Droop command and control loop for an inverter interfacing an intermittent energy source and an ac power grid

Abstract

The invention relates to a method for the primary control of an inverter interfacing an intermittent energy source and an AC power grid. Through each of its various aspects, the invention proposes a novel droop command and control loop or droop control loop (5) that adapts to the variations in active power available at the inverter (1). This loop may advantageously be implemented on each inverter that, among a plurality of generators connected to the grid, interfaces an intermittent energy source (2), while at the same time guaranteeing: a. correct allocation of active power among the generators, and b. correct signalling, via the frequency at the point of connection of each generator to the grid, of the fact that each generator has reached its active power limits, thereby allowing protective measures to be activated so that each generator of the plurality returns to its active power range.

Description

[0001]"Droop control loop for inverter interfacing an intermittent energy source and an AC electrical network" TECHNICAL FIELD OF THE INVENTION The present invention relates to a primary control method for an inverter that interfaces an energy source and an AC electrical network, the energy source being potentially intermittent. More specifically, the present invention relates to a droop control loop, hereinafter "droop control loop", which allows generators connected to an AC electrical network to contribute to maintaining balance or sharing active powers on the AC electrical network, in particular by controlling the frequency at an output point of each generator to the balance frequency of the AC electrical network.STATE OF THE ART Initially, the droop control loop strategy, hereinafter referred to as "droop strategy", was created to control synchronous machines that interface non-intermittent energy sources, such as fossil fuels. Later, the strategy was adapted to be implemented on inverters, as generators, also interfacing non-intermittent energy sources. However, this first adaptation is partly dysfunctional, especially if the droop strategy is implemented on inverters that interface, mainly, or even only, intermittent energy sources, such as solar, wind, etc.These malfunctions can even cause the disconnection of the inverters, which can endanger the stability of the AC power grid, if the available active powers of the other generators connected to the AC power grid are not sufficient to cover the electrical energy demand on the grid. The following paragraphs explain in more detail the operation of the conventional droop strategy and its possible malfunctions. Figure 1 represents an inverter 1 that interfaces a non-intermittent energy source 2 and the control-command scheme 5 corresponding to the conventional droop strategy. This strategy considers, as input, a measurement of the active power ^. ( ^ ) that inverter 1 supplies to the AC grid and uses it to calculate the setpoint ^ ^^^(^) of the frequency. The internal control loops 10 have the role of ensuring that the frequency at the connection point 4 of the inverter 1 to the alternating current network 3 follows the setpoint ^ ^^^ (^) calculated by droop control loop 5. It is assumed that network 3 potentially contains loads and other generators that interface non-intermittent energy sources, and that conventional droop strategies are also implemented on these generators. More precisely, the function used, by the conventional droop strategy, to calculate the value of the frequency ^ ^^^ (^) as a function of the active power ^(^) of the inverter takes the form of a linear function: ^ ^^^ ^^ ^^^ with a slope −^ ^ fixed, such that ^ ^ = This function can be represented by a profile {^, ^ ^^^} , as illustrated in Figure 2. Note that, for a certain droop strategy implemented on a generator, what is referred to as "profile {^, ^ ^^^} » is the set of steady-state operating points defined in the plane (^, ^ ^^^ ) and on which the generator can stabilize; this set of operating points is defined by the implemented droop strategy. During transients, the generator operating point may be outside this profile. As shown below, this representation helps to understand in particular how the conventional droop strategy allows a correct allocation of active powers when implemented on inverters interfacing non-intermittent energy sources. Figure 2 represents, on the same plane (^, ^ ^^^), the profiles corresponding to two inverters connected to the same AC power grid and controlled with the conventional droop strategy. Sub-indices 1 and 2 identify the droop strategy parameters that are relative to the first inverter and the second inverter, respectively. The available active power (^ ^^^^ ^ , ^ ^^^^ ^ ) at the output of each inverter coincides with the maximum assigned active power (^^^^ ^^^ ^^^ ^^^ ^ , ^ ^), to the inverter, since the power source that each inverter interfaces is non-intermittent. The key to the conventional droop strategy is that the frequency ^(^) is, in steady state, a unique quantity in each AC power grid 3. Transiently, two points in the same AC power grid may have different frequencies, but the electrical laws that govern the behavior of the AC power grid 3 push all points in the grid to have the same frequency ^(^) to reach steady state. This implies that, once steady state is reached, all generators in the grid will converge to the same frequency setpoint ^ ^^^ (^), and that this instruction will coincide with the frequency ^(^) at any point in the network 3. If the generators converge towards the same frequency instruction ^ ^^^(^), and more particularly, with reference to figure 2, if the first inverter reaches the operating point A and the second inverter reaches the operating point A' corresponding to the same set frequency , the proportion between the active power ^(^) produced and the assigned active power range ^ ^ ^^^ ^^^ , ^ ^^^ ^^^ ^ will be the same for all generators, that is, for the two inverters considered: This explains that the first attribute of the conventional droop strategy is that it allows a correct allocation of active powers in the case of non-intermittent energy sources. Moreover, the generators of network 3 will converge towards the set frequency ^ ^^^ (^) ≤ ^ ^^^ if and only if all generators { ^ } each provide active power greater than or equal to their available active power ^^ ^^^^ ^(^) ≥ ^ ^, ∀^ ^. Consequently, crossing the minimum frequency threshold ^ ^^^ can be used as a signal to activate protective measures, for example implemented using frequency-metric relays that allow disconnecting loads from the AC power grid 3. Conversely, the generators in grid 3 will converge towards the set frequency ^ ^^^ (^) ≥ ^ ^^^ if and only if all generators {^} each provide an active power less than or equal to their assigned minimum active power As a result, crossing the maximum frequency threshold ^ ^^^can be used as a signal to activate protective measures, for example implemented by frequency-metric relays that allow generators to be disconnected from the AC grid 3. We can then conclude that the conventional droop strategy has a second attribute: it signals, for example through the value of the set frequency ^ ^^^(^), if the generators reach their active power limits. And this signal can be used to activate protective measures of the AC power grid, in order to ensure that all generators return to their active power ranges. As we will develop below, this conventional droop strategy is not portable to the control of inverters 1 associated with intermittent energy sources, or at the very least does not allow the control of such inverters with the same attributes as those acquired for an inverter 1 associated with a non-intermittent energy source 2, and therefore does not allow to benefit from them. The traditional way to control inverters 1 that interface with intermittent energy sources 2, such as inverters for photovoltaic panels, is to control them so that the active power ^(^) that they supply to the grid 3 is always equal to the available active power ^ ^^^^(^) . Traditionally, droop control loops were not implemented for this type of inverter. Active power sharing schemes based on droop strategies and active power sharing schemes not based on droop strategies have since been developed to control inverters 1 that interface intermittent power sources 2. Droop-based schemes are for example described in the following papers: a. H. Liu, Y. Yang, X. Wang, P. C. Loh, F. Blaabjerg, W. Wang, and D. Xu, “An enhanced dual droop control scheme for resilient active power sharing among paralleled two-stage converters,” IEEE Transactions on Power Electronics, vol. 32, no. 8, pp. 6091–6104, Aug. 2017; b. H. Mahmood and J. Jiang, “Decentralized power management of multiple PV, battery, and droop units in an islanded microgrid,” IEEE Transactions on Smart Grid, vol.10, no.2, pp.1898–1906, Mar.2019; c. Wei Du, Qirong Jiang, Micah J.Erickson et Robert H. Lasseter, “Voltage-Source Control of PV Inverter in a CERTS Microgrid”, IEEE Transactions on Power Delivery, vol.29, no.4, pp 1726-1734, Août 2014 ; d. Zhe Chen; Robert H. Lasseter et Thomas M. Jahns, “Power Reserve for Grid- Forming PV Sources with Stability Enhancement in Mixed-Source Microgrids”, IEEE Power & Energy Society General Meeting (PESGM), pp 1-5, Août 2019 ; e. Z. Chen, R. H. Lasseter, and T. M. Jahns, “Active power reserve control for grid- forming PV sources in microgrids using model-based maximum power point estimation,” in 2019 IEEE Energy Conversion Congress and Exposition (ECCE), Sep.2019, pp.41–48 ; f. Z. Li, K. W. Chan, J. Hu, et J. M. Guerrero, “Adaptive droop control using adaptive virtual impedance for microgrids with variable pv outputs and load demands,” IEEE Transactions on Industrial Electronics, vol.68, no.10, pp.9630–9640, Oct. 2021 ; et g. N. L. Díaz, J. C. Vasquez, et J. M.Guerrero, “A communication-less distributed control architecture for islanded microgrids with renewable generation and storage,” IEEE Transactions on Power Electronics, vol.33, no.3, pp.1922–1939, Mar.2018. With an intermittent power source 2, the available power in the power source is variable. As a result, the available active power ^. ^^^^ (^) at the output of inverter 1 is variable. This is to adapt the conventional droop strategy to these variations in available active power ^ ^^^^ (^) that the droop control loops described in the articles listed above have been proposed. The droop control loops described in the first five articles cited above are based on different principles, but they give rise to profiles ^^, ^ ^^^ ^ equivalents composed of a fixed and immobile slope profile, to which is added a vertical profile which passes through ^ ^^^^(^) , as illustrated in Figure 9 by the profile ^^, ^ ^^^ ^ associated with the first undulator (identified by the index “1”), the vertical profile moving along the axis ^ as ^ ^^^^ (^) varies. The main problem with this type of profile is that it does not guarantee a correct allocation of active power. For example, referring to Figure 9, if we consider the steady-state operating point denoted A of the first inverter, this inverter provides active power equal to its available active power, while the second inverter whose operating point in the same steady-state is denoted A' would work well below its available active power ^ ^^^^ ^ (^). The droop control loops described in the last two articles cited above correspond to a droop profile with a slope that changes as the available active power ^ ^^^^(^) varies, so that the profile always passes through the points ^ ^ ^^^ ^^^ , ^ ^^^^ And , as illustrated in Figure 10 by the profile ^^, ^ ^^^^ associated with the first inverter (identified by the subscript “1”). The main problem with this type of droop profile is that its slope is strongly linked to the dynamic behavior of the generators. More precisely, steep slopes can cause oscillations in the electrical quantities that can destabilize the grid 3 (Cf. N. Pogaku, M. Prodanovic, and TC Green, “Modeling, analysis and testing of autonomous operation of an inverter-based microgrid,” IEEE Transactions on Power Electronics, vol. 22, no. 2, pp. 613–625, Mar. 2007; A. Firdaus and S. Mishra, “Mitigation of Power and Frequency Instability to Improve Load Sharing Among Distributed Inverters in Microgrid Systems,” IEEE Systems Journal, vol. 14, no. 1, pp. 1024–1033, 2020.). Furthermore, the droop control loops described in the last three articles cited above require an estimation / measurement of the available power ^ ^^^^(^). But this measurement / estimation may be wrong and differ from the actual available active power ^^^^^ ^é^^^^(^). This is dangerous in the case , because the inverter can then converge to an operating point with ^ ^^^^ (^) ≥ ^(^) > ^^^^^ ^é^^^^(^) , outside its active power range, and, given that then ^ ^^^ (^) ≥ ^ ^^^, no protective measures would be activated. Thus, the droop strategies proposed in the aforementioned articles: a. do not guarantee a correct allocation of active powers between generators, and / or b. cause oscillations in the electrical quantities that can destabilise the AC power grid, and / or c. do not allow correct signalling of activation of protective measures, in particular by means of the value of the frequency ^(^) at an output point of each generator (or equivalently at a connection point of each generator to the grid), when the generators reach, or even exceed, their active power limits.Active power sharing systems not based on droop strategies provide comparable results to those of droop strategies, including maintaining the balance of active power to which several generators contribute, correctly allocating active power between generators, and respecting the operating ranges of each generator connected to the AC grid. For example, the following articles describing such systems not based on droop strategies can be found in the scientific literature: a. E. Espina, J. Llanos, C. Burgos-Mellado, R. Cárdenas-Dobson, M. Martínez-Gómez, and D. Sáez, “Distributed Control Strategies for Microgrids: An Overview,” IEEE Access, vol. 8, pp. 193412–193448, 2020, doi: 10.1109 / ACCESS.2020.3032378, and b. DE Olivares et al., “Trends in Microgrid Control,” IEEE Transactions on Smart Grid, vol.5, no.4, pp.1905–1919, Jul.2014.The demonstrator described in “The Interflex consortium, 'D9.3 Demonstration results based on the KPI measurements and lessons learned from the demonstrations', Tech. Rep., 2019” and the industrial solution described in “C. Mahieux and A. Oudalov, 'Balancing act', ABB, 2015” also fall under such non-droop-based systems. However, these systems require communication links between the different generators and / or with a network server. The most important disadvantages related to this requirement are the cost associated with building these links and the potential reduction in system reliability in the event of a communication failure. On the contrary, droop control loops have the major advantage of not requiring, by nature, communication links, especially between the different generators, because they are implemented independently on each generator.It is thus clear from the above that there is a need for the control of inverters that interface intermittent energy sources with an AC power grid. Alternatives to existing systems, whether or not based on droop strategies, would be appreciated, in particular: a. if they did not necessarily implement communication links between the different generators and / or with a network server, and / or b. if they made it possible to guarantee a correct allocation of active power between generators, and / or c. if they made it possible to avoid the occurrence of oscillations in electrical quantities likely to destabilise the AC power grid. More particularly, with respect to existing systems based on droop strategies, a new control strategy that makes it possible to adapt to variations in available active power. ^^^^(^) typical of intermittent energy sources is desirable, which should be able to guarantee: a. a correct allocation of active power between the generators, and / or b. a correct signaling of activation of protective measures when the generators reach, or even exceed, their active power limits. SUMMARY OF THE INVENTION To achieve this objective, a first aspect of the invention provides a method of droop control, or "droop" method according to the English terminology, for an inverter intended to interface an intermittent energy source and an AC electrical network by regulating the frequency ^(^) at a connection point of the inverter to the network, the method comprising, as a function of: i. a constant ^ ^^^ defining the maximum network frequency limit, ii. a constant ^ ^^^ defining the minimum network frequency limit, iii. a constant ^ ^^^ ^^^defining the maximum active power assigned to the inverter, iv. a constant ^ ^^^ ^^^ defining the minimum active power assigned to the inverter, v. an active power measurement ^ ( ^ ) that the inverter supplies to the grid at time ^, and vi. a measurement or estimate of the available active power ^ ^^^^( ^ ) in the inverter at time ^, a. a step consisting of implementing: i. A first control-command branch, noted ^, configured to calculate a first frequency setpoint ^ ^^^^^ (^) verifying, or being a solution of, a first control-command function defined in a plane by the following equation: ii. A second control-command branch, noted ^ , comprising the implementation of a first proportional-integral regulator, referenced PI − 1, and a first saturation function and being configured to calculate a second frequency setpoint ^ ^^^^^(^) verifying, or being a solution of, a second control-command function defined in the plane by the following equation: max(^^^1(^), 0), with ^ ^ ^^^^, the proportional gain of the first proportional-integral regulator and ^ ^^^^^ , the integral gain of the first proportional-integral regulator, are parameters greater than zero and predetermined depending on the inverter concerned, and b. a step consisting of imposing a setpoint value ^ ^^^ (^) frequency at the point of connection of the inverter to the grid determined as a sum of the first and second frequency setpoints ^ ^^^( ^ ) = ^ ^^^^^( ^ ) + ^ ^^^^^( ^ ). A second aspect of the invention relates to a computer program product comprising instructions, which, when carried out by at least one processor, execute the steps of the droop control method according to the first aspect of the invention. A third aspect of the invention relates to a control unit or calculator for an inverter intended to interface an intermittent energy source and an alternating current grid by regulating the frequency ^(^) at the point of connection of the inverter to the grid, the control unit comprising electronic and / or microelectronic components configured to implement the droop control method according to the first aspect of the invention.A fourth aspect of the invention relates to an inverter for interfacing an intermittent energy source and an alternating current grid by regulating the frequency ^(^) at the point of connection of the generator to the grid and comprising at least one of: a. a readable non-transient medium comprising instructions, which when carried out by at least one processor, execute the steps of the droop control method according to the first aspect of the invention, and b. a control unit according to the third aspect of the invention.The inverter according to the fourth aspect of the invention may advantageously be free of a communication link with another generator or a server connected to the AC power grid, to ensure at least one, preferably each, of a correct allocation of active power between the generators and a correct signaling of activation of protective measures when the generators reach, or even exceed, their active power limits. Alternatively or in addition, the voltage at the connection point of the generator to the grid may be single-phase or polyphase. Through each of the different aspects of the invention, a novel droop control loop is proposed which adapts to variations in the active power available from the inverter.Accordingly, it can be implemented for an inverter interfacing an intermittent energy source, and advantageously on each inverter which, among a plurality of generators, interfaces an intermittent energy source, while ensuring: a. a correct allocation of active powers among the generators, and b. a correct signaling, by means of the frequency at the point of connection of each generator to the network, of the reaching by each generator of its active power limits, which allows the activation, possibly in a known manner, of protective measures, possibly known, so that each generator of the plurality returns to its active power range.BRIEF DESCRIPTION OF THE FIGURES The aims, objects, as well as the characteristics and advantages of the invention will emerge more clearly from the detailed description of an embodiment thereof which is illustrated by the following accompanying drawings in which: Figure 1 represents a functional diagram of a control-command system of an inverter interfacing a non-intermittent energy source with an alternating electrical network and illustrates the control-command type diagram of a control loop corresponding to a conventional droop strategy. Figure 2 graphically represents, in the plane ^^, ^. ^^^ ^, operating profiles of two inverters each interfacing a non-intermittent energy source with the same AC power grid, the inverters being controlled according to a conventional droop strategy. Figure 3 graphically represents, in the plane ^^, ^ ^^^^, operating profiles of two inverters interfacing for one, an intermittent energy source, and for the other, a non-intermittent energy source, with the same alternating electrical network, the inverters being controlled according to a conventional droop strategy. Figure 4 represents a functional diagram of a control-command system of an inverter interfacing an intermittent energy source with an alternating electrical network and illustrates a control-command type diagram of a control loop according to a first embodiment of the present invention. Figure 5 graphically represents, in the plane ^ ^, ^ ^^^^, operating profiles of two inverters interfacing, for one, an intermittent energy source, and, for the other, a non-intermittent energy source, with the same alternating electrical network, the inverter interfacing the non-intermittent energy source being controlled according to a conventional droop strategy and the inverter interfacing the intermittent energy source being controlled by a control loop as illustrated in Figure 4. Figure 6 represents a functional diagram of a control-command system of an inverter interfacing an intermittent energy source with an alternating electrical network and illustrates a control-command type diagram of a control loop according to a second embodiment of the present invention.Figure 7 represents a functional diagram of a control-command system of an inverter interfacing an intermittent energy source with an alternating electrical network and illustrates a control-command type diagram of a control loop according to a third embodiment of the present invention. It should be noted that the branch 12 represented is that corresponding to the first embodiment, but it could also be that corresponding to the third embodiment. Figure 8 graphically represents, in the plane ^^, ^. ^^^^, operating profiles of two inverters interfacing, for one, an intermittent energy source, and, for the other, a non-intermittent energy source, with the same alternating electrical network, the inverter interfacing the non-intermittent energy source being controlled according to a conventional droop strategy and the inverter interfacing the intermittent energy source being controlled by a control loop as illustrated in Figure 7. Figure 9 graphically represents, in the plane ^ ^, ^ ^^^^, operating profiles of two inverters interfacing, for one, an intermittent energy source, and, for the other, a non-intermittent energy source, with the same alternating current network, the inverter interfacing the non-intermittent energy source being controlled according to a conventional droop strategy and the inverter interfacing the intermittent energy source being controlled by a droop strategy adapted, according to known prior art, to an intermittent energy source. Figure 10 graphically represents, in the plane , operating profiles of two inverters interfacing, for one, an intermittent energy source, and, for the other, a non-intermittent energy source, with the same alternating electrical network, the inverter interfacing the non-intermittent energy source being controlled according to a conventional droop strategy and the inverter interfacing the intermittent energy source being controlled by a droop strategy adapted, according to another known prior art relative to that considered in figure 9, to an intermittent energy source. Figure 11 represents an electrical and control-command diagram of an embodiment of each of two droop control loops, one relating to the active power, noted ^, and illustrating a droop control loop according to an embodiment of the invention, and that relating to the reactive power, noted ^, which is not directly concerned by the present invention.The drawings are given as examples and are not limiting of the invention. Some constitute schematic representations of principle intended to facilitate the understanding of the invention. DETAILED DESCRIPTION OF THE INVENTION Before beginning a detailed review of embodiments of the invention, optional features which may possibly be used in combination or alternatively are set out below: According to an example of the first aspect of the invention, the implementation of the second control-command branch is furthermore a function of: a. a measurement of the direct voltage. at the inverter input, and b. a DC voltage setpoint ^ ^^ ^^^(^) at the input of the inverter, and further comprises the implementation of a first proportional regulator, referenced ^ − 1, the first proportional-integral regulator, the first saturation function and the first proportional regulator being configured to impose, at the second frequency setpoint ^ ^^^^^( ^ ) , a value defined by a variant of the second control-command function, said variant taking the form of the following equation: max ( ^^^1 ( ^ ) ′, 0 ) , with ^ ^ ^^^, the proportional gain of the first proportional regulator, is a parameter strictly greater than zero and predetermined according to the inverter concerned. The method according to this example of the first aspect of the invention makes it possible to prevent the second frequency setpoint ^ ^^^^^( ^ ) only the input voltage stabilizes then ^ ^^( ^ )of the inverter is not equal to its set value ^ ^^ ^^^( ^ ) . According to an example of the first aspect of the invention, the step of implementing the first and second branches further comprises, depending on: a. a measurement of the DC voltage at the inverter input, and b. a DC voltage setpoint ^ ^^ ^^^ (^) at the input of the inverter, the implementation of a third control-command branch, noted ^ , comprising the implementation of a second proportional-integral regulator, referenced ^^ − 2 , of a second saturation function and of a second proportional regulator, referenced ^ − 2 , and being configured to calculate a third frequency setpoint ^ ^^^^^ (^) verifying, or being a solution of, a third control-command function defined in the plane by the following equation: ^ ^^^^^, the proportional gain of the second proportional-integral regulator, ^ ^^^^^ , the integral gain of the second proportional-integral regulator are parameters greater than zero and predetermined depending on the inverter concerned, the proportional gain of the second proportional regulator, is a parameter strictly greater than zero and predetermined according to the inverter concerned, and: the setpoint value ^ ^^^ (^) frequency at the point of connection of the inverter to the grid is determined as the sum of the first, second and third frequency setpoints, ^ ^^^ (^) = ^ ^^^^^ (^) + ^ ^^^^^ (^) + ^ ^^^^^ (^). The method according to this example of the first aspect of the invention makes it possible to manage the possible eventuality according to which the measurement or estimation of the available active power ^ ^^^^(^) of an inverter 1 interfacing an intermittent energy source 2 may be erroneous and differ from the actual available active power, which is noted ^^^^^ ^é^^^^(^). According to an example of the first aspect of the invention, at least one, preferably each, of said parameters may be predetermined by an analysis of the time response of the inverter and / or by an analysis of the frequency response of the inverter and / or by implementing a pole placement method. According to an example of the first aspect of the invention, for a given inverter, each of the aforementioned parameters is substantially constant over time. According to an example of the first aspect of the invention, at least one of the second and third control-command branches further comprises the implementation of at least one subtractor and at least one adder.According to another example of the first aspect of the invention, the implementation of the first control-command branch may comprise the implementation of: a. two multipliers by a constant, this constant being defined by ^. ^ for one and by −^ ^ for the other, b. two subtractors, one to calculate ^ ( ^ ) − ^ ^^^ ^^^ and the other to calculate c. two adders, one to calculate ^ ^^^ ( ^ ) and the other to calculate ^ ^^^^^( ^ ). According to another example, the generator may be associated with, or comprise, one or more so-called internal control loops (as opposed in particular to the droop control loop proposed here, which is an external control loop). And the internal control loops of each generator are configured to control the frequency of the connection point of the generator to the alternating current grid; such internal control loops may be said to be configured in 'voltage source' mode (or 'grid-forming' in English). According to another example of the first aspect of the invention, the step of imposing the setpoint value ^ ^^^(^) of frequency at the point of connection of the inverter to the network comprises the supply of said value to at least one internal control loop of the inverter. It is specified that, in the context of the present invention, the following notations are adopted: a. ^(^) is the frequency at a point of an alternating current electrical network, and more precisely the frequency of the sinusoidal voltage at a point of connection of an inverter to an alternating current electrical network; b. ^ ^^^ (^) is a frequency setpoint for an inverter, calculated by its droop control loop and varying over time; c. ^ ^^^ is a constant defining the maximum frequency limit used to define the droop strategy; d. ^ ^^^is a constant defining the minimum frequency limit used to define the droop strategy; e. ^(^) is a measure of active power that an inverter provides to the AC grid to which it is connected, and varies over time; f. ^ ^^^ ^^^ is a constant defining the maximum assigned (or “nominal”) active power of an inverter; g. ^ ^^^ ^^^ is a constant defining the minimum assigned (or “nominal”) active power of an inverter; h. ^ ^^^^ is a measure or estimate of active power available in an inverter, and depends on the power source interfaced by the inverter. More specifically, ^ ^^^^ can be a constant if the inverter interfaces a non-intermittent power source, or can be time-variable, ^ ^^^^(^) , if the inverter interfaces an intermittent power source, with the AC power grid; i. is a measure of the DC voltage at the input of an inverter and varies over time; and j. ^ ^^ ^^^ is a set value of the DC voltage at the input of an inverter, and can be constant or variable over time, ^ ^^ ^^^ (^), depending on the constitution of the direct current circuit located upstream of the inverter and the way in which this circuit is controlled; it is preferable, if not necessary, that ^ ^^( ^ )does not deviate excessively from its setpoint to ensure the proper operation of the inverter and the energy source. In addition, the following terms are understood to mean: a. "generator" means any element capable of supplying to a network the power produced by an energy source by interfacing the energy source and the network. For example, a generator may be a synchronous machine that interfaces a gas (or hydraulic) turbine or an inverter that interfaces a photovoltaic panel or a wind turbine. As is clear from the detailed description that follows, the generators primarily discussed here not only interface the network and an intermittent energy source, but also actively control the amplitude and frequency of the sinusoidal voltage at the connection point between the generator and the network.Such generators operate in 'voltage source' mode (or 'grid-forming' in English), as opposed to 'current source' mode (or 'grid-feeding', or 'grid-following', in English). It may therefore be considered that each generator implemented within the framework of the present invention is a generator in 'voltage source' mode or equivalently a 'grid-forming' type generator; b. 'load', any electrical element connected to the AC electrical network to subtract active power from it, and more particularly that necessary for its operation; c. 'balance of active power(s)', the state of the network in which the sum of the active powers supplied to the network (by generators, for example) is equal to the sum of the active powers withdrawn from the network (by loads, for example), excluding losses; and d."droop control loop" or equivalently "droop control loop", a set of functions, typically digital and then implemented by a (micro)processor, but potentially analog and then implemented by electronic components, which make it possible to implement a droop strategy (or equivalently "droop control loop strategy") aimed at allowing a generator to participate in maintaining the active power balance of the AC electricity network to which it is connected. Normally, and this is the case here, two droop control loops are implemented on each generator: one which allows it to contribute to the active power balance and another to the reactive power balance. As is well known, the distinction between active power and reactive power only applies to the case of an AC electricity network, as opposed to a DC electricity network.Active power is the "useful" power (typically, the power produced by an energy source is transformed into active power once it passes into the AC power grid). Reactive power is necessary for the operation of the AC power grid, but it is not useful, it only goes "back and forth" in the grid. The appended figures representing functional diagrams, i.e., figures 1, 4, 6 and 7, represent only one of the two droop control loops, namely the control loop contributing to the balance of active powers which is the one concerned by the droop strategy proposed here. However, the droop control loop contributing to the balance of reactive powers is also present, although not represented in the aforementioned figures.In contrast to the latter, Figure 11 illustrates an electrical and control diagram in which, for example, an embodiment of each of the two droop control loops, namely that relating to the active power, noted ^ , and that relating to the reactive power, noted ^ , is represented; each of the two loops is recognizable in Figure 11 in that it involves one or the other of the active power ^ and that relating to the reactive power ^; e. "droop profile", a profile defined in the plane (^, ^. ^^^) as a set of steady-state operating points at which a given generator can stabilize, this set of operating points being defined by the implemented droop strategy. During transients, i.e. outside the steady state, the operating point of each generator may be outside this profile; f. "frequency at a point on a network", the frequency of the sinusoidal voltage at this point on the network. As indicated, malfunctions of the conventional droop strategy appear if it is implemented on a generator that interfaces an intermittent energy source with the AC electrical network. These malfunctions are notably due to the fact that the available active power ^ ^^^^( ^ ) , then variable in time ^, can be lower than the maximum active power assigned ^ ^^^ ^^^to the generator. To illustrate these malfunctions, consider figure 3 which represents, as in figure 2, the profiles {^, ^ ^^^} of two inverters controlled with the conventional droop strategy, but this time, unlike the case discussed in the introduction, a first inverter that interfaces an intermittent energy source, the second inverter being associated with a non-intermittent energy source. According to the illustrated scenario where ^ ^^^^ ^^^ ^^ ^ (^) < ^ ^ ^ , the AC grid can stabilize, i.e. reach a steady state, at the operating point A of the first inverter and at the operating point A' of the second inverter. The first inverter then provides an active power very close to its available active power ^ ^^^^ ^ , while the active power of the second inverter is well below its available active power ^ ^^^^ ^ . As a result, the second inverter will have a better dynamic response to disturbances in the AC grid, such as in the event of an increase(s) in demand.Furthermore, if the AC grid were to stabilize at operating point B of the first inverter and at operating point B' of the second inverter, the first inverter would then have to provide active power greater than its available active power ^ ^^^^ ^ , which would cause it to disconnect from the grid. It is therefore clear from this example that the conventional droop strategy does not guarantee a correct allocation of active power between generators connected to the grid regardless of the steady state to be achieved, when at least one of the generators interfaces an intermittent energy source with the AC grid. Moreover, even if the first inverter were 'overutilized', so that the grid could reach the steady state, the frequency at operating point B would be higher than ^. ^^^and, consequently, no protective measure can be activated solely on the basis of the value of the frequency ^(^). It can then be concluded that, in this case, the value of the frequency ^(^) does not correctly signal whether the generators are reaching their active power limits. The invention described below proposes a control-command strategy which allows an inverter to contribute to maintaining the active power balance of the AC electricity network to which it is connected, by controlling the frequency at an output point of the inverter. This strategy can therefore be called a droop strategy.And like the droop strategies known from the prior art, the strategy currently proposed has the major advantage of not requiring, in order to contribute to maintaining the active power balance of the AC electricity network, any communication link, in particular between the different generators and / or loads connected to the AC electricity network and / or with a server connected to the network.Unlike the droop strategies known from the prior art, in the case where one of the generators connected to the network is an inverter interfacing an intermittent energy source, the droop strategy proposed here makes it possible to guarantee a correct allocation of active powers among the different generators connected to the alternating electricity network and makes it possible to guarantee correct signaling, by means of a measurement of the frequency ^(^), of the reaching of their active power limits by each generator, which allows the activation of protective measures in particular configured so that, if a generator leaves its active power range, situated by definition between its minimum assigned power ^. ^^^ ^^^ and its available power ^ ^^^^(^), it returns to its active power range. As already announced above, the droop strategy proposed here is, like the droop strategies known from the prior art, "decentralized", that is to say that it can be implemented independently on each generator connected to the AC electricity network. Not only does its implementation therefore not require communication links, as already indicated above, but also the possible activation of the protection measures by implementing the droop strategy proposed here does not require communication links, in particular between the different generators and loads connected to the AC electricity network or with a network server.Let us also note that the droop strategy proposed here, regardless of its embodiments described below, is suitable for any alternating current electrical network, regardless in particular of its number of phases (these networks are most often single-phase or three-phase, but may, alternatively, comprise two phases or more than three phases). With reference to Figures 4 and 5, a first embodiment of the droop control loop 5 according to the first aspect of the invention is described below. Figure 4 shows an inverter 1 which, as a generator, interfaces an intermittent energy source 2, and the control diagram which corresponds to the aforementioned first embodiment of the droop loop 5 proposed here. This diagram is composed of the first and second “branches”, referenced 11 and 12 and noted ^ and ^ , which are described below.The droop control loop 5 illustrated in Figure 4 makes it possible to implement, like the other embodiments described below, a droop control method or "droop" method for an inverter 1 intended to interface an intermittent energy source 2 and an AC electrical network 3 by regulating the frequency ^(^) of the connection point 4 of the inverter 1 to the network 3. The network 3 potentially contains, and in a non-imitative manner, loads and / or other generators which interface intermittent energy sources and / or generators which interface non-intermittent energy sources. It is assumed that all the generators operate in 'voltage source' mode (or "grid-forming" in English), and that functional droop control loops are implemented on each of these generators.For example, the conventional droop control loop can be implemented on inverters that interface non-intermittent power sources and the proposed droop control loop 5 can be implemented on inverters that interface intermittent power sources. If there are generators that operate in grid-feeding mode, their operation can be likened to that of loads drawing negative active power from the grid. In order to understand the operation of the droop control loop 5 illustrated in Figure 4, Figure 5 represents the profiles {^, ^. ^^^} of a first inverter and a second inverter connected to the same alternating current network: the first inverter interfaces an intermittent energy source and is controlled with the droop control loop 5 illustrated in FIG. 4; the second inverter interfaces a non-intermittent energy source and is controlled with a conventional droop control loop. Note here that the droop control loop 5 illustrated in FIG. 4, as well as those illustrated in FIGS. 6 and 7, are preferably each implemented in the form of a computer program product comprising instructions, which, when carried out by at least one processor, execute the steps of the corresponding droop control method. The hardware medium on which each of the computer program products can be executed more particularly comprises a microprocessor and / or an analog card and / or a computer.Furthermore, the three blocks illustrated at the top of each of Figures 4, 6 and 7, respectively entitled "electrical conditioning circuit and intermittent energy source", "inverter" and "electrical filters" are preferably hardware components or comprise such components, and for example switches, capacitors, resistors etc. According to the illustration of Figure 4, this also being true for the illustrations given by Figures 6 and 7, the conditioning circuit of the intermittent energy source 2 transmits, to the inverter 1, a direct current, and the inverter 1 transforms this direct current into an alternating current, to transmit the latter to electrical filters 100, configured to filter the alternating current produced by the inverter 1 before transmitting it to the alternating electrical network 3, so that the alternating current produced by the inverter 1 can be distributed to the various loads connected to the network 3.The method is implemented based on a plurality of input parameters including: a. a constant ^. ^^^ defining the maximum network frequency limit 3, b. a constant ^ ^^^ defining the minimum network frequency limit 3, i.e. a constant ^ ^^^ ^^^ defining the maximum active power assigned to inverter 1, d. a constant ^ ^^^ ^^^ defining the minimum active power assigned to inverter 1, e. a measurement of active power ^(^) that inverter 1 supplies to network 3 at time ^, and f. a measurement or estimate of the available active power ^ ^^^^( ^ )in the inverter 1 at time ^. The various aforementioned measurements are preferably all carried out locally, that is to say at or around the inverter 1, in particular to justify the “decentralized” aspect of the method according to the first aspect of the invention. These measurements may involve one or more measurement processing units. These units may themselves implement, or even comprise, measuring devices, and in particular for measuring alternating or direct current and / or alternating or direct voltage, as represented in Figures 4, 6 and 7 by the “measurement processing” blocks and the long dashed lines which connect said blocks to the locations of the system 0 where the measurements can be carried out, typically on either side of the inverter 1 and at the output of the electrical filters 100.The various means of carrying out the measurements necessary for implementing the method according to the first aspect of the invention are not further described here, since they are deemed to be known to those skilled in the art. Based on the aforementioned input parameters of the method according to the first aspect of the invention, the latter firstly comprises a step consisting of implementing: a. A first control-command branch 11, denoted ^, which is configured to calculate a first frequency setpoint ^. ^^^^^( ^ ) , and b. A second control-command branch 12, denoted ^, which is configured to calculate a second frequency setpoint ^ ^^^^^( ^ ) . A second step in the process is to add the first and second frequency setpoints to obtain the final frequency setpoint, ^ ^^^( ^ ), which is sent to the internal control loops 10 shown in Figure 4. The mathematical expression corresponding to this step is: As illustrated in Figure 4, the first control-command branch 11 can be implemented by means of several control-command functions, and it can comprise: a. two multipliers by a constant, this constant being defined by ^ ^ for one and by −^ ^ for the other, b. two subtractors, one to calculate ^ ( ^ ) − ^ ^^^ ^^^ and the other to calculate c. two adders, one to calculate ^ ^^^ ( ^ ) and the other to calculate ^ ^^^^^( ^ ) The mathematical expression that defines the first frequency instruction given by the first branch 11 is: The second control-command branch 12 can be implemented by means of several control-command functions, and it can include: a. a subtractor to calculate ^ ^^^ ^^^ − ^(^) b. a proportional-integral regulator 121, denoted PI-1 in Figure 4, and c. a saturation function 122. The mathematical expression which defines the second frequency setpoint given by the second branch 12 is: where ^ ^ ^^^^, the proportional gain of the first proportional-integral regulator 121 and ^ ^^^^^ , the integral gain of the first proportional-integral regulator 121, are parameters greater than zero and predetermined depending on the inverter 1 concerned. It should be noted that the first term of the function ^^^1 ( t ) (ie, ^ ^ ^^^^ ∗ ^^ ^^^ ^^^ − ^ ( ^ ) ^) is called the "proportional term of regulator 121", and that the second term of the function ^^^1 ( t )(ie, ^ ^^^^^ ∗ ∫ ^ ^ ^^^ ^^^ − ^ ( ^ )^ ∗ ^^) is called the “integral term of the regulator 121”. The function ^^^1(t) is a proportional-integral regulator equation with a so-called “parallel” architecture. Other architectures of the PI regulator 121, for example a “series” architecture, would also be valid. The person skilled in the art is here deemed to know how to adapt the function ^^^1(t) to said other architectures, for an identical technical effect and a result of the same nature. The two gains and ^ ^ ^^^^are parameters predetermined before implementation of the control-command method according to the first aspect of the invention. They are preferably predetermined individually for each inverter 1 on which the control-command method is to be implemented, since the dynamic response of the inverter 1 depends on its characteristics and on the AC electrical network 3 to which it is connected. Different adjustment processes can be used, and for example, the analysis of the time response of the system 0, the analysis of the frequency response of the system 0, the pole placement method, or other processes deemed to be known to the person skilled in the art, and which are therefore not detailed further here. These processes are deemed to be known to the person skilled in the art, and are therefore not detailed further here. For the inverter 1 of Figure 4, if ^ ^^^ ^^^ − ^ ( ^ ) < 0, the second frequency setpoint converges to a zero value. Indeed, if ^ ^^^ ^^^ − ^( ^ ) < 0: a. the proportional term of the regulator 121 becomes strictly less than zero, and b. the integral term of the regulator 121 decreases. As a result, the variable ^^^1(^) decreases and eventually becomes strictly negative. The second frequency setpoint ^ ^^^^^( ^ ) is equal to ^^^1 ( ^ ) as long as the latter is greater than or equal to 0, but the second frequency instruction ^ ^^^^^( ^ ) remains equal to 0 when ^^^1(^) becomes strictly negative, by means of the saturation function 122 illustrated in Figure 4. Consequently, if ^ ^^^ ^^^ − ^(^) < 0 , the second frequency setpoint ^ ^^^^^ (^) will converge to a zero value and the value of ^ ^^^ (^) will depend only on the first branch The operation of the first branch 11 can be explained in the manner proposed below by describing two operating scenarios, it being understood that these two scenarios are given for illustrative purposes and not as a limitation, other scenarios being of course potentially observable. In the first scenario, the network 3 of figure 4 is initially in a steady state such that the inverter 1 injects active power ^ ( ^ ) with ^ ^^^ ^^^ − ^ ( ^ ) < 0 and ^ ^^^^( ^ ) − ^ ( ^ ) > 0 and its set frequency is higher than the minimum network frequency (^ ^^^( ^ ) > ^ ^^^). Assuming that a disturbance in the network, such as an increase in the active power consumed by the loads, occurs, the generators connected to the network 3 then increase their active powers immediately, in accordance with their 'voltage source' operating mode. On each generator, the increase in active power therefore causes a decrease in the setpoint frequency defined by its droop control loop. More precisely, on inverter 1 in Figure 4, and according to the first branch 11 described previously, the frequency decreases following a linear function of slope −^ ^. The electrical laws of the network force the setpoint frequencies of all generators connected to network 3 to converge towards the same value. By means of the droop loops implemented on each generator, if the total active power drawn from the network is, apart from losses, strictly less than the sum of the available active powers of all generators, then each generator stabilizes at an operating point where: a. its setpoint frequency has a value strictly greater than ^ ^^^ , and b. its active power has a value less than or equal to its available active power. More precisely, inverter 1 in Figure 4 converges to an operating point where ^ ( ^ ) < ^ ^^^^ (^) , which, according to the first branch 11 defined previously, translates as follows: Using a complementary graphical explanation, after a disturbance in the network such as an increase in the active power consumed by the loads, the droop control loops force the generators to converge towards steady-state operating points that belong to specific profiles in the plane (^, ^ ^^^). These profiles are constructed so that each generator increases its active power according to its operating margin, ensuring that no generator converges to an operating point above its available active power, if there are other generators in the network whose active powers are less than or equal to their available active powers. Moreover, if a generator converges to an operating point where it provides active power equal to its available active power or less than its available active power, but close to this limit, this necessarily implies that each of the other generators in the network also provides active power equal to its available active power or less than its available active power, but close to this limit. More precisely, for inverter 1 in Figure 4, if ^ ^^^ ^^^ − ^ ( ^ ) < 0 and with ^ ^^^^( ^ ) − ^ ( ^ )> 0, this profile is defined by the action of the first branch 11, and consists of a segment of slope −^ ^ which starts at point {^ ^^^ ^^^ , ^ ^^^ (^)} . According to the first branch 11 defined previously, this implies that the slope segment −^ ^ ends at the point It is important to note that if ^ ^^^^( ^ ) varies, this segment moves from right to left or from left to right in the plane (^, ^ ^^^ ) following the point and keeping its slope −^ ^ . This is explained by the dependence of the variable ^ ^^^ (^) relative to the available active power ^ ^^^^( ^ ) (ie, ^ ^^^ ( ^ ) =^ ^^^ + ^ ^ × ^^ ^^^^( ^ ) − ^ ^^^ ^^^ ^ ), which translates the fact that ^ ^^^ (^) increases or decreases if ^ ^^^^(^) increases or decreases, respectively. With reference to the operating profiles illustrated in Figure 5, the slope segment −^ ^is represented by the element referenced 1001 associated with the first inverter. A possible steady-state state for this network corresponds to operating point A for the first inverter and operating point A' for the second inverter. If a disturbance such as an increase in the active power consumed by the loads occurs, the droop control loops of the first and second inverters force them to converge towards operating points on the drawn profiles, such as operating point B for the first inverter and operating point B' for the second inverter. It can be seen that, between the steady-state state corresponding to operating points A and A' and the steady-state state corresponding to operating points B and B', the first and second inverters have increased their active powers, but neither has exceeded its available active power.Furthermore, if one of the first and second inverters supplies the network with active power equal to or less than its available active power, but close to this limit, this necessarily implies that the other of the first and second inverters also supplies active power equal to or less than its available active power, but close to this limit (Cf. points B and B'). In the second scenario which serves to explain the operation of the first branch 11, the network 3 of figure 4 is initially in a steady state such that the inverter 1 injects active power ^. ( ^ ) with ^ ^^^ ^^^ − ^ ( ^ ) < 0 and ^ ^^^^( ^ ) − ^ ( ^ ) > 0 and its set frequency is strictly higher than the minimum network frequency (^ ^^^ (^) > ^ ^^^). Assuming that a disturbance in the network, such as an increase in the active power consumed by the loads occurs, the generators connected to the network 3 then increase their active powers immediately, in accordance with their 'voltage source' operating mode. On each generator, the increase in active power therefore causes a decrease in the setpoint frequency defined by its droop control loop. More precisely, on inverter 1 in Figure 4, and according to the first branch 11 described previously, the frequency decreases following a linear function of slope −^ ^. The electrical laws of the network force the set frequencies of all generators connected to network 3 to converge towards the same value. By means of the droop loops implemented on each generator, if the total active power drawn from the network is, apart from losses, strictly greater than the sum of the active powers available from all generators, then in each generator: a. the set frequency evolves towards values ​​strictly lower than ^ ^^^ , and b. the active power evolves towards values ​​strictly higher than the available active power. More precisely, the active power of inverter 1 illustrated in Figure 4 evolves towards operating points such that ^ ( ^ ) > ^ ^^^^ (^), which, according to the first branch 11 defined previously, translates as follows: Crossing the minimum frequency threshold ^ ^^^is then observed which can be used as a signal to activate protective measures, for example implemented using frequency-metric relays which allow to disconnect loads connected to the electrical network and return the network to a functional state where all the generators inject into the network active powers less than or equal to their available active powers. As explained previously, for inverter 1 in Figure 4, the variable ^ ^^^^^( ^ ) is zero, or converges to a zero value, if ^ ^^^ ^^^ − ^ ( ^ ) < 0. Consequently, the second branch 12 has no influence on the set frequency ^ ^^^( ^ ) that if ^ ^^^ ^^^ − ^ ( ^ )≥ 0. The operation of the second branch 12 can be explained in the manner proposed below, by describing two operating scenarios, it being understood that these two scenarios are given for illustrative purposes and not as a limitation, other scenarios being of course potentially observable. In the first scenario, the network 3 of figure 4 is initially in a steady-state state such that the inverter 1 is at the operating point corresponding to the activation limit of the second branch 12, that is to say with: Further assuming that ^ ^^^^ (^) < ^ ^^^ ^^^ , the value of the reference setpoint is given by the following expression: ^ ^^^ (^) = ^ ^^^ (^) < ^ ^^^. Finally, it is assumed that there is at least one other generator whose active power injected into the network is strictly greater than its assigned minimum active power. With reference to the droop profiles illustrated in Figure 5, the initial state described above corresponds to the operating point C for the first inverter and to the operating point C' for the second inverter. If a disturbance in the network, such as a decrease in the active power consumed by the loads, occurs, then the generators connected to network 3 decrease their active powers immediately, in accordance with their 'voltage source' operating mode. On each generator, the decrease in active power therefore causes an increase in the setpoint frequency defined by its droop control loop. More precisely, on inverter 1 in Figure 4, the aforementioned decrease in active power implies that ^ ^^^ ^^^ − ^ ( ^ )> 0. Consequently, and according to the second branch 12 defined previously: a. the proportional term of the regulator 121 becomes strictly greater than zero, and b. the integral term of the regulator 121 increases progressively. Consequently, the variables ^^^1 ( ^ ) become strictly greater than zero and gradually increase. This implies a progressive increase in the value of the frequency setpoint ^ ^^^ (^). It should be noted that the progressive increase in the value of the frequency setpoint ^ ^^^( ^ ) only stops when the input of the integral term of regulator 121 becomes zero, that is, when ^ ^^^ ^^^− ^(^) = 0. By the electrical laws that govern the network, the progressive increase in the frequency setpoint of inverter 1 forces the other generators in the network to reduce the active power they inject. This allows inverter 1 to increase its active power ^(^) and for it to become equal to its minimum assigned active power. When ^ ^^^ ^^^ − ^ ( ^ ) = 0, the integral term of the regulator 121 stabilizes; it neither increases nor decreases. Consequently, the variables ^^^1 ( ^ ) , and the frequency setpoint ^ ^^^( ^ )also stabilize. The electrical laws of the network force the set frequencies of all generators connected to network 3 to converge towards the same value. By means of the droop loops implemented on each generator, if the total active power drawn from the network is, apart from losses, strictly greater than the sum of the minimum active powers assigned to all generators, then each generator stabilizes at an operating point where: a. its set frequency has a value strictly less than ^ ^^^ , and b. its active power has a value greater than or equal to its assigned minimum active power. More precisely, on inverter 1 of Figure 4, and according to the second branch 12 described above, ^ ( ^ ) becomes exactly equal to ^ ^^^ ^^^due to the influence of the integral term of the regulator 121. It is important to note that, throughout the process described in the paragraphs above, the first branch 11 of the inverter 1 is active and increases or decreases the first frequency setpoint ^ ^^^^^ (^) , when the active power ^(^) decreases or increases, respectively, but the effect of the integral term of the regulator 121 is dominant, making the active power ^ converge ( ^ ) towards ^ ^^^ ^^^ in steady state, and, according to the definition of the first branch 11 given previously, converging the first frequency instruction ^ ^^^^^ (^) towards ^ ^^^(^). Using a complementary graphical explanation, after a disturbance in the network such as a decrease in the active power consumed by the loads, the droop control loops force the generators to converge to steady-state operating points that belong to specific profiles in the plane (^, ^ ^^^ ). These profiles are constructed so that each generator decreases its active power according to its operating margin, ensuring that no generator converges to an operating point below its assigned minimum active power, if there are other generators in the network whose active powers are greater than or equal to their assigned minimum active powers. More precisely, for inverter 1 in Figure 4, if ^ ^^^ ^^^ − ^(^) ≥ 0, the droop profile ^^, ^ ^^^ ^ is defined by the action of the second branch 12, and consists of a vertical segment that starts at point ^^ ^^^ ^^^ , ^^^^ ( ^ ) ^ and ends at the point ^^ ^^^ ^^^ , ^ ^^^ ^. This profile ensures that the active power ^ ( ^ ) does not remain lower than ^ ^^^ ^^^ if there are other generators whose active power is higher than their assigned minimum active power. As explained before, and according to the definition of the first branch 11, the variable ^ ^^^ (^) depends on the available active power ^ ^^^^( ^ ) (ie, ^ ^^^ ( ^ ) =^ ^^^ + ^ ^ × ^ ^ ^^^^( ^ ) − ^ ^^^ ^^^^ ). Thus, the variable ^ ^^^ ( ^ ) increases or decreases, if the available active power ^ ^^^^( ^ ) increases or decreases, respectively. It is important to note that the aforementioned vertical segment corresponding to the second branch 12: a. lengthens when the variable ^ ^^^ (^) decreases and b. contracts if ^ ^^^ ( ^ )increases; and in this way, this vertical segment always starts at point ^^ ^^^ ^^^ , ^ ^^^ ( ^ ) ^. It is also important to note that the intersection points between the segments corresponding to the first branch 11 and the second branch 12 coincide at ^^ ^^^ ^^^ , ^ ^^^( ^ )^. With reference to the droop profiles illustrated in Figure 5, the aforementioned vertical segment is represented by the element referenced 1002 associated with the first inverter. A possible steady-state state for this network corresponds to the operating point C for the first inverter and to the operating point C' for the second inverter. If a disturbance such as a decrease in the active power consumed by the loads occurs, the droop control loops of the first and second inverters force them to converge to operating points on the drawn profiles, such as at the operating point D for the first inverter and at the operating point D' for the second inverter.It can be seen that, between the steady-state state corresponding to operating points C and C' and the steady-state state corresponding to operating points D and D', the first inverter maintains the same active power, equal to its assigned minimum active power, while the second inverter decreases its active power, without it becoming lower than its assigned minimum active power. In the second scenario intended to explain the operation of the second branch 12, the network 3 of Figure 4 is initially in a steady-state state such that the inverter 1 injects an active power such that ^. ^^^ ^^^ − ^ ( ^ ) = 0 and its set frequency given by the expression: such that ^ ^^^( ^ ) ≥ ^ ^^^ (^) and that ^ ^^^( ^ ) < ^ ^^^. Finally, it is assumed that there is at least one other generator whose active power injected into the network is strictly greater than its assigned minimum active power. If a disturbance in the network, such as a decrease in the active power consumed by the loads, occurs, the generators connected to network 3 decrease their active powers immediately, in accordance with their 'voltage source' operating mode. On each generator, the decrease in active power causes an increase in the setpoint frequency defined by its droop control loop. More precisely, on inverter 1 in Figure 4, a decrease in active power causes ^ ^^^ ^^^ − ^ ( ^ )> 0. Consequently, and according to the second branch 12 defined previously: a. the proportional term of the regulator 121 becomes strictly greater than zero, and b. the integral term of the regulator 121 increases progressively. Consequently, the variables ^^^1(^) become strictly greater than zero and gradually increase. This implies that the value of the frequency setpoint ^ ^^^( ^ ) increases gradually. The electrical laws of the network force the set frequencies of all generators connected to network 3 to converge towards the same value. By means of the droop loops implemented on each generator, if the total active power drawn from the network is, apart from losses, strictly less than the sum of the minimum active powers assigned to all generators, then in each generator: a. the set frequency evolves towards values ​​strictly greater than ^ ^^^, and b. the active power evolves towards values ​​strictly lower than the minimum assigned active power. More precisely, on the inverter 1 of figure 4, and according to the second branch 12 described above, this is done by means of the integral term of the regulator 121. If ^ ( ^ ) remains strictly less than ^ ^^^ ^^^ , the integral term of regulator 121 continues to increase the value of ^ ^^^ (^) and the set frequency ends up becoming strictly greater than ^ ^^^ . It is important to note that, throughout the process described in the paragraphs above, the first branch 11 of inverter 1 is active and increases or decreases the first frequency setpoint ^ ^^^^^ (^) , when the active power ^(^) decreases or increases, respectively, but the effect of the integral term of the regulator 121 is dominant, and the frequency setpoint^ ^^^(^) gradually increases. Crossing the maximum frequency threshold ^ ^^^can be used as a signal to activate protective measures, for example implemented using frequency-metric relays that allow disconnecting generators connected to the electrical network and returning the network to a state where all generators inject into the network active powers greater than or equal to their assigned minimum active powers. It is concluded that the combination of the first and second branches 11 and 12 gives rise to a droop control loop 5 which retains the attributes which are those of the conventional droop control loop 5 when it is applied to a non-intermittent energy source 2.More particularly, the combination of the first and second branches 11 and 12 ensures a correct allocation of active power between the generators, because, via the first branch 11, no generator converges to an operating point located above its available active power if there are other generators in the network whose active powers are less than or equal to their available active powers. Via the first branch 11, if a generator converges to an operating point where it provides an active power equal to its available power or less than its available active power, but close to this limit, this necessarily implies that each of the other generators in the network also provides an active power equal to its available active power or less than its available active power, but close to this limit.By means of the second branch 12, no generator converges to an operating point below its assigned minimum active power if there are other generators in the network whose active powers are greater than or equal to their assigned minimum active powers. Correct signaling of the inverters having reached their active power limits by measuring the frequency is achieved, because, by means of the first branch 11, if the total active power drawn from the network is, apart from losses, strictly greater than the sum of the available active powers of all the generators, then in each generator: a. the set frequency evolves towards values ​​strictly lower than ^. ^^^, and b. the active power evolves towards values ​​strictly higher than the available active power. By means of the second branch 12, if the total active power drawn from the network is, apart from losses, strictly lower than the sum of the minimum active powers assigned to all the generators, then in each generator: a. the set frequency evolves towards values ​​strictly higher than ^ ^^^, and b. the active power evolves towards values ​​strictly lower than the assigned minimum active power. A second embodiment of the control-command method according to the first aspect of the invention is described below with reference to Figures 5 and 6. This second embodiment can be considered as a variant of the first embodiment described above. Indeed, the second embodiment is in accordance with the first, with the difference that the implementation of the second control-command branch 12: a. is furthermore a function of: i. a measurement of the DC voltage at the input of inverter 1, and ii. a set value of the direct voltage ^ ^^ ^^^ (^) at the input of the inverter 1, and b. further comprises the implementation of a first proportional regulator 123, referenced ^ − 1 in Figure 6. Note that the set value of the DC voltage ^ ^^ ^^^(^) is the set value of the voltage at the input of inverter 1. It can be constant or variable over time, depending on how the DC circuit located upstream of inverter 1 is constructed and controlled. It is preferable that the measurement of the DC voltage does not deviate excessively from its set value to ensure the proper functioning of the inverter 1 and the energy source 2 that it interfaces with. The mathematical expression that defines the second branch 12 according to the second embodiment is: the proportional gain of the first proportional regulator 123, is a parameter strictly greater than zero and predetermined according to the inverter 1 concerned. It should be noted that the first term of the function ^^^1 ( t ) ′ (ie, ^ ^ ^^^^ ∗ ^^ ^^^ ^^^ − ^ ( ^ )+ ^^^2(^)^) is called the "proportional term of the regulator 121", and that the second term of the function ^^^1 ( t ) ′ (ie, ^ ^^^^^ ^^) is called the "integral term of regulator 121". The function ^^^1(^) ^ is a Proportional-Integral regulator equation with a so-called "parallel" architecture. As already stated previously, other PI regulator architectures, for example a "series" architecture, would also be valid. The person skilled in the art is here deemed to know how to adapt the function ^^^1(t)′ to said other architectures, for an identical technical effect and a result of the same nature. The proportional gain ^ ^^^^ of the first proportional regulator 123 is a predetermined parameter before implementation of the control-command method according to the present variant of the first aspect of the invention. It is preferably predetermined for each inverter 1 on which the control-command method is to be implemented, because the dynamic response of the inverter 1 depends on its characteristics and on the AC electrical network 3 to which it is connected. Different adjustment processes can be used, and for example, the analysis of the time response of the system, the analysis of the frequency response of the system, the pole placement method, or other processes deemed to be known to the person skilled in the art, and which are therefore not detailed further here. These processes are deemed to be known to the person skilled in the art, and are therefore not detailed further here.The variant described here of the second branch 12 performs all the functions provided by the second branch 12 of the first embodiment which is described above. Additionally, it further ensures that, in steady state, the measurement of the continuous voltage ^. ^^( ^ ) is equal to the set value of the DC voltage ^ ^^ ^^^( ^ ) , if inverter 1 stabilizes at a steady-state operating point with ^(^) equal to ^ ^^^ ^^^ . This additional function of the variant of the second branch 12 is ensured by the term ^^^2 ( ^ ) , illustrated in Figure 6, which is added to the input of the Proportional-Integral regulator 121. As explained before, the second frequency setpoint ^ ^^^^^( ^ ) does not stabilize until the input of the integral term ^^ − 1 becomes equal to zero. In the variant of the second branch 12, the input of this integral term is ^ ^^^ ^^^ − that this input is equal to zero, the active power ^ ( ^ ) must be equal to the minimum assigned power ^ ^^^ ^^^ and the measurement of the direct voltage ^ ^^( ^ ) must be equal to the DC voltage setpoint ^ ^^ ^^^( ^ ) . The integral term of regulator 121 then ensures that, in steady state, ^ ^^( ^ ) is equal to ^ ^^ ^^^( ^ ) if ^ ( ^ ) is equal to ^ ^^^ ^^^ . Note that, usually, a control loop (outside the droop control loop) is provided which ensures that the voltage measurement continues ^ ^^( ^ ) is equal to the set value of the DC voltage . However, this 'ordinary' control loop saturates and shuts down when ^(^) ≤ ^ ^^^ ^^^ , hence the usefulness of the present variant of the second branch 12. It is important to ensure that the voltage continues ^ ^^( ^) be equal to or close to the set value of the DC voltage ^ ^^ ^^^ (^) , otherwise malfunctions of the inverter 1 and / or the energy source 2 that it interfaces may occur. It should be considered that the measurement or estimation of the available active power ^ ^^^^ (^) of an inverter 1 interfacing an intermittent energy source 2 may be erroneous and differ from the actual available active power, which is noted ^^^^^ ^é^^^^(^) . This is problematic, even dangerous, in the case where ^^^^^ ^é^^^^(^ ) < ^ ^^^^ (^), because inverter 1 can converge to an operating point with ^^^^^ ^é^^^^(^ ) < ^ ( ^ ) ≤ ^ ^^^^( ^ ) , outside its active power range. According to the definition of the first branch 11 given previously, at such an operating point, the set frequency of inverter 1 would be ^ ^^^( ^ ) ≥ ^ ^^^. As a result, no protective measure would be activated, and this would cause the capacitors placed behind the inverter 1 to gradually discharge until it is disconnected. To address this possibility, a third embodiment of the method according to the first aspect of the invention is proposed, which is described below with reference to Figures 7 and 8. This third embodiment is consistent with at least one of the first and second embodiments, with the difference that the method according to the third embodiment further comprises a step of implementing a third control-command branch 13, denoted ^ in Figure 7, configured to calculate a third frequency setpoint ^ ^^^^^( ^ ) . The implementation of this third branch 13 is furthermore a function of: a. a measurement of the direct voltage at the input of inverter 1, and b. a set value of the direct voltage ^ ^^ ^^^(^) at the input of the inverter 1. The implementation of this third branch 13 may further comprise the implementation of: a. a second proportional-integral regulator 131, referenced ^^ − 2 in Figure 7, b. a second saturation function 132, and c. a second proportional regulator 133, referenced ^ − 2 in Figure 7. The step used to calculate the final frequency setpoint consists of adding the first, second and third frequency setpoints. The equivalent mathematical expression is: The mathematical expression that defines the third branch 13 is: ^ ^^^^^ (^) = min(^^^3(^), 0), where: ^ ^ ^^^^, the proportional gain of the second proportional-integral regulator 131, ^ ^^^^^ , the integral gain of the second proportional-integral regulator 131 are parameters greater than zero and predetermined depending on the inverter 1 concerned, and the proportional gain of the second proportional regulator 133, is one of the parameters strictly greater than zero and predetermined according to the inverter 1 concerned. It should be noted that the first term of the function ^^^3(t) (ie, ^ ^ ^^^^ is called the "proportional term of regulator 131", and that the second term of the function ^^^3 ( t ) (ie, ^ ^^^^^ ∗ ∫ ^^ ^^^^ ^^^ (^) − ^ ( ^ ) ^ ) is called the "integral term of the regulator 131". The function ^^^3(t) is a Proportional-Integral regulator equation with a so-called "parallel" architecture. As previously, other architectures of the PI regulator, for example a "series" architecture, would also be valid. The person skilled in the art is here deemed to know how to adapt the function ^^^3(t) to said other architectures, for an identical technical effect and a result of the same nature. The three gains ^ ^ ^^^^, ^ ^^^^^ And , are parameters predetermined before implementation of the control-command method according to the first aspect of the invention. They are preferably predetermined individually for each inverter 1 on which the control-command method is to be implemented, since the dynamic response of the inverter 1 depends on its characteristics and on the AC electrical network 3 to which it is connected. Different adjustment processes can be used, and for example, the analysis of the time response of the system, the analysis of the frequency response of the system, the pole placement method, or other processes deemed to be known to the person skilled in the art, and which are therefore not further detailed here. These processes are deemed to be known to the person skilled in the art, and are therefore not further detailed here. The role of the third branch 13 is significant only in the case where ^^^^^ ^é^^^^(^ ) < ^(^) ≤ ^ ^^^^(^). Below we explain how the third branch 13 works in other possible situations. If ^(^) > ^ ^^^^ (^) , the first control branch command 11 already ensures that the frequency setpoint ^ ^^^ (^) decreases below the minimum frequency limit ^ ^^^ and thus allowing the activation of protective measures based solely on the value of the frequency ^(^). The third control branch 13 contributes to the reduction of the frequency setpoint ^ ^^^ (^) by decreasing the third frequency instruction ^ ^^^^^ (^) as set out below. Always if ^ ( ^ ) > ^ ^^^^( ^ ) , but also if: a. ^ ^^( ^ ) < ^ ^^ ^^^( ^ ) (if external control loop that usually controls ^ ^^( ^ ) saturates) or b. ^ ^^( ^ ) = ^ ^^ ^^^( ^ ) (if external control loop that usually controls ^ ^^( ^ ) does not saturate), the input of the PI 131 regulator, equal − ^ ^^ ^^^ (^)^ − ^(^) , is then negative, and: c. The proportional term of ^^ − 2 becomes strictly less than zero, and d. The integral term of ^^ − 2 decreases. As a result, the variables ^^^3(^) and ^ ^^^^^ (^) become strictly less than zero and gradually decrease. If ^ ( ^ ) < ^ ^^^^( ^ ) and ^ ( ^ ) < ^^^^^ ^^^^^^(^) , the third frequency instruction ^ ^^^^^( ^ ) is saturated to zero and the action of the third control-command branch 13 disappears. In this situation, ^ ^^ (^) = ^ ^^ ^^^ (^) , because the external control loop that usually controls ^ ^^( ^ ) , is not saturated because ^ ( ^ ) < ^^^^^ ^é^^^^(^ ). The input of the PI 131 regulator, equal ^ ^^ ^^^( ^ )^ − ^ ( ^ ) , is then strictly greater than zero, the proportional term of ^^ − 2 becomes strictly greater than zero, and the integral term of ^^ − 2 gradually increases. As a result, the variable ^^^3 ( ^ ) increases and eventually becomes strictly positive. The variable ^ ^^^^^( ^ ) is equal to ^^^3 ( ^ ) as long as the latter is less than or equal to 0, but ^ ^^^^^( ^ ) remains equal to 0 when ^^^3 ( ^ ) becomes strictly positive, through the saturation function 132 of figure 7. We conclude that the role of the third branch 13 is important only in the case where ^^^^^ ^é^^^^(^ ) The operation of the third branch 13 in this situation can be explained by considering the operating scenario described below, it being understood that this scenario is given for illustrative purposes and not as a limitation, other scenarios being of course potentially observable. In this scenario, the network 3 of figure 7 is initially in a steady state where the inverter 1 injects active power ^ ( ^ ) with ^ ( ^ ) > ^ ^^^ ^^^ and ^ ( ^ ) < , and its set frequency is ^ ^^^( ^ ) > ^ ^^^. It is also assumed that . It is further assumed that there is at least one other generator whose active power injected into the network is strictly lower than its available active power. Considering that a disturbance in the network, such as an increase in the active power consumed by the loads, occurs, then network 3 would reach, in steady state, a new state for which the active power of inverter 1 would be such that ^^^^^ ^é^^^^(^ ) < ^ ( ^ ) ≤ ^ ^^^^( ^ ) , if only the first and second branches 11 and 12 were implemented on inverter 1. According to the previously given definition of the first branch 11, the set frequency would still be ^ ^^^( ^ ) ≥ ^ ^^^and, consequently, no protective measures would be activated. The third branch 13 advantageously makes it possible to avoid the occurrence of such a malfunction, as explained below. Since ^^^^^ ^é^^^^(^ ) < ^ ( ^ ) , the capacitors located upstream of the inverter discharge, which causes the voltage to gradually drop ^ ^^( ^ ) . The input of the PI 131 regulator, equal by becoming negative, the proportional term of ^^ − 2 becomes strictly less than zero, and the integral term of ^^ − 2 gradually decreases. As a result, the variables ^^^3 ( ^ ) and ^ ^^^^^ (^) become strictly less than zero and decrease. And consequently, the frequency setpoint ^ ^^^ (^) decreases. The benefit of decreasing the frequency setpoint ^ ^^^( ^ ) , when ^ ^^^^ ^^^( ^ ) − ^ ( ^ )< 0, is that the other inverters of network 3 are then pushed to increase the active power that they supply (due to the electrical laws that govern the behavior of network 3). This allows the first inverter 1 to decrease its active power ^ ( ^ ) and that the latter becomes equal to ^ ^^^^ ^^^ (^) (ie ^ ^^^^ ^^^( ^ ) − ^ ( ^ ) = 0 ). When ^ ^^^^ ^^^( ^ ) − ^(^) = 0 , the integral term of ^^ − 2 stabilizes; it neither increases nor decreases. Consequently, the variables ^^^3 ( ^ ) and ^ ^^^^^ (^) also stabilize. The key to the third branch of control-command 13 is then to know at what value the auxiliary variable ^ stabilizes ^^^^ ^^^ (^) . Two conditions define the value of ^ ^^^^ ^^^ (^) : a. to stop the voltage drop ^ ^^( ^ ), it is necessary that the active power ^(^) of the first inverter 1 decreases and becomes equal to ^^^^^ ^é^^^^(^), and b. to stabilize the integral term of the PI regulator 131 and the third setpoint ^ ^^^^^ (^) , it is necessary that ^ ( ^ ) becomes equal to the auxiliary variable ^ ^^^^ ^^^ (^). Consequently, the third control branch 13 will force the first inverter 1 to go to the only steady-state operating point that meets the two previous conditions: ^ ( ^ ) = ^^^^^ ^^^ (^ ) = ^ ^^^^ Note that, so that ^ ^^^^ ^^^ (^) is equal to ^^^^^ ^^^^^^(^), the PI 131 regulator will leave the voltage ^ ^^ (^) stabilize slightly below the set value ^ ^^ ^^^ (^), since To avoid the tension ^ ^^( ^ ) does not deviate too far from the set value ^ ^^ ^^^( ^ ), it is better to implement a gain sufficiently large. For example, if the set value ^ ^^ ^^^( ^ ) is constant and to avoid that the tension ^ ^^( ^ ) does not stabilize below a threshold called ^ ^^ ^^^ , you have to choose a gain equal to or greater than: where ^ ^^^ ^ ^^^ is an estimate of the maximum possible difference between the ^ actual available active power ^^^^^ ^é^^^^ and the available active power ^ ^^^^ (^) as measured or estimated. Note that, usually, a control loop (outside the droop 5 control loop) is provided which ensures that the voltage measurement continues ^ ^^( ^ ) is equal to the set value of the DC voltage ^ ^^ ^^^( ^ ) . However, this 'ordinary' control loop saturates and shuts down when ^(^) ≥ ^^^^^ ^^^^^^(^), which is why ^ ^^(^) may stabilize below ^ ^^ ^^^( ^ ) . Using a complementary graphical explanation, and with reference to Figure 8, after a disturbance in the network such as an increase in the active power consumed by the loads, the droop control loops force the generators to converge towards steady-state operating points that belong to specific profiles in the plane (^, ^ ^^^ ). The proposed droop loop 5 results in a profile {^, ^ ^^^} which prevents inverter 1 from converging to an operating point with ^^^^^ ^é^^^^(^ ) < ^ ( ^ ) if there are other generators in the network whose active powers are less than or equal to their available active powers. In other words, if ^^^^^ ^é^^^^(^) < ^(^) ≤ ^ ^^^^ (^), this profile {^, ^ ^^^} according to the third embodiment of the invention is defined by the action of the third branch 13 and consists of a vertical segment 1003 which goes from the point ^ ^^^ ^ at the point of intersection between the segment 1001 corresponding to the first branch 11 and the vertical placed at ^ ( ^ ) = ^^^^^ ^é^^^^ (^ ) (that is, the point It should be noted that this vertical segment 1003 moves from right to left or from left to right, when ^^^^^ ^é^^^^(^ ) varies. It should also be noted that, when this segment 1003 is activated (i.e. when the active power enters the interval ^^^^^ ^é^^^^(^ ) < ^ ( ^ ) ≤ ^ ^^^^( ^ ) ), the segment 1001 corresponding to the first branch 11 contracts, so that the end points of the two segments corresponding to the first and third branches 11 and 13 coincide. Considering the profiles {^, ^ ^^^} of Figure 8, the initial state of the scenario described above could correspond, for example, to point A for the first inverter and to point A' for the second inverter. After a disturbance such as an increase in the active power consumed by the grid loads, and if only the first and second branches 11 and 12 were implemented on the first inverter, the grid could reach a new steady-state state corresponding to point B for the first inverter and to point B' for the second inverter. The action of branch 13 forces the first inverter to decrease its active power, which brings the grid to a state that can correspond, for example, to point C for the first inverter and to point C' for the second inverter.The third control branch 13 operates correctly in systems where the energy source 2 is not directly connected to the capacitors placed upstream of the first inverter 1, i.e. to the extent that the voltage ^. ^^(^) is not directly applied to the source. This is, for example, the case of two-stage photovoltaic power plants, which include a DC / DC converter placed between the photovoltaic array and the capacitors. As already specified above, Figure 11 represents an electrical and control diagram of an embodiment of each of two droop control loops, one (referenced 5) relating to the active power, noted ^, and illustrating a droop control loop according to any one of the embodiments of the invention, and that relating to the reactive power, noted ^, which is not directly concerned by the present invention. A comparison between them of the diagram represented by Figure 11 and each of the diagrams represented in Figures 4, 6 and 7 makes it possible to illustrate the structure of a system 0 consisting of an inverter 1 and an energy source 2.More particularly, the electrical and control-command diagram of Figure 11 makes it possible to illustrate the same system 0 in a different way than Figures 4, 6 and 7. The invention is not limited to the embodiments previously described and extends to all the embodiments covered by the invention. In particular, if the proposed droop loop, according to any of the embodiments described above, is essentially dedicated to the interfacing of an intermittent energy source with the AC electrical network, it should be noted that it can also be used for the interfacing of a non-intermittent energy source with the AC electrical network. Furthermore, the different mathematical functions used to define the different branches described above could be replaced by other functions giving equivalent results.For example, the saturation function 122, defined mathematically as max(aux1(t),0) could be written, equivalently, using logical functions. Similarly, if the aforementioned equations are each given in continuous form as a function of the time variable t, they could be described, equivalently, in continuous or discrete form, for example using the Laplace transform or the z transform, known to those skilled in the art.

Claims

CLAIMS Method for controlling statism for an inverter (1) intended to interface an intermittent energy source (2) and an alternating electrical network (3) by regulating the frequency ^(^) at a connection point (4) of the inverter (1) to the network (3), the method comprising, as a function of: i. a constant ^ ^^^ defining the maximum network frequency limit (3), ii. a constant ^ ^^^ defining the minimum network frequency limit (3), iii. a constant ^ ^^^ ^^^ defining the maximum active power assigned to the inverter (1), iv. a constant ^ ^^^ ^^^ defining the minimum active power assigned to the inverter (1), see an active power measurement ^ ( ^ ) that the inverter (1) supplies to the network (3) at time ^, and vi. a measurement or estimate of the available active power ^ ^^^^( ^ )in the inverter (1) at time ^, ^ a step consisting of implementing: i. A first control-command branch (11), noted ^, configured to calculate a first frequency setpoint ^ ^^^^^ (^) verifying a first control-command function (1001) defined in a plane by the following equation: ii. A second control-command branch (12), denoted ^, comprising the implementation of a first proportional-integral regulator (121) and a first saturation function (122) and being configured to calculate a second frequency setpoint ^ ^^^^^ (^) verifying a second control-command function (1002) defined in the plane by the following equation: ^ ^ ^^^^ , the proportional gain of the first proportional-integral regulator (121) and ^ ^^^^^, the integral gain of the first proportional-integral regulator (121), are parameters greater than zero and predetermined depending on the inverter (1) concerned, and ^ a step consisting of imposing a setpoint value ^ ^^^ (^) frequency at the connection point (4) of the inverter (1) to the network (3) determined as a sum of the first and second frequency setpoints Method according to the preceding claim, in which the implementation of the second control-command branch (12): ^ is furthermore a function of: i. a measurement of the direct voltage at the input of the inverter (1), and ii. a set value of the direct voltage ^ ^^ ^^^(^) at the input of the inverter (1), and ^ further comprises the implementation of a first proportional regulator (123), the first proportional-integral regulator (121), the first saturation function (122) and the first proportional regulator (123) being configured to impose, at the second frequency setpoint ^ ^^^^^( ^ ) , a value defined by a variant of the second control-command function (1002), said variant taking the form of the following equation: ^ ^ ^^^, the proportional gain of the first proportional regulator (123), is a parameter strictly greater than zero and predetermined as a function of the inverter (1) concerned. Method according to any one of the preceding claims, in which the step of implementing the first and second branches (11 and 12) further comprises, as a function of: i. a measurement of the DC voltage at the input of the inverter (1), and ii. a set value of the direct voltage ^ ^^ ^^^ (^) at the input of the inverter (1), ^ the implementation of a third control-command branch (13), noted ^ , comprising the implementation of a second proportional-integral regulator (131) of a second saturation function (132) and of a second proportional regulator (133) and being configured to calculate a third frequency setpoint ^ ^^^^^ (^) verifying a third control-command function (1003) defined in the plane by the following equation: ^ ^^^^^( ^ ) = min ( ^^^3 ( ^ ) , 0 ) , with ^ ^ ^^^^, the proportional gain of the second proportional-integral regulator (131), ^ ^^^^^, the integral gain of the second proportional-integral regulator (131) are parameters greater than zero and predetermined depending on the inverter (1) concerned, the proportional gain of the second proportional regulator (133), is a parameter strictly greater than zero and predetermined according to the inverter (1) concerned, and in which: the set value ^ ^^^ (^) of frequency at the connection point (4) of the inverter (1) to the network (3) is determined as the sum of the first, second and third frequency setpoints, ^ ^^^ (^) = ^ ^^^^^ (^) + ^ ^^^^^( ^ ) + ^ ^^^^^(^). A method according to any preceding claim, wherein at least one, preferably each, of said parameters may be predetermined by an analysis of the time response of the inverter and / or by an analysis of the frequency response of the inverter and / or by implementing a pole placement method. A method according to any preceding claim, wherein the step of imposing the setpoint value ^ ^^^(^) of frequency at the point of connection of the inverter to the network comprises supplying said value to at least one internal control loop (10) of the inverter (1). Computer program product comprising instructions, which when carried out by at least one processor executes the steps of the droop control method according to any one of the preceding claims. Control unit for an inverter (1) intended to interface an intermittent energy source (2) and an alternating current grid (3) by regulating the frequency ^(^) at a point of connection of the inverter (1) to the network (3), the control unit comprising electronic and / or microelectronic components configured to implement the droop control method according to any one of claims 1 to 5.Inverter (1) intended to interface an intermittent energy source (2) and an alternating electrical network (3) by regulating the frequency ^(^) at a connection point of the inverter (1) to the network (3) and comprising at least one of: ^ a readable non-transient medium comprising instructions, which when carried out by at least one processor, executes the steps of the droop control method according to any one of claims 1 to 5, and. ^ a control unit according to claim 7. Inverter (1) according to the preceding claim, free from a communication link with another generator or a server connected to the alternating current network. Inverter (1) according to any one of the two preceding claims, in which a voltage at the point of connection of the inverter to the network is single-phase or polyphase.

Images