Calculation unit and method for generating a lookup table and power distribution device using a lookup table

Abstract

Calculation unit and method for generating a lookup table and power allocation device using a lookup table. This description relates to a method for generating a lookup table comprising allocation factors (α1, …, αk) of a total power to be delivered to an electrical supply network or of a total power to be received from the network, the method comprising: - receiving a past evolution of a signal (N) from the network over a period (T); - receiving a consumption model for each of a plurality of devices in a system; and - generating, by the calculation device, the lookup table by calculating, for each of the devices and for a plurality of states of the system, the allocation factors (α1, …, αk) using a predictive model control algorithm based on the consumption model of each of the devices and the past evolution of the signal (N) over the period.Figure for the summary: Fig. 3.

Description

Title of the invention: Calculation unit and method for generating a lookup table and power distribution device using a lookup table. Technical field

[0001] This description relates generally to the field of electronic devices and systems for energy management in electrical power supply networks. Previous technique

[0002] A recurring difficulty faced by power supply networks is to balance supply and demand at all times in order to avoid overloading the transmission network, which could lead to general power outages.

[0003] The system frequency is the most important indicator of an instantaneous power imbalance on the network. Indeed, an increase in consumption causes an increase in power demand on synchronous generating machines and thus causes a slowing of their rotational speed. Conversely, a production surplus, and therefore an increase in frequency, will lead to a reduction in the instantaneous power requirement.

[0004] In Europe, it is the role of the transmission system operator (TSO) to ensure network stability by organizing and supervising actions and mechanisms associated with frequency control. Other countries have entities that play a similar role. In relatively large transmission networks, there are generally three frequency control layers used to address frequency deviations: primary, secondary, and tertiary. Primary Frequency Control (PFC), also known as Frequency Containment Reserve (FCR), is predominantly ensured by the static capability of large generators.Secondary frequency control, or automatic frequency restoration reserve (aFRR), provides backup to primary frequency control in maintaining the network frequency. The secondary reserve, for example, has a longer response time compared to the primary reserve. However, the advent of demand-response (DR) technology in the development of today's smart grids has made the requesting site a new player in the stability of a power supply network. Indeed, industrial sites often want... accept a certain degree of flexibility in their power consumption in exchange for financial incentives such as reduced energy costs.

[0005] A set of devices, for example present in different geographical sites, is grouped for a common management of a power delivered or received by each of the devices.

[0006] There is a need for systems to optimize response dynamics and / or energy storage management of devices and there is a need to improve interactions between devices in the power supply network to save energy. Summary of the invention

[0007] One embodiment provides a method for generating a lookup table comprising distribution factors for the total power to be delivered to an electrical supply network or the total power to be received from the network, the method comprising: - the reception, by a computing device, of a past evolution of a network signal, the signal being a control signal of a power to be delivered to the network or received from the network, over a period; - the reception, by the computing device, of a consumption model for each of a plurality of devices in a system, each device being configured to receive and / or generate a partial power; - the generation, by the computing device, of the lookup table by calculating, for each device and for a plurality of system states, the distribution factors using a predictive model control algorithm based on the consumption model of each device and the past evolution of the signal over the period; and - the transmission, by the calculation device, of the correspondence table to a system control device configured to control the partial power delivered to or received from the network by each of the system devices.

[0008] According to one embodiment, the method further includes the control, by the system control device, of the power delivered to the network or received from the network by each of the system devices, using the distribution table.

[0009] According to one embodiment, a system state comprises a signal value and a state of each of the system devices, the state of each system device being: - a state of charge if the device is of a first type configured to deliver and receive power; - the power delivered if the device is of a second type configured to deliver power; or - power received if the device is of a third type configured to receive power.

[0010] According to one embodiment, the calculation of the correspondence table includes the generation of a database, the generation of the database including, for a plurality of times in the period, the calculation, for each of the devices, for a state of the system and at a time in the plurality of times, of the distribution factors using a predictive model control algorithm based on the consumption model of each of the devices.

[0011] According to one embodiment, the generation of the database further includes, for a plurality of times in the period, the generation of a forecast of an evolution of the signal over a time interval following a time.

[0012] According to one embodiment, the calculation of the correspondence table further includes a statistical analysis of the database.

[0013] According to one embodiment, the lookup table is configured to associate the distribution factors with a value of the system state.

[0014] According to one embodiment, the consumption model of each device includes a relationship between a power delivered or received by the device and a consumption or a production cost.

[0015] Another embodiment provides a computing unit configured to generate a lookup table comprising allocation factors for the total power to be delivered to an electrical supply network or the total power to be received from the network, the computing device being configured to: - receive a past evolution of a network signal, the signal being a function of a frequency difference of the network, over a period; - receive a consumption model for each of a plurality of devices in a system, each device being configured to receive and / or generate a partial power; - generate the lookup table by calculating, for each device and for a plurality of system states, the distribution factors using a predictive model control algorithm based on the consumption model of each device and the past evolution of the signal over the period; and - transmit the lookup table to a system control device configured to control the partial power delivered to or received from the network by each of the system devices.

[0016] Another embodiment provides a system comprising: - the above-mentioned calculation unit; and - the system control device configured for: - receive a signal from the network, the signal being a function of a frequency difference of the network; - receive a measurement of the state of each of the system's devices; - establish, for each of the system's devices, a partial power to be delivered or received, using the correspondence table; and - to transmit, to each of the devices in the system, the partial power to be delivered or received.

[0017] According to one embodiment, the state of a device is: - a state of charge if the device is of a first type configured to deliver and receive power; - the power delivered if the device is of a second type configured to deliver power; or - power received if the device is of a third type configured to receive power.

[0018] According to one embodiment, the device of the first type is a battery. Brief description of the drawings

[0019] These features and advantages, as well as others, will be described in detail in the following description of particular embodiments, given by way of non-limiting example, in relation to the accompanying figures, among which:

[0020] [Fig.1] is a diagram illustrating a power supply system according to an embodiment of the present description;

[0021] [Fig.2] is a diagram illustrating a computing device configured to generate a lookup table according to an embodiment of the present description;

[0022] [Fig.3] is a flowchart illustrating steps of a process for generating the lookup table according to an embodiment of the present description;

[0023] [Fig.4] is a diagram illustrating an example of a predictive model control algorithm according to an embodiment of the present description;

[0024] [Fig. 5] is a diagram illustrating a method for distributing power between assets, according to an embodiment of the present description; and

[0025] [Fig.6] is a graph illustrating an example of an evolution of a total power to be delivered to a network or received from the network. Description of the implementation methods

[0026] The same elements have been designated by the same reference numerals in the different figures. In particular, the structural and / or functional elements common to the different embodiments may have the same reference numerals and may have identical structural, dimensional and material properties.

[0027] For the sake of clarity, only the steps and elements necessary for understanding the described embodiments have been shown and are detailed. In particular, the evaluation of a frequency deviation in an AC power supply network, and the evaluation of the state of charge of an energy reservoir or battery, are known to a person skilled in the art and are not detailed in this description. The implementation and use of a predictive model control algorithm are also known to a person skilled in the art and are not detailed in this description.

[0028] In the following description, an asset is defined as a device configured to deliver or receive electricity and configured to modulate the received or delivered electrical energy relative to an operating power. In particular, a generator is defined as an asset configured to deliver electricity and configured to modulate the delivered electrical energy relative to an operating power, and a variable load is defined as an asset configured to receive electricity and configured to modulate the received electrical energy relative to an operating power. A tank is defined as an asset configured to both receive and deliver power.

[0029] Unless otherwise specified, when referring to two elements connected together, this means directly connected without intermediate elements other than conductors, and when referring to two elements coupled together, this means that these two elements can be connected or linked through one or more other elements.

[0030] Unless otherwise specified, transmission means direct or indirect transmission, for example via a monitoring and control interface or a programmable logic controller.

[0031] In the following description, when reference is made to absolute position qualifiers, such as the terms "front", "back", "top", "bottom", "left", "right", etc., or relative position qualifiers, such as the terms "above", "below", "superior", "inferior", etc., or to orientation qualifiers, such as the terms "horizontal", "vertical", etc., reference is made, unless otherwise specified, to the orientation of the figures in a normal position of use.

[0032] Unless otherwise specified, the expressions "approximately", "roughly", and "in the order of" mean within 10%, preferably within 5% or within 5.

[0033] Fig. 1 is a diagram illustrating a power supply system 100 according to an embodiment of the present description.

[0034] In the example of [Fig.1], the system 100 comprises a first customer site 102, a second customer site 104 and a central energy management system 106.

[0035] Each of the customer sites 102, 104 in the system 100 comprises at least one asset, with electrical energy being received by the customer site 102, 104 from an alternative power supply network and / or electrical energy being delivered by the customer site 102, 104 to the network. For example, one or more power supply contracts are in place involving the customer sites 102, 104 and network operators, establishing a commercial relationship between the entities. Although two customer sites 102, 104 are illustrated in the example in [Fig. 1], in alternative embodiments there may be a single customer site or more than two customer sites managed by the central power management system 106.

[0036] Each of the customer sites 102, 104 includes at least one asset which is, for example, adapted to meet current and future needs and the system 100 includes at least two assets distributed between the customer sites.

[0037] In the example of [Fig.1], the customer site 102 includes an energy storage tank 110 (“RES1”), for example a battery, for example a lithium-ion battery or for example a high-capacity battery bank, a hydraulic dam, etc.

[0038] In the example of [Fig. 1], the customer site 102 further includes a variable load 113 (“LB1”). A variable load is a device configured to consume a variable power over time, for example an electrolyzer, an electric heat generation source, a rotating machine (fan, pump), or a reversible hydraulic turbine, etc.

[0039] The reservoir 110 and the variable charge 113 are for example coupled to a first programmable logic controller (“PLC1” from the English “Programmable Logic Controller”) 119, for example a battery management system (“BMS” from the English “Battery Management System”).

[0040] The customer site 102 further includes, for example, an on-site control unit, not shown. The control unit includes, for example, a monitoring and control interface and a communication and control interface, not shown. The monitoring and control interface is, for example, a programmable logic controller (PLC), configured, for example, to implement power management, similar to the role of a power management system (PMS), or configured to implement a request-response (DR) box. The monitoring and control interface is, for example, coupled to one or more first on-site measuring devices configured to monitor active and reactive power for one and / or all three phases, and / or other electrical properties.For example, the first on-site measurement devices include one or more electrical measuring devices monitoring an electric current supplied to the site and / or received by the site, and one or more electrical measuring devices monitoring a current. Electrical power delivered by or received from reservoir 110, one or more electrical measuring devices monitoring the phases of electrical signals, and one or more AC (alternating current) frequency meters monitoring the frequency of electrical supply voltages present on the electrical network, as seen by site 102. Other properties that can be measured include electrical energy and power factor. The first monitoring and control interface is, for example, configured to receive measurement data from the first measuring devices on site. The control unit is, for example, configured to provide a communication interface between customer site 102, and in particular the first programmable logic controller 119, and the central management system 106, for example, via the communication and control interface.The communication and control interface is configured, for example, to communicate with the central energy management system 106, and in some cases with customer equipment, via the Internet. For example, although not illustrated in [Fig. 1], the connection between the communication and control interface and the Internet is made either via a switched communication network, such as an ADSL modem (asymmetric digital subscriber line), or via a wireless connection, for example, a cellular communication network. The connection between the communication and control interface and the Internet is, for example, a secure connection, for example, via the use of a VPN (Virtual Private Network).

[0041] In the example of [Fig. 1], the customer site 104 includes a reservoir 120 (“RES2”), for example a battery, a variable load 123 (“LB2”) and a generator 126 (“GSET1”). A generator is a device configured to produce power that varies over time, for example a generator set, an industrial load shedding system, etc.

[0042] Customer site 104 includes, for example, a second programmable logic controller 129 (“PLC2”). This second programmable logic controller 129 plays a similar role to the first programmable logic controller 119 at customer site 102 and will not be described in detail again. Customer site 104 also includes, for example, an on-site control unit (not shown) connected to the central management system 106.

[0043] Although in the example of [Fig.1] only one control unit 119, 129 is provided for each site 102, 104, in alternative embodiments some sites could include several control units.

[0044] The central management system 106 includes, for example, a control and data acquisition system 138, for example implemented by a management system of distributed energy resources (DERMS) or which is for example a supervisory control and data acquisition (SCADA) system, for example configured to receive data and transmit control signals to customer sites 102, 104, for example to programmable logic controllers 119, 129. In addition, the control and data acquisition system 138 is for example responsible for the acquisition and storage of data measurements from sites 102, 104, and also information concerning the state of the sites, for example a state of charge (SOC) of tanks 110, 120, power delivered by a generator 126, power received by a variable load 113, 123. The control and data acquisition system 138 includes for example a control and planning circuit 140 (“Planning Monitoring”) and a central interface 142 (“Pool gateway”).The central interface 142 is configured, for example, to receive and send data to sites 102, 104 and to the control and planning circuit 140.

[0045] A reservoir 110, 120 is configured to store energy. When a reservoir delivers power, the amount of energy stored in the reservoir decreases. When a reservoir receives power, the amount of energy stored in the reservoir increases. The ratio between the amount of energy stored in the reservoir at a given time and its maximum capacity is called its state of charge, for example, expressed as a percentage.

[0046] The central energy management system 106 also includes, for example, a distributed energy resource management system (DERMS) 150, which is, for example, a computer platform configured to organize the allocation of resources between loads or batteries of various customer sites of the system 100. The DERMS 150 also provides, for example, an interface with a market server and with an electricity network operator server.

[0047] For example, the market server provides information on electricity prices for current and / or future periods, and information on activations requested by the electricity grid operator. The electricity grid operator server corresponds, for example, to an IT platform of an electricity grid operator supplying electricity to customer sites 102, 104. In Europe, the electricity grid operator corresponds, for example, to the transmission system operator (TSO) and / or the distributed system operator (DSO). For example, the electricity grid operator server provides activation orders to the DERMS 150, and the DERMS 150 provides control data, such as monitoring data and / or load states, to the electricity grid operator server.

[0048] The DERMS 150 is, for example, configured to determine how energy use at customer sites 102, 104 can be adjusted in light of electricity prices for current and / or future periods in order to reduce energy costs and / or generate revenue at the customer sites, for example, through reductions in raw material charges or by balancing services in the grid. For example, the DERMS 150 is arranged to transmit control signals to customer sites 102, 104 indicating periods during which charging a tank 110, 120 from the electricity grid should be prioritized, for example, due to relatively low electricity prices, and periods during which discharging a tank 110, 120 to the electricity grid should be prioritized, for example, due to relatively high electricity prices.

[0049] In other embodiments, not illustrated in [Fig. 1], each site comprises, for example, one or more reservoirs and / or one or more variable loads and / or one or more generators. One or more variable loads and / or one or more generators are activated, for example, when a reservoir of system 100 has a relatively low or relatively high state of charge. A generator or a variable load is designated by a counterpart.

[0050] In some embodiments, an asset is a device configured to deliver or receive power. The production and / or consumption of power by each asset is, for example, modeled by a consumption model, for example, a dynamic model establishing a relationship between variables of the asset and, for example, being a function of a production cost or a consumption of raw materials. The consumption model is, for example, configured to favor a counterparty consuming relatively few raw materials or a counterparty with a relatively low production cost.

[0051] The power supply network, not shown, is configured, for example, to deliver alternating voltage or current at a target frequency, for example 50 Hz, to consumers. The power delivered by the network to the consumers is generated, for example, by a set of assets configured to deliver power to the network. A frequency deviation Af from the network frequency is, for example, an indicator of a change in consumer demand. System 100 is configured, for example, to deliver or receive a total power P_POOL that varies over time to respond to the change in demand. For example, System 100 is configured to: - evaluate or receive from the network, for example via the control and data acquisition system 138, a control signal N of a power to be delivered to or received from the network; and - deliver or receive the P_POOL power as a function of the N signal. Variations in the P_POOL power follow, for example, variations in the N signal and allow, for example, to compensate for a temporary increase or decrease in the power available on the network.

[0052] For example, according to one convention, P_POOL is positive when system 100 delivers power. The tanks of system 100 are then discharging. Conversely, P_POOL is negative when system 100 receives power. The tanks of system 100 are then charging. Another convention could be used.

[0053] According to one embodiment, the system 100 is configured to be engaged in automatic Frequency Restoration Reserve (aFRR). For example, the system 100 is configured to be able to deliver or receive a commitment power (“P_R2”) for a predefined duration, on demand. The commitment power P_R2 corresponds, for example, to the sum of a commitment power P_R2_1, ..., P_R2_k from each of the k reservoirs of the system 100, for example, reservoirs 110 and 120 in the example of [Fig. 1]. A reservoir's commitment power corresponds, for example, to the maximum power delivered or received by the reservoir. The commitment power of each reservoir is, for example, set at the beginning of an aFRR engagement, for example, by the DERMS 150.The engagement power of each tank varies, for example, from one aFRR engagement to another and is, for example, fixed over the duration of an aFRR engagement.

[0054] Each of the active k of the system 100 is for example configured to deliver or receive a partial power PI, ..., Pk, for example defined by the control and data acquisition system 138 or the programmable logic controller 119 or 129.

[0055] The signal N corresponds, for example, to a percentage of the commitment power to be delivered to or received from the network. The sign of the signal N indicates, for example, the activity of the asset, such as a power delivery or reception. For example, the signal N takes values ​​between -1 and +1. When the signal N is equal to -1, system 100 is, for example, configured to receive a power equal to P_R2. When the signal N is equal to +1, system 100 is, for example, configured to deliver a power equal to P_R2. The power P_POOL is, for example, defined by:

[0056] [Math.l] P POOL - N. PJ12. (1)

[0057] Tanks 110 and 120 are, for example, configured to receive or deliver a partial power PI and a partial power P2 respectively. Variable loads 113 and 123 are, for example, configured to receive a partial power P3 and P4 respectively. Generator 126 is, for example, configured to generate a partial power P5. The power ratings of the assets are calculated so that system 100 receives or delivers the total power P_POOL, in accordance with signal N and its engagement in aFRR, the total power corresponding to the sum of the partial powers of each of the system's assets, i.e., in the example of [Fig. 1]: P_POOL = P1 + P2 + P3 + P4 + P5.

[0058] The control and data acquisition system 138 or the programmable logic controller 119 or 129 is, for example, configured to define the partial power levels so as to favor the use of the tanks and to discourage the use of the counterparts, for example to reduce production costs or the quantity of raw materials consumed. The partial power levels are, for example, defined to promote a homogeneous use of the counterparts.

[0059] Figure 2 is a diagram illustrating a computing device 200 configured to generate a lookup table (LUT) according to an embodiment of the present description. For example, the computing device 200 is separate from the system 100 of Figure 1, but is capable of communicating with one or more subsystems of the system 100, such as the control and data acquisition system 138.

[0060] The computing device 200 includes, for example, a processor 210 (“CPU”, from the English “Central Processing Unit”), a memory 220 (“MEM”) and a communication interface 230 (“COM”).

[0061] The computing device 200 is, for example, configured to receive a past evolution of the signal N or the total power P_POOL over a period T. For example, the period T is between one week and ten years, for example between eight months and two years, and is, for example, approximately equal to one year. The past evolution of the signal N or the total power P_POOL over the period T is, for example, received via the communication interface 230, for example a wired connection, a USB (Universal Serial Bus) connection, a wireless connection, etc. The past evolution of the signal N or the total power P_POOL over the period T is, for example, recorded in the memory 220 of the computing device 200.

[0062] The computing device 200 is further configured, for example, to receive the consumption model of each of the assets in a system, for example, system 100 in [Fig. 1]. The consumption model of the assets is received, for example, via the communication interface 230 and stored in memory 220. The consumption model of the assets allows, for example, the computing device 200 to calculate An evolution of an asset's parameters, such as its state of charge, as a function of the partial power received or delivered by the asset. The asset consumption model is used, for example, by the calculation device 200 to order and / or select assets to activate from among all assets, for example, based on an operating cost or consumption.

[0063] The computing device 200 is for example configured to generate the LUT lookup table, for example via the processor 210. According to embodiments, the LUT lookup table matches a total power distribution instruction P_POOL between the system assets to a plurality of system states.

[0064] The distribution instruction includes, for example, distribution factors al, ..., ak of the total power P_POOL, for example such that:

[0065] [Math.2] k ai ~ 1 (2) z=i

[0066] where the system comprises k assets and ai is the distribution factor of an asset i. We then have Pi = ai.P_POOL with Pi the partial power of asset i.

[0067] A system state includes, for example, a value of the signal N and, for example, a state of each of the system's assets. The state of an asset is defined, for example, by a state of charge and / or a power delivered or received. For example, the state of a tank includes its state of charge, the state of a generator includes its power delivered, and the state of a variable load includes its power received.

[0068] Thus, for a state of the system and a value of the signal N, the LUT lookup table matches, for example, a set of distribution factors al, ..., ak of the total power P_POOL.

[0069] The computing device 200 is for example configured to calculate the distribution factors al, ..., ak, for example using a model predictive control algorithm (“MPC”), for example via the processor 210. The calculation of the distribution factors al, ..., ak is for example carried out on the basis of the consumption model of each of the devices and the past evolution of the total power over the period T.

[0070] The computing device 200 is configured for example to transmit, for example via the communication interface 230, the LUT lookup table to a system control device, for example the control and data acquisition system 138 of the system 100 described in relation to [Fig.1] or a programmable logic controller of a customer site, for example the programmable logic controller 119 or 129.

[0071] Fig. 3 is a flowchart illustrating steps of a process 300 for generating the LUT lookup table according to an embodiment of the present description.

[0072] The process 300 is for example executed by the computing device 200, described in relation to [Fig.2], for example via the processor 210.

[0073] The computing device 200 has access, for example through its memory 220, to the evolution of the signal N over the period T, to the consumption model of each of the system's assets and for example to a state of each of the system's assets.

[0074] The period T is, for example, decomposed into a multitude of instants t, two consecutive instants being separated by a time step At, for example between 1 s and 10 min and, for example, between 4 s and 10 s. The period T is, for example, further decomposed into forecast horizon periods T_p, for example corresponding to twenty time steps At. A control horizon period T_c is also defined, the control horizon period T_c being less than the forecast horizon period T_p, for example equal to five time steps At.

[0075] According to one embodiment, the state of each of the system's assets is initiated, for example, at a load state of 50% for each of the tanks, at a partial power output of 0 W for each of the generators, and at a power input of 0 W for each of the variable loads. According to other embodiments, the state of each of the system's assets is initiated differently. The state of each of the system's assets is, for example, calculated and recorded for a multitude of instants t by the computing device 200 during the process 300.

[0076] In a step 310 (“GENERATE PREDICTED SIGNAL”), the computing device 200 is configured for example to generate, for each instant of a forecast horizon period T_p, a forecast N' of an evolution of the signal N.

[0077] According to one embodiment, the calculation device 200, for example from the evolution of the signal N over a period preceding the forecast horizon period T_p, generates the forecast N' of the evolution of the signal N.

[0078] The forecast N' is for example generated using an estimate of a probability of being selected on a bid and / or a probability of being called upon to deliver or receive power based on the concurrent bids of other systems on the market.

[0079] According to one embodiment, the prediction N' is for example considered perfect during the execution of process 300.

[0080] In a step 320 (“GENERATE OPTIMAL FACTORS”), for example following step 310, the computing device 200 is configured, for example, to generate a set of distribution factors al, ..., ak for each instant t of a period control horizon T_c included in the forecast horizon period T_p, two consecutive control horizon periods T_c being for example separated by the time step At.

[0081] According to one embodiment, for each instant t of a control horizon period T_c, the computing device 200 executes a predictive model control algorithm to generate a set of distribution factors al, ..., ak. The predictive model control algorithm is, for example, executed using the state of each of the system's assets at time t, the consumption model of each of the system's assets, and the forecast N' over the control period T_c.

[0082] According to one embodiment, the state of each of the system's assets is recorded, for example in memory 220, for the first instant of the control horizon period T_c. The state of each of the system's assets is calculated, for example by the computing device 200, for each of the other instants of the control horizon period T_c. The state of each of the system's assets for an instant t' is calculated, for example, using the set of distribution factors al, ..., ak calculated for the instant preceding t' and using the consumption model of each of the system's assets.

[0083] According to one embodiment, a state of each of the system assets calculated for the beginning of the second instant of the control horizon period T_c is used for the generation of a set of distribution factors al, ..., ak for the first instant of a control horizon period following the control horizon period T_c.

[0084] According to one embodiment, the duration of the last control horizon periods T_c is reduced. For example, the last four control horizon periods T_c are respectively four, three, two and one time step.

[0085] In a step 330 (“STORE IN DATABASE”), for example following step 320, the computing device 200 is configured for example to add each set of allocation factors al, ..., ak calculated in step 320, indexed by the state of each of the system's assets and the value of the forecast N', to a database.

[0086] The database is for example stored in memory 220 of the computing device 200.

[0087] Steps 320 and 330 are repeated for example 333 for each control horizon period T_c included in the forecast horizon period T_p.

[0088] According to one embodiment, steps 310, 320 and 330 of process 300 are for example repeated 336 for each forecast horizon period T_p of period T.

[0089] In a step 340 (“STATISTICAL ANALYSIS”), for example following the last repetition of step 330, the computing device 200 is configured for example to perform a statistical study on the database.

[0090] At input to step 340, one or more sets of distribution factors al, ..., ak has for example been calculated for each instant of each forecast horizon period T_p of the period T. For example, the period T is relatively large and a relatively large number of system states are represented, once or several times.

[0091] In particular, if the control horizon period T_c includes c time steps At, between 1 and c sets of distribution factors al, ..., ak are generated for each instant t.

[0092] According to embodiments, a relatively large number of sets of distribution factors al, ..., ak, for example greater than or equal to 3000, is generated during steps 310 to 330.

[0093] In particular, several sets of distribution factors al, ..., ak are generated for the same state of the system or for relatively close states of the system.

[0094] The statistical analysis is configured for example to select or calculate a set of distribution factors al, ..., ak for a multitude of system states, for example if a set of distribution factors al, ..., ak is statistically significant.

[0095] Statistical analysis is carried out for example by a data partitioning method, for example by a centroid-based method such as a k-mean or k-medoid algorithm, or for example by a hierarchical grouping method or an expectation maximization algorithm.

[0096] According to one embodiment, the computing device 200 is configured to generate the LUT lookup table by matching a set of distribution factors al, ..., ak obtained using statistical analysis to a corresponding system state.

[0097] The LUT lookup table is, for example, represented by a matrix having k+1 dimensions, where k is the number of assets in the system. For example, the lookup table gives a set of distribution factors al, ..., ak for a state of each asset and for a value of the signal N.

[0098] According to one embodiment, the forecast N' depends on a probability of system activation and the current activations of the system's assets. Values ​​of N are, for example, compared to the forecast N' to test the robustness of the process 300 and the LUT lookup table. The calculations performed by the computing device 200 are, for example, carried out in a closed loop with, for example, a feedback loop configured to correct the ai and to take into account an interaction between the system and the market in real time.

[0099] Fig. 4 is a diagram illustrating an example of a 400 model predictive control (MPC) algorithm according to an embodiment of the present description.

[0100] Algorithm 400 is executed for example by the computing device 200, for example in steps 310 and 320 of process 300.

[0101] According to one embodiment, a prediction p, for example the prediction N' of the evolution of the signal N detailed in relation to [Fig. 3], is generated by a block 410 ("PREDICTION"). The prediction p is, for example, provided as input to a predictive control function of model 420.

[0102] The predictive control function of model 420 includes, for example, an optimizer 422 (“OPT”) and a complete model 424 of a system, for example, all or part of the system 100 described in relation to [Fig. 1]. The optimizer 422 is, for example, configured to generate a setpoint c. The optimizer 422 is, for example, an integer linear optimizer, a stochastic optimizer, an optimizer configured to use dynamic programming, etc.

[0103] The setpoint c includes for example a set of distribution factors al, ..., ak and the forecast p. The optimizer 422 is for example configured to transmit the setpoint c to the complete model 424.

[0104] The complete model 424 of the system includes a partial model 426 of the system, for example configured to generate a theoretical response y_th of the system for an input e. The partial model 426 of the system is, for example, generated using the consumption model of each of the system's assets and / or is, for example, a digital twin of the system 100 and / or is, for example, generated using differential equations. The input e corresponds, for example, to the setpoint c adjusted by a disturbance ("DIST"). The theoretical output y_th corresponds, for example, to a cost of the system, for example, an environmental cost, a quantity of raw materials consumed, a financial cost, etc. For example, the theoretical output y_th is calculated by summing an operating cost of each of the assets, for example, defined by the consumption model of each of the assets.The DIST disturbance corresponds, for example, to a measurement error, high-frequency noise adding a bias to measured values, etc.

[0105] According to one embodiment, the optimizer 422 is configured to generate the setpoint c so as to reduce the theoretical response y_th. The optimizer 422 is, for example, configured to take into account the theoretical response y_th by means of a second feedback loop 430, connecting an output of the partial model 426 to an input of the optimizer 422.

[0106] According to one embodiment, the forecast p is generated over the forecast horizon period T_p. The setpoint c comprises, for example, a set of distribution factors al, ..., ak for each instant of the control horizon period T_c and the forecast p, for example, over the control horizon period T_c. The setpoint c is generated, for example, during state 320 of process 300.

[0107] For example, the system's reservoirs have a lower operating cost than the system's variable loads and generators, and the use of the reservoirs is, for example, preferred. The setpoint c allows, for example, relatively low utilization of the counterparties over the period T.

[0108] The [Fig.5] is a diagram illustrating a method 500 of distributing the total power P_POOL between the assets of a system, for example all or part of the system 100 described in relation to the [Fig.1], according to an embodiment of the present description.

[0109] The process 500 is implemented for example by a control device, for example the control and data acquisition system 138 described in relation to [Fig.1] or a programmable logic controller of a customer site, for example the programmable logic controller 119 or 129.

[0110] According to one embodiment, a lookup table is stored in a memory of the control device. For example, the LUT lookup table, generated at the end of process 300 described in relation to [Fig. 3], is transmitted to the control device.

[0111] According to some embodiments, each of the system's assets is, for example, configured to transmit its state to the control device, for example periodically, for example with a time interval of between 1 s and 30 s, for example every 4 s. The states are, for example, transmitted to the control device via one or more devices, for example via a programmable logic controller and / or the control and data acquisition system 138. The state of an asset is, for example, a state of charge for tanks, a power received for variable loads, and a power delivered for generators. The control device is, for example, configured to generate a set of allocation factors a1, ..., a1 and, for example, to transmit a allocation factor, or a power to be delivered or received, to each asset of the system.The distribution factor or the power to be delivered or received is for example transmitted via one or more devices, for example via a programmable logic controller, for example the programmable logic controller 119, 129 described in relation to [Fig.1], and / or the control and data acquisition system 138. The power Pi to be delivered or received from an asset i is for example equal to ai.P_POOL. .

[0112] The control device is, for example, configured to receive: - the signal N, of the network, described in relation to [Fig.1], for example corresponding to an evolution of the total power P_POOL to be delivered or received from the system; and - the state (“SOC, P”) of each of the system assets.

[0113] The control device is configured for example to generate, from the LUT lookup table, the set of distribution factors al, ..., ak.

[0114] For example, if the state of the system, i.e. the value of the signal N and the values ​​of the state of each of the assets, is represented in the LUT lookup table, then the set of distribution factors al, ..., ak is the corresponding one in the LUT lookup table.

[0115] For example, if the system state is not represented in the LUT lookup table, then the control device is configured to calculate the set of distribution factors al, ..., ak, for example by performing an operation on sets of distribution factors al, ..., ak corresponding to similar system states. The operation is, for example, interpolation, averaging, etc. The choice of the operation performed depends, for example, on the computing capabilities of the control device, its memory, etc.

[0116] Fig. 6 is a graph illustrating an example of the evolution of total power to be delivered to or received from a network.

[0117] Graph 600 illustrates an evolution of the signal N, for example received from the network, over time (“TIME”). The signal N is for example between the value “-1”, corresponding to a total power P_POOL to be received equal to the commitment power P_R2, and the value “1”, corresponding to a total power P_POOL to be delivered equal to the commitment power P_R2.

[0118] The total power P_POOL is distributed among the system assets.

[0119] One advantage of the various embodiments described is that the processes are These methods are relatively modular and can, for example, be implemented on any asset with a consumption model. Another advantage is that the process described in relation to [Fig. 3] and configured to generate the LUT lookup table is performed only once for a system, for example, before an aFRR commitment. The LUT lookup table is valid for the duration of the aFRR commitment, and the calculations are performed only once. The LUT lookup table is stored in a controller, such as the control device. One advantage is that the controller has relatively small computing and memory capacity, and the LUT lookup table allows the controller to generate sets of allocation factors a1, ..., a1 configured so that the operating cost, for example, environmental or financial, is relatively low.Another advantage is the relatively easy implementation of the processes described at client sites, for example, the relatively easy implementation of the LUT lookup table, because relatively few calculations are performed at the client site. Another... The advantage is that no server is needed for the control device to generate a set of distribution factors al, ..ak.

[0120] Another advantage of generating the LUT lookup table by the computing device 200 is relatively easy maintenance of the algorithm.

[0121] Various embodiments and variants have been described. Those skilled in the art will understand that certain features of these various embodiments and variants could be combined, and other variants will become apparent to them. In particular, although the described embodiments detail the generation of a set of distribution factors a1, ..., ak, a set of powers P1, ..., Pk to be delivered or received for each of the assets, for example with P1 = a1.P_POOL, could be generated.

[0122] Although consecutive forecast horizon periods T_p have been described, a person skilled in the art would be able to adapt the processes to overlapping forecast horizon periods T_p.

[0123] Moreover, although an implementation of the computing device 200 by software has been described, in other embodiments, a complete or partial implementation by a dedicated circuit will be possible, such as by a custom integrated circuit (ASIC, from the English "Application Specified Integrated Circuit") or a programmable integrated circuit (FPGA, from the English "Field Programmable Gate Array").

[0124] Finally, the practical implementation of the embodiments and variants described is within the reach of a person skilled in the art, based on the functional indications given above.

Claims

Demands

1. A method for generating a lookup table (LUT) comprising allocation factors (al, ...ak) of a total power (P_POOL) to be delivered to an electrical supply network or of a total power (P_POOL) to be received from the network, the method comprising: - the reception, by a computing device (200), of a past evolution of a signal (N) from the network, the signal (N) being a control signal of a power to be delivered to the network or received from the network, over a period (T); - the reception, by the computing device (200), of a consumption model for each of a plurality of devices in a system, each device being configured to receive and / or generate a partial power; - the generation, by the computing device, of the lookup table (LUT) by calculating, for each of the devices and for a plurality of system states, the allocation factors (al, ..., ak) using a predictive model control algorithm based on the consumption model of each of the devices and the past evolution of the signal (N) over the period (T); and - the transmission, by the computing device (200), of the lookup table (LUT) to a system control device configured to control the partial power delivered to or received from the network by each of the system devices.

2. A method according to claim 1, further comprising the control, by the system control device, of the power delivered to the network or received from the network by each of the devices of the system, using the load table (LUT).

3. A method according to claim 1 or 2, wherein a system state comprises a signal value (N) and a state of each of the system devices, the state of each system device being: - a charging state if the device is of a first type configured to deliver and receive power; - a delivered power if the device is of a second type configured to deliver power; or - a received power if the device is of a third type configured to receive power.

4. A method according to claim 3, wherein the calculation of the lookup table (LUT) comprises the generation of a database, the generation of the database comprising, for a plurality of times in the period (T), the calculation, for each of the devices, for a state of the system and at a time in the plurality of times, of the distribution factors (al, ..., ak) using a predictive model control algorithm based on the consumption model of each of the devices.

5. Method according to claim 4, wherein the generation of the database further comprises, for a plurality of times in the period (T), the generation of a prediction (N', p) of an evolution of the signal (N) over a time interval following a time (t).

6. A method according to claim 4 or 5, wherein the calculation of the lookup table (LUT) further includes a statistical analysis of the database.

7. A method according to any one of claims 3 to 6, wherein the lookup table (LUT) is configured to associate the distribution factors (al, ..., ak) with a value of the system state.

8. A method according to any one of claims 3 to 7, wherein the consumption model of each device includes a relationship between a power delivered or received by the device and a consumption or production cost.

9. A computing unit configured to generate a lookup table (LUT) comprising allocation factors (al, ..., ak) of a total power (P_POOL) to be delivered to a power grid or a total power (P_POOL) to be received from the grid, the computing device being configured to: - receive a past evolution of a signal (N) from the grid, the signal (N) being a function of a frequency deviation of the grid, over a period (T); - receive a consumption model for each of a plurality of devices in a system, each device being configured to receive and / or generate a partial power; - generate the lookup table (LUT) by calculating, for each of the devices and for a plurality of system states, the allocation factors (al, ..., ak) using a predictive model control algorithm based on the consumption model of each of the devices and the past evolution of the signal (N) over the period (T); and - transmit the lookup table (LUT) to a system control device configured to control the partial power delivered to or received from the network by each of the system devices.

10. System comprising: - the calculation unit according to claim 9; and - the system control device configured to: - receive a signal (N) from the network, the signal (N) being a function of a frequency deviation of the network; - receive a measurement of a state of each of the devices of the system; - establish, for each of the devices of the system, a partial power to be delivered or received, using the lookup table (LUT); and - transmit, to each of the devices of the system, the partial power to be delivered or received.

11. System according to claim 10, wherein the state of a device is: - a state of charge if the device is of a first type configured to deliver and receive power; - a power delivered if the device is of a second type configured to deliver power; or - a power received if the device is of a third type configured to receive power.

12. System according to claim 11, wherein the device of the first type is a battery.

Images