the measuring units and broadcasts some of or all of this information, along with the cluster reference power, to some of or all controllable power units in real time. This centralized processing capability ensures that each power unit has up-to-date information about the cluster's current status and target power levels, allowing them to dynamically adjust their power setpoints accordingly. For example, in a renewable energy grid, the global processing circuitry can continuously update and share the cluster reference power based on fluctuations in solar or wind generation, ensuring that the entire cluster operates efficiently and remains balanced.

[0169] In one embodiment of the present disclosure, the system comprises a combination of at least one controllable power unit and at least one un-controllable power unit or at least one power unit that cannot adapt its / their power consumption or production. This configuration enables the system to effectively manage power-sharing dynamics within a mixed cluster of adaptive and non-adaptive units, ensuring stability and efficiency despite the presence of units with fixed behavior.

[0170] By incorporating both controllable and uncontrollable power units, the system ensures that power-sharing decisions account for the static contributions or demands of the non-adaptive units. For example, in a hybrid grid, uncontrollable power units such as conventional generators or fixed-output renewable sources can provide baseline power, while controllable units such as batteries or adjustable loads dynamically respond to variations in demand or generation. The system coordinates the controllable units to maintain balance within the cluster, compensating for the fixed nature of the uncontrollable units and ensuring optimal utilization of available resources.

[0171] The present disclosure also relates to any combination of any of the embodiments relating to both method and system described herein.

[0172] Reference numeral list

[0173] 100 system for dynamic power sharing

[0174] 101 measurement of the global power consumption at the point of common coupling (PCC)

[0175] 102 cluster reference power (Pref)

[0176] 103 power consumption or production of the cluster (Pmeas)P7345PC00

[0177] 29

[0178] 104 external system such as an electric power system

[0179] 105 electric wires / circuitry connecting units of the system

[0180] 106 first controllable power unit

[0181] 107 second controllable power unit

[0182] 108 last controllable power unit

[0183] 109 power measurement of first controllable power unit

[0184] 110 power measurement of second controllable power unit

[0185] 111 power measurement of last controllable power unit

[0186] 112 controller for first controllable power unit

[0187] 113 controller for second controllable power unit

[0188] 114 controller for last controllable power unit

[0189] 115 broadcasting cluster reference power to the controllable power units 116 broadcasting the power consumption or production of the cluster to the controllable power units

[0190] 200 feedback loop of a controllable power unit connected to a system for dynamic power sharing

[0191] 201 controller

[0192] 202 controllable power unit

[0193] 203 measurement of power consumption or production of controllable power unit

[0194] 204 local priority factor

[0195] 205 power setpoint

[0196] 206 power consumption or production of the cluster

[0197] 207 cluster reference power

[0198] 208 compensation factor

[0199] 209 power consumption or production of controllable power unit

[0200] 210 connection point to electrical system

[0201] 300 flowchart showing an example of a possible implementation of the method of the present disclosure

[0202] 301 obtaining the global reference power and the global power consumption or production of the cluster

[0203] 302 obtaining the local priority factor and the local compensation factor for the controllable power unit

[0204] 303 obtaining the local power consumption or production of the controllable power unitP7345PC00

[0205] 30

[0206] 304 calculate and update local power setpoint of the controllable power unit 305 feedback loop step checking whether the production or consumption is complete and reiterating the procedure if not

[0207] Examples

[0208] This section provides details on several non-limiting examples and embodiments of the present disclosure. The following examples focus on the case of power consumption for convenience, but the principles and formalism of the present examples apply mutatis mutandis to the case of power production. In the case of production, the algorithm still applies as formulated below, as long as the formulation is consistently applied to all the units of the cluster. In the power profiles, power consumption is, by convention, characterized by positive values and power production is characterized by negative values.

[0209] General details:

[0210] When allocating the cluster power amongst each other, the changes should respect the threshold Pref, and at the same time make full use of the available power, i.e., reducing the difference between Pmeasand Pref. Hence, every controllable power unit may run an algorithm that calculates its own power setpoint by taking into account the global variables Pmeasand Pref. In addition, each controllable power unit may have local measurements, i.e., they each know their local power consumption. The power setpoint Pset,yof a controllable power unit y, where y runs from 1 to N and N is the numbers of controllable power unit in the cluster, is calculated based on the following control logic comprising four terms:

[0211] Pset,y=Ay + By + Cy + Dy.

[0212] The terms are defined as follows:

[0213] Ay is the base power allocated for each controllable power unit. The sum of all base powers in the cluster must be smaller than or equal to the cluster reference power. This condition can be expressed as Pref> Sy=i Ay.P7345PC00

[0214] 31

[0215] Byexploits available cluster power that is not utilized or not produced by other controllable power units. This happens if one of more controllable power units are idle, or if one or more controllable power units do not utilize their designated base power.

[0216] Cyserves as an offset / storage that preserves differences of the setpoint and Ay. It enables the exploitation of unused cluster power; the utilized power capacity is added to the base power Ay.

[0217] Dy is a compensation term that reduces the power setpoint to make space for other controllable power units. It effectively counteracts Cyand can be activated continuously or event / trigger based.

[0218] Further details:

[0219] In one embodiment of the control concept, the following demonstrates how the base elements Ay, By, Cy, and Dycan be implemented. This embodiment focuses on the case of power consumption for convenience.

[0220] In the present case, the base power Aywill be defined as a share of the cluster reference power Pref, defined via a normalized local priority factor Ayas Ay= Ay• Pref, where Aye [0,1], The present disclosure also comprises implementations in which the local priority factors may be outside this range, but the present example focuses on normalized priority factors in this range for concreteness and to ensure stable operation. For instance, to respect the cluster limit and meet the condition Pref> 2y=i Ay, the normalized local priority factors of all units should satisfy the following normalization condition: Jy=1Ay1- If one or more controllable power units are idle, or if one or more controllable power unit do not fully utilize the designated base power, active controllable power units can utilize this power. This is captured by the term Bywhich exploits the difference between cluster reference and measured power based on the normalized local priority factor: By= Ay• (Pref- Pmeas).

[0221] To preserve this utilized power, the term Cyserves as an offset of the current power setpoint with respect to the base power, taking into account the actual power consumption of each controllable power unit, Pconsy, ■-©- >cy= Peons, y>■ Pref- IfaunitP7345PC00

[0222] 32

[0223] exploits unutilized cluster power, the value of Bydecreases over time, while Cyincreases. Effectively, this means that the value of Byis transferred to Cy.

[0224] Finally, if a controllable power unit utilizes power above or below its base power, the unit should be able to make space for other controllable power units that so far have not utilized their designated power value. This may happen in case of new power requests, or if the power capability of one or more controllable power units changes due to changes in external parameters such as for example temperature, or conditions affecting delivery or extraction of power and services related to provisions to grid operations. This mechanism of making space can be accomplished by the compensation term Dy, which partially or fully contracts Cy.

[0225] In the present case, Dyis implemented as a compensation term that subtracts a proportional share of the measured consumption above Ay. The compensation term comprises a compensation factor, ^y, such as Dy= ^y• (Ay• Pref- Pcons,y). The compensation term Dycan either be continuously applied, or activated by time-based or event-based triggers. In this context, the compensation factor ^yallows for tuning the response Dy.

[0226] One possible principle for deriving the normalized local priority factor Ayis an urgencybased evaluation system, where each controllable power unit estimates its average power needed to achieve a given energy goal. In this case, we can calculate the local priority factors Ayas the required power rates Ay= AE / At, where AE is the remaining energy to fulfil a given energy goal and At is the remaining time to achieve the goal. The normalized local priority factors can now be calculated to saturate the above normalization condition as Ay= Ay / £y=1Ay, where £y=1Ayis the sum of the local priority factors of all individual controllable power units. In the following, we will refer to Ayas the “local priority factor”, while Ayis defined as the “normalized local priority factor”.

[0227] There are various options for how to achieve this value. Firstly, the local priority factors Ayof the individual power units can be broadcasted, alongside the global quantities Pref and Pmeas. Secondly, the individual priorities can be collected in a database, and only the final sum £y=1Aymay be broadcasted. While both these approaches mayP7345PC00

[0228] 33

[0229] require communication, the update rate can be less frequent than the interval in which the control script of each charger is executed. Thirdly, the sum £y=1Aycan also be estimated using the installed capacity of the controllable power units in the cluster as individual Ay. For instance, if a cluster has 10 controllable power units and power ratings of 22 kW each, the sum Sy=i Aycan be considered to be 220 kW as a worstcase assumption. As such, no communication besides Prefand Pmeasis required, while the condition Jy=1Ay1, is fulfilled. Due to the interrelation of Ay, By, Cy, and Dy, the controllable power units still collectively coordinate their power consumption although no communication of the individual urgency / priority values is required.

[0230] Another approach for sharing the available power Prefis to apply equal sharing, where all priorities are set to the same value. This value can, for instance, be chosen as Ay= 1 / N, which directly fulfils the condition Jy=1Ay1-

[0231] Finally, with the exemplary formulations of Ay, By, Cy, and Dygiven above, the equation for the charger setpoint Pset,Ycan be simplified to:

[0232] "

[0233]

[0234] The above expression shows how the power setpoint of the individual controllable power units can be uniquely determined by the local power consumption, the global cluster reference power, the measured cluster power consumption, as well as the local priority factor and the local compensation factor. A central point is, however, that the chargers collectively respect the cluster reference power without exchanging any data on their power consumption. The only quantities that need to be communicated to all controllable power units are the cluster reference power Prefand the measured cluster consumption Pmeas. Hence, there is no single entity providing setpoints to each controllable power unit since the decision-making and consensus reaching is done by the individual controllable power units.

[0235] Flowchart descriptionP7345PC00

[0236] 34

[0237] An example of a flowchart illustrating one possible embodiment of the presently disclosed method is shown in FIG. 3. The flowchart (300) contains procedural steps to be implemented by each of the controllable power units during dynamic power sharing. The first three steps (301-303) concern obtaining input variables for the method, i.e., the global reference power, the global power consumption of the cluster, the local priority factor, the local compensation factor, and the local power consumption of the controllable power unit. These steps of obtaining input variables can also be taken in a different order than that illustrated in FIG. 3. For instance, the first three steps (301-303) can also take place in parallel rather than sequential. In case the local priority factors are communicated between the controllable power units, the flowchart may contain an additional step concerning this communication.

[0238] The fourth step (304) concerns calculating and updating the local power setpoint of the controllable power unit based on the provided input variables of the preceding steps and as exemplified in the above equations. In one variation of the presently disclosed method, there may be an additional step after the calculation of the power setpoint concerning forcing the power setpoint to stay below a maximum power setpoint and / or above a minimum power setpoint.

[0239] The fifth step (305) concerns checking whether production of consumption of power is done. If the production or consumption is done, the method can be truncated for the relevant controllable power unit. A reason for being done may for instance be that a power charging cycle is complete. If the production or consumption is not done, one returns to the procedural steps of obtaining input variables, and as such the method implements a feedback control loop. The frequency of the power setpoint updates can be arbitrary and range from discrete time intervals to infinitesimally small time intervals.

[0240] The flowchart captures the steps relevant for calculating and updating the power setpoint using real-time control. In some variations of the presently disclosed method, additional communication steps for other purposes, such as communicating logged data at discrete intervals, can take place outside the flowchart of FIG. 3.

[0241] Simulations of power profiles

[0242] In the following, several examples of an embodiment of the presently disclosed method are provided. Specifically, the simulations described in the following section focus onP7345PC00

[0243] 35

[0244] an embodiment in which the controllable power units are charging stations configured for charging electrical vehicles. These examples are further illustrated in FIG. 4 to FIG.

[0245] 10.

[0246] A first example / simulation is provided in FIG 4. Here, the cluster comprises two power units, each with a power capability of up to 20kW. For unit 1, the local priority factor Ayis initialized to 0 and configured to step to 20 at t = 0 (t being time measured in seconds). From t = 100s and onwards, the local priority factor of unit 1 is set to 0. For unit 2, the local priority factor is initialized to 0 and configured to step to 20 at t = 10s. The reference power reference power for the cluster is fixed to 20 kW and kept constant. The compensation factors of both units are constant at ^y= 0.1. The control algorithm in each unit applies the normalized local priority factor Ay, which is calculated based on the local priority factor Ayof unit y, divided by the sum of all local priority factors £y=1Ayas Ay= Ay / £y=1Ay, where N = 2. The power setpoint for each unit is calculated and updated every 0.1s.

[0247] At t = 0s, the local priority factor of unit 1 becomes greater than 0. Consequently, it increases the power, since the cluster power is below the reference value of 20 kW. Since unit 1 has a power capability of 20 kW, it can fully meet the cluster reference power of 20 kW, since it is the only active unit. Between t = 0s and t = 10s, unit 2 has a local priority factor of 0 and, therefore, stays inactive. At t = 10s, unit 2 increases its priority factor to 20. Since now both units have the same local priority factor, they will aim at sharing the cluster reference power equally. As seen in FIG. 4, unit 1 is reducing its power to make space for unit 2. Eventually, they reach the same power of 10 kW each. During this process, the cluster power is constant and equal to the cluster reference power. At t = 100s, the priority of unit 1 is set to 0, after which the respective unit reduces its power to 0 kW. Simultaneously, unit 2 increases its power to ensure that the cluster power meets the cluster reference power. Eventually, unit 2 has a power of 20 kW. During this transition, the cluster power is constant and equal to the cluster reference power.

[0248] A second example / simulation is provided in FIG. 5. Here, the cluster comprises two power units, wherein the compensation factors of both units are constant at ^y= 0.01. The parameters and procedure are otherwise identical to those of FIG. 4. The effect of the smaller compensation factor is that the re-allocation of the cluster power betweenP7345PC00

[0249] 36

[0250] unit 1 and unit 2 after the time instances at t = 10s and t = 100s happens in smaller steps compared to FIG. 4, resulting in slower re-allocation processes. Yet, throughout the dynamic processes, the cluster power is constant and equal to the cluster reference power.

[0251] A third example / simulation is provided in FIG. 6. Here, the cluster comprises two power units, wherein the power of unit 1 is suddenly reduced to 0 at t = 100s. The parameters and procedure are otherwise identical to those of FIG. 5. Potential causes of such a sudden drop in power can be if a unit suddenly disconnects from the electrical system, or if the calculated setpoint for the respective unit is overwritten / overruled, e.g. through protection equipment. While in FIG. 4 and FIG. 5 the power of unit 1 was gradually decreased and the power of unit 2 gradually increased, in FIG. 6 the sudden power drop of unit 1 results in a short dip of the cluster power to 10 kW. Subsequently, unit 2 rapidly increases its power to 20 kW. Hence, the cluster power is again constant and equal to the cluster reference power.

[0252] A fourth example / simulation is provided in FIG. 7. Here, the cluster comprises two power units, wherein the units do not communicate their local priority factors with each other. The parameters are otherwise identical to those of FIG. 5. Here, the control algorithm in each unit applies the normalized local priority factor Ay, which is calculated based on the local priority factor Ayof unit y, divided by the assumed sum of all local priority factors £y=1Ay. Since the local priority factors of the other units are unknown, the sum is calculated based on the power ratings of all units. Applied to the present example, this means that £y=1Ay= 40 kW. Other variations of this example comprise making the normalizing factor change in time by for example using forecasts.

[0253] In the present example, the normalized local priority factors are lower than in FIG. 5 during the intervals between t = 0s and t = 10s, as well as t > 100s since the denominator is larger. It is based on a fixed value (40 kW) and not dynamically adjusted as in the case where the other local priority factors are communicated. A consequence is of this is that the cluster power is not fully utilized. In the present simulation, however, the cluster power in steady state at t = 150s is 19.804 kW, which corresponds to 99.02% of the cluster reference power. Despite the priorities not being shared among the units, the units still share the available cluster power according to their priorities.P7345PC00

[0254] 37

[0255] A fifth example / simulation is provided in FIG. 8. Here, the cluster comprises two power units, wherein power setpoints for each unit is calculated and updated every 0.15s. The parameters and procedure are otherwise identical to those of FIG. 7. The difference from FIG. 7 is thus that the power setpoints are updated every 0.15s and not every 0.1s. This is reflected in the time it takes to re-allocate the cluster power after the changes in priority at t = 10s and t = 100s, respectively (50% longer compared to FIG.

[0256] 7). Like in FIG. 7, the units still share the available cluster power according to their priorities, despite the priorities not being shared among the units.

[0257] A sixth example / simulation is provided in FIG. 9. Here, the cluster comprises two power units, wherein the local priority factor of unit 1 is set to 10 at t = 0s. The parameters and procedure are otherwise identical to those of FIG. 7. The difference from FIG. 7 is that at t = 0s, the local priority factor of unit 1 is set to 10, while in FIG. 7 the local priority factor of unit 1 was set to 20. The differences in the local priority factors of unit 1 and unit 2 causes the units to share the cluster power accordingly. Since unit 2 has a local priority factor with twice the value of unit 1 , it has a power twice as high compared to unit 1. Hence, the units share the cluster power in ratios of 1 / 3 (unit 1) and 2 / 3 (unit 2) of the cluster reference power. Once the local priority factor of unit 1 is reduced to 0, it reduces the power, while unit 2 gradually increases its power. Despite the priorities not being shared among the units, the units still share the available cluster power according to their priorities.

[0258] A seventh example / simulation is proved in FIG. 10. In this simulation, the cluster comprises three power units, wherein the local priority factors of units 1, 2, and 3 turn on from zero to finite values in a step-wise manner at times t = 0s, t = 10s, and t = 100s, respectively. The reference power for the cluster changes throughout the simulation. Initially, the value is set to 20 kW (consumption); at t = 80s, the reference power is set to -20 kW (production), and at t = 120 s, the reference power is increased with a rate of 10 kW / s until it reaches 20 kW at t = 124s from which it is kept constant. The three power units are specified as follows: unit 1 is a uni-directional unit (can only consume power) and has a maximal power rating of 20 kW, unit 2 is a bi-directional unit (can consume and inject power, such as a battery) and has a maximal power rating of 20 kW, and unit 3 is a bi-directional unit and has a maximal power rating of 5 kW. The compensation factors of all three units are constant at = 0.01. The unitsP7345PC00

[0259] 38

[0260] communicate their local priority factor with each other. The control algorithm in each unit applies the normalized local priority factor Ay, which is calculated based on the local priority factor Ayof unit y, divided by the sum of all local priority factors: Ay= Ay / £y=iAywith N = 3. The setpoint for each unit is calculated and updated every 0.10s.

[0261] In the present example, the following sequence of events take place. At t = Os, the local priority factor of unit 1 is set to 10. Since this is the only active unit at this time, it starts consuming 20 kW at this time, which corresponds to the reference power of the cluster. At t = 10s, the local priority factor of unit 2 is set to 20. In response, unit 1 decreases its power consumption to make space for unit 2, which can ramp up its power consumption. The normalized local priority factors of unit 1 and unit 2 are 1 / 3 and 2 / 3, respectively. Hence, the power consumption of unit 1 and 2 will converge towards these fractions (6.67 kW and 13.33 kW, respectively) of the available cluster power. At t = 100s, the local priority factor of unit 3 is set to 20. Since units 2 and 3 have the same local priority factors, they would share the cluster reference evenly if they had the same power rating. However, unit 3 has a maximum power of 5 kW. Hence, unit 2 and unit 3 inject 15 kW and 5 kW, respectively.

[0262] Finally, at t = 120s, the reference power is increased at a rate of 10kW / s from -20kW to +20kW. Since the power is positive (consumption), all three units can contribute to fulfilling the cluster reference power. Given their local priority factors of 10, 20, and 20, the normalized local priority factors are 1 / 5 (unit 1), 2 / 5 (unit 2) and 2 / 5 (unit 3). If the three units had sufficient power ratings, they would therefore share the cluster reference power with 4 kW (unit 1), 8 kW (unit 2), and 8 kW (unit 3). However, the power rating of unit 3 is limited to 5 kW. As a result, unit 3 consumes 5 kW. The remaining 15 kWare shared among units 1 and 2, with relative shares corresponding to the ratio of their priorities. Since the local priority factor of unit 2 is twice as high as the local priority factor of unit 1 , it will consume twice the power. Hence, unit 2 consumes 10 kW, while unit 1 consumes 5 kW. Throughout the re-allocation process of the cluster power, e.g., during and after the increase of the reference value from -20 kW to + 20kW, the units manage to continuously fulfil the cluster reference power as seen by the matching curves of cluster power and cluster reference power.P7345PC00

[0263] 39

[0264] Items

[0265] 1. An autonomous decision-making method for dynamical power sharing for a cluster of one or more controllable power units, comprising:

[0266] obtaining for each controllable power unit a local power setpoint using a global reference power of the cluster, and a global power consumption or production of the cluster;

[0267] continuously calculating and updating the power setpoint for each controllable power unit using a feedback loop using the global reference power of the cluster, the global power consumption or production of the cluster, a local priority factor for each controllable power unit, and a local compensation factor for each controllable power unit.

[0268] 2. The method according to item 1 , wherein at least two of the controllable power units are connected through a point of common coupling.

[0269] 3. The method according to any one of the preceding items, wherein the compensation factors are defined to subtract or add a share of the measured power consumption or production above or below a base power, wherein the base power is defined as the cluster reference power weighted by the associated local priority factor.

[0270] 4. The method according to any one of the preceding items, wherein the compensation factors are proportional such that they subtract or add a proportional share of the measured power consumption or production above or below the base power for each controllable power unit.

[0271] 5. The method according to item 3 or 4, wherein the compensation factors are defined to subtract or add a share of the previous power setpoint of the associated controllable power unit.

[0272] 6. The method according to any one of the preceding items, wherein the compensation factors are trigger-based such that they after defined events subtract or add a fixed or proportional share of the global power consumption or production above or below the base power for one or more of the controllable power units.P7345PC00

[0273] 40

[0274] 7. The method according to any one of the preceding items, further comprising communication of the local priority factors or their sum between the controllable power units.

[0275] 8. The method according to any one of the preceding items, further comprising dynamically updating at least one of the local priority factors.

[0276] 9. The method according to any one of the preceding items, further comprising using a sum of all local priority factors, each defining the allocated share of the cluster reference power for the individual electrical power units, to respect a cluster power consumption or production limit.

[0277] 10. The method according to item any one of the preceding items, wherein all the local priority factors are fixed to be the value reciprocal to the number of controllable power units for each of the controllable power units.

[0278] 11. The method according to any one of the preceding items, wherein at least one local priority factor is calculated using technical specifications of an incoming unit upon connecting to a controllable power unit, or using electrically linked units, or using a set of sub-units, or using one or more user requests, or using power ratings of units, or a combination thereof.

[0279] 12. The method according to any one of the preceding items, further comprising a subsequent step after the power setpoint for each of the controllable power units have been calculated, the additional step involving forcing the power setpoint to stay below a maximum power setpoint and / or above a minimum power setpoint.

[0280] 13. The method according to any one of the preceding items, wherein the controllable power units are configured to consume or produce at least a predetermined fraction of the available cluster power.

[0281] 14. The method according to any one of the preceding items, wherein the global cluster reference power is dynamically updated in response to conditionsP7345PC00

[0282] 41

[0283] affecting delivery or extraction of power, for instance controlled by availability of local renewable energy sources, peak shaving, participation in aggregated portfolios, electricity prices, services related to provisions to grid operations, or combinations thereof.

[0284] 15. The method according to any one of the preceding items, further comprising broadcasting the global cluster reference power and the global cluster power consumption or production to the individual controllable electrical power units using a communication system.

[0285] 16. The method according to any one of the preceding items, wherein the local priority factors are broadcasted using a communication system.

[0286] 17. The method according to item 15 or 16, wherein the communication system is Wi-Fi, Ethernet, TCP / IP, Modbus, MQTT, OPC UA, Profibus, Profinet, CAN bus, BACnet, Zigbee, LoRaWAN, HART, RS-232, RS-485, DeviceNet, EtherCAT, AS-lnterface, lO-Link, Bluetooth, Z-Wave, DNP3, KNX, EnOcean, Fieldbus, LonWorks, Serial Modem, ISM band communication, M-Bus, Power Line Communication (PLC), Thread, 6L0WPAN, Sigfox, Cellular (3G, 4G, 5G), SCADA, or combinations thereof.

[0287] 18. The method according to any one of the preceding items, wherein the global power consumption or production is obtained as the sum of the measured power consumption or production from each of the controllable power units and wherein the measured power consumption or production of each controllable power units or their sum is communicated to all the controllable power units.

[0288] 19. The method according to any one of the preceding items, configured for systems comprising a combination of at least one controllable power unit and at least one un-controllable power unit, wherein the method is applied to a subset of the controllable power units.

[0289] 20. The method according to any one of the preceding items, wherein the method is configured for dynamical current sharing instead of dynamical power sharing.P7345PC00

[0290] 42

[0291] 21. The method according to any one of the preceding items, wherein the controllable power units are charging stations configured for charging electrical vehicles.

[0292] 22. The method according to item 21, further comprising using the local priority factors to set an urgency of a power consumption or production request from the one or more charging stations.

[0293] 23. The method according to any one of the preceding items 1-20, wherein the controllable power units are residential living units configured for consuming, producing, or delivering electrical power.

[0294] 24. The method according to any one of the preceding items, wherein the controllable power units are wind turbines, solar panels, batteries, organic matter fuel burning cells, hydropower units, electrolysis cells, or combinations thereof, configured for producing, delivering, or storing electrical power.

[0295] 25. A system comprising processing circuitry configured to perform the method according to any one of the preceding items.

[0296] 26. A system configured for dynamical power sharing for a cluster of one or more controllable power units, the system comprising:

[0297] local measuring units configured for measuring a local power consumption or production for each of the controllable power units; local processing circuitry in each of the respective controllable power units configured for calculating a local power setpoint for each of the controllable power units, wherein the measuring units and the local processing circuitry are configured for obtaining and continuously updating local power setpoints for each of the respective controllable power units to achieve dynamical power sharing.

[0298] 27. The system according to item 26, further comprising a measuring unit configured for measuring a global cluster power consumption or production, or measuring units for each of the controllable power units configured for measuring the global cluster power consumption or production as the sum ofP7345PC00

[0299] 43

[0300] the power consumption or production of the individual power units.

[0301] 28. The system according to any one of the preceding items 26-27, further comprising global processing circuitry configured for receiving and broadcasting the cluster power consumption or production and a cluster reference power.

[0302] 29. The system according to any one of the preceding items 26-28, comprising a combination of at least one controllable power unit and at least one uncontrollable power unit or at least one power unit that can not adapt its / their power consumption or production.

Claims

P7345PC0044Claims1. An autonomous decision-making method for dynamical power sharing for a cluster of one or more controllable power units, comprising:obtaining for each controllable power unit a local power setpoint using a global reference power of the cluster, and a global power consumption or production of the cluster;continuously calculating and updating the local power setpoint for each controllable power unit using a feedback loop using the global reference power of the cluster, the global power consumption or production of the cluster, a local priority factor for each controllable power unit, and a local compensation factor for each controllable power unit.

2. The method according to claim 1 , wherein at least two of the controllable power units are connected through a point of common coupling.

3. The method according to any one of the preceding claims, wherein the local priority factor for each controllable power unit defines a respective proportional share of the global reference power of the cluster allocated to said controllable power unit.

4. The method according to any one of the preceding claims, wherein the compensation factors are defined to subtract or add a share of the global power consumption or production above or below a base power, wherein the base power is defined as the cluster reference power weighted by the local priority factor.

5. The method according to any one of the preceding claims, wherein the compensation factors are proportional such that they subtract or add a proportional share of the global power consumption or production above or below the base power for each controllable power unit.

6. The method according to any one of the preceding claims, wherein the compensation factors are determined such that the controllable power units adjust their respective local power setpoints to accommodate a change in composition of the cluster, while maintaining the cluster reference power.P7345PC00457. The method according to any one of the preceding claims, wherein the compensation factors are determined such that the controllable power units adjust their respective local power setpoints to accommodate changes in local power consumption or production driven by changes in local operating conditions.

8. The method according to any one of the preceding claims, wherein the compensation factors are determined and applied continuously over time, or in response to a triggering event9. The method according to any one of the preceding claims, further comprising determining, for each controllable power unit, a balancing term configured to allocate unused cluster power that is not utilized or not produced by the controllable power units.

10. The method according to any one of the preceding claims, further comprising determining, for each controllable power unit, a preservation term configured to target the controllable power unit to maintain a local power setpoint at least at a current power level of the controllable power unit.11 . The method according to claim 2, 9 and 10, wherein the local power setpoint is determined as Psetv= Av+ Bv+ Cv+ Dv., where Avis the base power, Bvis the balancing term, Cyis the preservation term and Dyis the compensation factor.

12. The method according to any one of the preceding claims, used in a multiphase system, comprising the step of continuously calculating and updating the local power setpoint for each of a number of phases of the multi-phase system, wherein the global reference power of the cluster comprises a reference power for each phase, and the global power consumption comprises a power consumption for each phase.

13. The method according to any one of the preceding claims, wherein at least one local priority factor is calculated using technical specifications of an incoming power unit upon connecting to a controllable power unit, or using electrically linked units, or using a set of sub-units, or using one or more user requests, orP7345PC0046using power ratings of units, or a combination thereof.

14. The method according to any one of the preceding claims, further comprising a subsequent step after the power setpoint for each of the controllable power units have been calculated, the additional step involving forcing the power setpoint to stay below a maximum power setpoint and / or above a minimum power setpoint.

15. The method according to any one of the preceding claims, further comprising communication of the local priority factors or their sum between the controllable power units.

16. The method according to any one of the preceding claims, further comprising using a sum of all local priority factors, each defining the allocated share of the cluster reference power for the individual electrical power units, to respect a cluster power consumption or production limit.

17. The method according to any one of the preceding claims, wherein the global cluster reference power is dynamically updated in response to conditions affecting delivery or extraction of power, for instance controlled by availability of local renewable energy sources, peak shaving, participation in aggregated portfolios, electricity prices, services related to provisions to grid operations, or combinations thereof.

18. The method according to any one of the preceding claims, further comprising broadcasting the global cluster reference power and the global cluster power consumption or production to the individual controllable electrical power units using a communication system.

19. The method according to any one of the preceding claims, wherein the local priority factors are broadcasted using a communication system.

20. The method according to any one of the preceding claims, wherein the global power consumption or production is obtained as the sum of the power consumption or production from each of the controllable power units andP7345PC0047wherein the power consumption or production of each controllable power units or their sum is communicated to all the controllable power units.

21. The method according to any one of the preceding claims, wherein the controllable power units are charging stations configured for charging electrical vehicles.

22. The method according to claim 21, further comprising using the local priority factors to set an urgency of a power consumption or production request from the one or more charging stations.

23. The method according to any one of the preceding claims, wherein the controllable power units are wind turbines, solar panels, batteries, fuel cells, hydropower units, electrolysis cells, or combinations thereof, configured for producing, delivering, or storing electrical power.

24. A system comprising processing circuitry configured to perform the method according to any one of the preceding claims.

25. A system configured for dynamical power sharing for a cluster of one or more controllable power units, the system comprising:local measuring units configured for measuring a local power consumption or production for each of the controllable power units; local processing circuitry in each of the respective controllable power units configured for calculating a local power setpoint for each of the controllable power units, wherein the measuring units and the local processing circuitry are configured for obtaining and continuously updating the local power setpoints for each of the respective controllable power units to achieve dynamical power sharing.

Images