Control of converters in an AC grid powered by renewable energy sources.
The method stabilizes AC grids by regulating load power based on measured grid voltage, addressing voltage and frequency fluctuations in island grids powered by renewables, ensuring grid stability through converter control.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- ABB (SCHWEIZ) AG
- Filing Date
- 2023-05-30
- Publication Date
- 2026-06-04
AI Technical Summary
Island grids and microgrids powered by renewable energy sources face fluctuations in grid voltage and frequency due to environmental conditions, necessitating effective load power regulation to prevent voltage collapse.
A method for controlling converters that supply power to loads in AC grids, using a controller to regulate load voltage based on measured AC grid voltage and load voltage, independent of communication between source and load converters, by determining a load voltage reference and adjusting converter output power to stabilize the grid.
Minimizes fluctuations in grid voltage and frequency by adapting load power to available energy, maintaining grid stability without communication links, applicable to various converter topologies.
Smart Images

Figure 2026518318000001_ABST
Abstract
Description
[Technical Field]
[0001] The present invention relates to the field of AC electric grids that operate without a stable connection to a large-scale grid, such as island grids or microgrids. More specifically, the present invention relates to a method, computer program, computer-readable medium, and controller for controlling a converter for a load connected to an AC grid powered by at least one renewable energy source. Furthermore, the present invention relates to a system comprising such a renewable energy source, an AC electric grid, and a converter. [Background technology]
[0002] Hydrogen production plants may be located in remote, isolated areas, without being connected to the national power grid, using renewable energy sources such as wind and solar power. In such cases, since there is no common voltage source, wind and / or solar inverters are responsible for forming the grid and maintaining its voltage amplitude and frequency under different operating conditions. Therefore, since the energy production of such microgrids depends solely on environmental conditions, such as constantly fluctuating solar irradiance and wind speed, it may be necessary to match energy consumption with production to avoid grid voltage collapse.
[0003] Generally, load power regulation in island operations is achieved by controlling the grid voltage, i.e., the magnitude and frequency of the output voltages of the converters forming the grid. However, this usually results in fluctuations in grid voltage and frequency whenever environmental conditions change, which can be undesirable.
[0004] Load power regulation within such grids, primarily as described in CN110299722A, focuses on DC distribution and its challenges and problems, as, for example, batteries, fuel cells, electrolyzers, and photovoltaic arrays either generate DC or behave as DC loads. Furthermore, a common assumption is that unlimited energy sources are available in the form of a large-scale distribution grid, in parallel with renewable energy sources.
[0005] Furthermore, conventionally, grid formation control has been implemented by regulating the grid voltage on the side of the converter of the renewable energy source. For example, U.S. Patent No. 11005270(B2) proposes a grid formation control method for a group of solar inverters in island operation based on AC grid voltage regulation via a frequency-power characteristic curve. [Overview of the Initiative]
[0006] The objective of the present invention is to simplify the operation of island grids and / or microgrids powered by renewable energy sources.
[0007] This objective is achieved by the subject matter of the independent claim. Further exemplary embodiments are evident from the dependent claims and the following description.
[0008] One aspect of the present invention relates to a method for controlling a converter that supplies power to a load, the load being connected to an AC grid via the converter, the AC grid being powered by at least one renewable energy source.
[0009] A converter may be adapted to convert AC current from the grid into DC or AC current supplied to a load. The converter may include thyristors and / or transistors for switching the current. This method may be implemented by the converter's controller.
[0010] In general, a load may be any device and / or system adapted to receive variable input power. For example, the load may be an electrolyzer, i.e., a device adapted to produce hydrogen from water by electrolysis. As a further example, the load may be a device and / or system having an internal battery for balancing variable input power, such as a data center.
[0011] At least one renewable energy source may comprise a photovoltaic panel, a wind turbine, or a hydro turbine. Each renewable energy source may comprise a source converter for supplying an AC current to the grid. One or more source converters of the renewable energy source may be static converters, i.e., such converters are controlled to supply all the electric power generated by at least one renewable energy source to the AC grid, particularly independently of the natural fluctuations of the renewable energy.
[0012] The AC grid may be a grid independent of a large-scale grid, also called a microgrid, and / or may have a floating voltage and / or power. The AC grid need not be of a single-phase type and may be a three-phase grid. Also, the converter supplying power to the load, and / or the converter of at least one renewable energy source may be a three-phase converter.
[0013] According to one embodiment of the invention, the method comprises receiving a measured AC grid voltage measured within the AC grid and a measured load voltage measured at the output of the converter. The controller may receive measurement signals from voltage sensors at the input and output of the converter and / or may obtain measurements of the AC grid voltage and the load voltage.
[0014] According to one embodiment of the invention, the method further comprises determining the magnitude of the fundamental wave positive-phase component of the measured AC grid voltage, which may be regarded as the fundamental and / or grounded sine wave component of the measured AC grid voltage.
[0015] According to one embodiment of the present invention, the method further comprises determining a load voltage reference from magnitude and nominal load voltage reference, wherein the load voltage reference decreases when the magnitude decreases and / or increases when the magnitude increases. In this way, the load voltage reference is based on the magnitude of the fundamental positive-sequence component of the measured AC grid voltage. Specifically, a scaling factor may be determined from the magnitude, which is multiplied by the nominal load voltage reference to calculate the load voltage reference.
[0016] According to one embodiment of the present invention, the method further comprises determining a voltage error by subtracting a measured load voltage from a load voltage reference. The voltage error can be used as an indicator of how much the input power from the AC grid deviates from the nominal input power.
[0017] According to one embodiment of the present invention, the method further comprises controlling the output power of a converter using a voltage error. For example, the firing angle of the converter's thyristors may be determined and shifted based on the voltage error. The duty cycle of the converter may also be set based on the voltage error.
[0018] This method can be considered a control method for regulating the power of devices such as electrolyzers connected to an AC microgrid that relies entirely on renewable energy sources. This method allows for load power regulation without communication between the renewable energy source source converter and the load converter by monitoring the AC grid voltage. This method is independent of the converter topology and can be applied to any converter type controller adapted to reduce load power. With this method, fluctuations in the magnitude and frequency of the AC grid voltage are minimized because the load adapts itself according to energy availability.
[0019] According to one embodiment of the present invention, the method further comprises scaling the AC grid voltage such that, when the AC grid voltage is equal to the nominal voltage, the scaled AC grid voltage has a phase voltage peak of 1. The AC grid voltage is scaled to correspond to a nominal phase voltage peak of 1. In this way, the following method steps become independent of the nominal voltage.
[0020] According to one embodiment of the present invention, the method further comprises converting an AC grid voltage (optionally scaled) into a two-component space vector, and extracting the fundamental positive-sequence component of the AC grid voltage from the space vector. The phase voltages of the AC grid voltage can be converted into complex-valued space vectors using the Clarke transform. The fundamental positive-sequence component can be readily extracted from the space vector.
[0021] According to one embodiment of the present invention, the method further comprises low-pass filtering the magnitude of the fundamental wave positive-sequence component. The fundamental wave positive-sequence component may be low-pass filtered to remove fast changes in the AC grid voltage that may be caused by fast changes in renewable energy sources. Here, the term “fast” may refer to the time scale that is filtered out by the corresponding low-pass filter.
[0022] According to one embodiment of the present invention, the method further comprises low-pass filtering of the measured load voltage. The measured load voltage may be low-pass filtered to remove the power consumption of a rapidly changing load. Here again, the term “rapid” may refer to the time scale that is filtered out by the corresponding low-pass filter.
[0023] According to one embodiment of the present invention, the method further comprises applying a function to the magnitude of the fundamental positive-sequence component to determine the amplified magnitude, and the load voltage reference is determined from the amplified magnitude. The load voltage reference does not need to be proportional to the magnitude, nor does it need to be linearly dependent on the magnitude. This function can be considered as a magnitude-dependent amplification function, which is used to set the strength of the control intervention according to the magnitude.
[0024] The amplified magnitude can be a scaling factor and / or can be multiplied by the nominal voltage reference. The amplified magnitude is used to scale the nominal voltage reference in order to determine the load voltage reference.
[0025] According to one embodiment of the present invention, the function is a power function having an exponent of 1.5 to 3.5. For example, the exponent may be 2 or 3. The power function may be selected based on the relationship between load voltage and load power. For example, an electrolyzer is usually not a linear system, and as a result, its power reduction is not exactly inversely proportional to the square of the voltage. In the case of an electrolyzer system, good performance may be achieved by setting the exponent to 3. In any case, the power function may be selected based on a load model to achieve optimal performance with a particular load.
[0026] According to one embodiment of the present invention, the method further comprises limiting the amplified magnitude to 0 to 1. This is to ensure that the load voltage reference remains within 0 and the nominal voltage reference.
[0027] According to one embodiment of the present invention, the method further comprises determining a control variable from a voltage error by applying a PI controller to the voltage error. For example, the control variable may be used to control the duty cycle of a load converter. For example, the control variable may be the duty cycle reference of the load converter or the phase control angle of a thyristor rectifier. Specifically, the final phase control angle of a thyristor rectifier may be determined from the control variable.
[0028] According to one embodiment of the present invention, the AC grid is powered by at least one renewable energy source. The maximum power generated by the at least one renewable energy source is less than 10 MW, meaning that the system with the AC grid may be a small or medium-sized system. The grid may be considered an island grid. As already mentioned, the AC grid may be isolated from a large grid that stabilizes voltage, power, phase angle, and / or frequency.
[0029] According to one embodiment of the present invention, the load is an electrolyzer. Specifically, hydrogen production in the electrolyzer can be easily configured according to the available power.
[0030] According to one embodiment of the present invention, the converter is an active rectifier, and the measured load voltage is a DC voltage. The load may be a DC load. In the case of an AC load, the measured load voltage may be the DC link voltage of the DC link between the load rectifier and the load inverter.
[0031] A further aspect of the present invention relates to a method for controlling power supplied to at least two loads connected to an AC grid powered by at least one renewable energy source. Each of the loads is controlled by the method described herein. Furthermore, the loads can be controlled independently of one another. This may mean that the power control of one load does not depend on control variables determined for another load. Specifically, multiple parallel load converters that can be connected to the same grid connection point can be controlled by the method without additional control parameter adjustments.
[0032] Further aspects of the present invention relate to a computer program adapted to carry out the method described herein when executed by a processor, and a computer-readable medium in which such a computer program is stored. The computer program may be part of the control software of a controller of a load converter.
[0033] Computer-readable media may be floppy disks, hard disks, USB (Universal Serial Bus) storage devices, RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory), or flash memory. Computer-readable media may also be data communication networks that enable the download of program code, such as the Internet. Generally, computer-readable media may be non-temporary or temporary media.
[0034] Further aspects of the present invention relate to a controller for controlling power supplied to a load via a converter, the controller being adapted to carry out the method described herein. It should be noted that the method may be implemented at least partially in hardware, for example, in a DSP or FPGA.
[0035] A further aspect of the present invention relates to an electrical system comprising at least one renewable energy source, a load, a converter for supplying power to the load, an AC grid interconnecting the at least one renewable energy source with the load via the converter, and a controller for controlling the converter, such as those described herein.
[0036] Please understand that the characteristics of the methods described above and below may be the characteristics of the systems described above and below, and vice versa.
[0037] These and other aspects of the present invention will become apparent from and be illustrated by the embodiments described below.
[0038] The subject matter of the present invention will be described in more detail below with reference to exemplary embodiments shown in the accompanying drawings. [Brief explanation of the drawing]
[0039] [Figure 1] A schematic diagram of an electrical system according to one embodiment of the present invention. [Figure 2] A block diagram showing a method and controller according to one embodiment of the present invention. [Figure 3] A schematic diagram of an electrical system according to a further embodiment of the present invention. [Figure 4] A schematic diagram of an electrical system according to a further embodiment of the present invention. [Modes for carrying out the invention]
[0040] The symbols used in the drawings, and their meanings, are listed in a concise format within the explanation of the symbols. As a general rule, the same symbol is assigned to identical parts within a drawing.
[0041] Figure 1 shows an electrical system 10 comprising an AC grid 12, a renewable energy source 14, and a load 16 in the form of a hydrogen electrolyzer.
[0042] The renewable energy source 14 includes a source converter 18 that supplies AC current to the AC grid. The load 16 includes a load converter 20 that takes power from the AC grid and supplies power to the load 16.
[0043] As shown in the figure, the AC grid 12 may be a three-phase grid, and the converters 18 and 20 may be three-phase converters. The AC grid 12 may be a microgrid, that is, it may be isolated from a larger distribution grid that can be used to stabilize the power and frequency within the AC grid 12.
[0044] Furthermore, the battery system 22 may be connected to the AC grid, for example, via a further converter 24. The battery system 22 may be used to balance the power within the AC grid. However, such balancing may not be possible when the batteries in the battery system 22 are empty or full.
[0045] Figure 1 also shows the controller 26 of the load converter 20, which controls the load converter 20 to reduce its output power to the load 16 when the voltage in the grid drops.
[0046] Figure 2 shows a block diagram of the controller 26, which also illustrates the methods implemented by the controller 26.
[0047] In this method, the power supplied to the load 16, such as an electrolyzer, is regulated. The load 16 is connected to an AC grid 12 that can rely entirely on a renewable energy source 14, such as a solar light source or a wind power source. In the AC grid 12, the AC grid voltage can be formed solely by a static converter 18, and the amount of usable energy is entirely dependent on environmental conditions such as solar irradiance and / or wind speed, which can be constantly changing.
[0048] With this method, the load power can be adjusted to match the available power by controlling the operating point of the load converter 20, particularly without a communication link between one or more source converters 18 and the load converter 20. This technique can eliminate, or at least reduce, fluctuations in grid voltage and frequency due to changing environmental conditions, and thus keep the AC grid 12 more stable at all times.
[0049] This method is not specific to converter topology and can thus be used with any type of converter 20 capable of reducing the load power. The converter 20 may be, for example, a thyristor rectifier, a diode and active voltage source rectifier with an optional buck DC-DC converter as a second conversion stage, a pulse-width modulated current source rectifier, a diode rectifier fed by a transformer with a remotely adjustable tap changer. Specifically, a conventional 12-pulse thyristor rectifier may be used and, hereinafter, such a converter may be referred to as an example.
[0050] In block 30, the measured AC grid voltage v abc is received, measured within the AC grid 12, for example at the connection point, and / or at the input of the converter 20 to the grid 12. In the case of a three-phase grid 12, it may be sufficient to measure only the line-to-line voltage between two lines of the AC grid 12 that can be converted into three-phase voltages.
[0051] The AC grid voltage v abc is scaled such that when the AC grid voltage v abc is equal to the nominal voltage, the scaled AC grid voltage v abc,pu has a phase voltage peak of 1. The three-phase voltage v abc is scaled to a value per unit (p.u.) such that the nominal phase voltage peak corresponds to the value of the scaled AC grid voltage v abc,pu .
[0052] In block 32, the scaled AC grid voltage v abc,pu is converted into a space vector v xy having two components. The space vector v xy is in the stationary coordinate system and can be generated using a standard Clarke transformation (abc to xy).
[0053] In block 34, the fundamental positive-phase component v abc of the AC grid voltage v xy,1+ is the space vector v xyIt is extracted from the load voltage reference v. dc * This is done to avoid the effects of voltage harmonics and phase imbalance in the system (see below).
[0054] In block 36, the measured AC grid voltage v is represented by |u|, etc. abc The fundamental wave positive-sequence component v xy,1+ Size v pk The magnitude v is determined. Therefore, the magnitude v pk This is used as the quantity to be monitored.
[0055] In the optional block 38, the fundamental wave positive-sequence component v xy,1+ Size v pk The data is filtered. This filtering may include low-pass filtering (LPF), period averaging, and / or moving averaging.
[0056] In blocks 40 to 46, the fundamental wave in-sequence component v xy,1+ Size v pk or filtered size
[0057]
number
[0058] Based on this, load voltage reference v dc * This is generated.
[0059] In block 40, to determine the amplified magnitude, the fundamental wave positive-sequence component v xy,1+ Size v pk or filtered size
[0060]
number
[0061] The function is applied to it.
[0062] This function is the AC grid voltage v abc When the load voltage reference v drops below its nominal value, dc * This defines how rapidly it decreases. For example, the function is a power function u with an exponent of 1.5 to 3.5. x Therefore, good performance can be achieved by setting the exponent x to 2 or 3.
[0063] In block 42, the amplified magnitude is limited to 0 to 1. The output of block 42 is the load voltage v dc * This is a scaling factor, and is limited to values between 0 and 1.
[0064] In block 44, the nominal load voltage reference v provided by block 46 is used. dc The instantaneous load voltage reference v is multiplied by a limited scaling factor. dc * This will be decided.
[0065] In block 48, the measured and optionally filtered load voltage v dc Use the load voltage reference v dc * The voltage error e is determined by subtracting it from the given value.
[0066] Block 50 shows the measured load voltage v measured at the output of converter 20. dc Receive. Optionally, measure the load voltage v to reduce undesirable oscillation and noise. dc The load voltage is low-pass filtered (LPF).
[0067]
number
[0068] It may also be done.
[0069] In block 52, the control variable y is determined from the voltage error e by applying a PI controller to the voltage error e. The voltage controller may be a standard PI controller that outputs the control variable y. The control variable y may be a duty cycle reference for the converter 20, and / or the control variable y may control the duty cycle of the converter 20.
[0070] As shown in block 54, in the case of a thyristor rectifier, the control variable y can be used to determine the phase control angle α of the thyristor. Since a larger phase control angle α reduces the load voltage of the thyristor rectifier, this angle is inverted by subtracting the control variable from 180° (provided by block 56) to obtain the phase control angle α.
[0071] Finally, the converter 20 is controlled by a control variable y having a phase control angle α, in particular, by generating corresponding switching signals and applying them to the semiconductor switches of the converter 20.
[0072] Figure 3 shows an embodiment of a system 10 comprising a 1 MVA photovoltaic inverter 18 that supplies power to a 12-pulse thyristor rectifier 20 from a renewable energy source 14, along with an electrolyzer as a load 16. A photovoltaic array with an open-circuit voltage of approximately 1200 V is connected to the DC input of the photovoltaic inverter 18. The photovoltaic inverter 18 operates in grid-forming mode and generates a three-phase 50 Hz grid voltage with a line-to-line rms voltage of 690 V at the output of an LCL type supply filter 60. The 12-pulse thyristor rectifier 20 is connected to the grid interconnection point PCC through a Y-delta-Y transformer 62 that steps down the 690 V primary voltage to 400 V on the secondary side. Each rectifier bridge 20a, 20b is connected to an AC-side three-phase filter inductor (L ac ) and DC inductors (L) at both terminals on the DC side dc The individual rectifier bridge DC terminals are connected in parallel, and the electrolyzer 16 is connected between the common positive and negative rails of the rectifier 20.
[0073] During steady-state operation with 100% solar irradiance and a nominal load of 1MW, the DC input voltage of the solar inverter 18 is approximately 1100V, which is appropriate for the inverter 18 to maintain the nominal grid voltage. When the irradiance suddenly drops to 25% while the load power remains the same, the DC voltage of the inverter 18 begins to drop. The controller of the solar inverter 18 attempts to keep the voltage magnitude and frequency at the PCC constant, but this is only possible if the voltage of the solar array 14 remains higher than the peak of the generated line-to-line voltage. At some point, the DC voltage becomes too low, and therefore the amplitude of the grid voltage must be reduced. This is detected by the controller 26 of the load converter 20, and the controller increases the phase control angle α and begins to reduce the load power to prevent grid voltage collapse.
[0074] Figure 4 shows one embodiment of the system 10, which comprises two loads 16, each connected to the AC grid 12 via a load converter 20. Each of the loads 16 is designed as shown in Figure 3 and connected to the grid connection point PCC.
[0075] The power supplied by the solar inverter 18 is now shared by two 12-pulse thyristor rectifiers 20, each having an independent controller 26. There is no inter-unit communication between the converters 20 and their controllers 26. Each controller 26 implements this control method independently.
[0076] Although the present invention is illustrated and described in detail in the drawings and the foregoing description, such illustrations and descriptions should be considered illustrative or typical and not limiting, and the present invention is not limited to the embodiments disclosed. Other variations of the disclosed embodiments can be understood and brought about by a person skilled in the art by practicing the claimed invention by examining the drawings, this disclosure and the appended claims. In the claims, the term “equipped with” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude plural. A single processor or controller or other unit may satisfy the functions of several items described in the claims. The mere fact that several means are described in different dependent claims does not indicate that a combination of these means cannot be used advantageously. Reference numerals in the claims should not be considered limiting in scope. [Explanation of symbols]
[0077] 10 Electrical Systems 12 AC grid 14 Renewable energy sources 16 load 18 Source Converters 20 Load Converters 22 Battery Systems 24 Battery Converters 26 Controllers 30-56 Controller Blocks v abc Measured AC grid voltage v abc,pu Scaled AC grid voltage v xy Spatial vectors v xy,1+ Fundamental wave positive phase component v pk Magnitude of the fundamental wave positive-sequence component
[0078]
number
[0079] Filtered size v dc * Load voltage reference e Voltage error v dc Measured load voltage
[0080]
number
[0081] Filtered load voltage y is a control variable. α Phase control angle 60 LCL filter 62 Transformers PCC grid connection point 20a Rectifier Bridge 20b Rectifier Bridge
Claims
1. A method for controlling a converter (20) that supplies power to a load (16) connected to an AC grid (12) powered by at least one renewable energy source (14), The measured AC grid voltage (v) measured within the AC grid (12) abc ), and the measured load voltage (v) measured at the output of the converter (20) dc ) receiving, The measured AC grid voltage (v abc The fundamental wave positive-sequence component (v) xy,1+ ) size (v pk ) to decide, the size (v pk ), and the nominal load voltage reference (v dc * ), to determine a load voltage reference (v dc * ), where the load voltage reference (v pk ), when the size (v dc * ) decreases, decreases, The aforementioned load voltage reference (v dc * ) from the measured load voltage (v dc The voltage error (e) is determined by subtracting ), A method comprising controlling the output power of the converter (20) using the voltage error (e).
2. The AC grid voltage (v abc When ) is equal to the nominal voltage, the scaled AC grid voltage (v abc,pu ) such that the AC grid voltage (v abc The method according to claim 1, further comprising scaling ).
3. The AC grid voltage (v abc ) is a two-component spatial vector (v xy Converting to ) and The aforementioned spatial vector (v xy ) from the AC grid voltage (v abc The aforementioned fundamental wave positive-sequence component (v) xy,1+ The method according to claim 1 or 2, further comprising extracting ) and
4. The aforementioned fundamental wave positive-sequence component (v xy,1+ The aforementioned size (v pk ) to be low-pass filtered, and / or The measured load voltage (v dc ) Low-pass filtering The method according to any one of claims 1 to 3, further comprising:
5. To determine the amplified magnitude, the function is defined as the fundamental wave positive-sequence component (v xy,1+ ) size (v pk The method according to any one of claims 1 to 4, further comprising applying to the above, wherein the load voltage reference is determined from the amplified magnitude.
6. The method according to claim 5, wherein the function is a power function having an exponent of 1.5 to 3.
5.
7. The method according to claim 5 or 6, further comprising limiting the amplified magnitude to 0 to 1.
8. The system further comprises applying a PI controller to the voltage error (e) to determine a control variable (y) from the voltage error (e), and controlling the converter (20) using the control variable (y), The method according to any one of claims 1 to 7, wherein the control variable (y) controls the duty cycle of the converter.
9. The AC grid (12) is powered by at least one renewable energy source (14) and / or The maximum power generated by the at least one renewable energy source (14) is less than 10 MW, and / or The method according to any one of claims 1 to 8, wherein the AC grid (12) is an island grid.
10. The load (16) is an electrolyzer and / or The converter (20) is an active rectifier, and the measured load voltage (v dc The method according to any one of claims 1 to 9, wherein ) is a DC voltage.
11. A method for controlling the power supplied to at least two loads (16) connected to an AC grid (12) powered by at least one renewable energy source (14), Each of the above loads (16) is controlled by the method described in any one of claims 1 to 10, The loads (16) are controlled independently of each other.
12. A computer program adapted to carry out the method described in any one of claims 1 to 11 when executed by a processor.
13. A computer-readable medium storing the computer program described in claim 12.
14. A controller (26) for controlling a converter, the controller being adapted to carry out the method according to any one of claims 1 to 11.
15. At least one renewable energy source (14) and A converter (20) for supplying power to the load (16), An AC grid (12) interconnects the at least one renewable energy source (14) with the converter (20), A controller (26) according to claim 14 for controlling the converter and An electrical system (10) equipped with the following: