Wec controller, method, and system

Abstract

A controller (10) for a gyroscope structure (2), which is associated with a floating hull (3) and equipped with an electric converter (9) suitable for converting the rotational energy of the floating hull (3) into electrical energy, receives a disturbed output state (x) including the operating variables of the gyroscope structure (2) as input to determine a drive signal (u) for the electric converter (9), the drive signal (u) comprising: a first signal portion (v) determined using a predictive control model of the gyroscope structure (2) calculated based on the disturbed output state (x); and a second signal portion (v * The tube convergence is determined using the parameter deviation (r) of the operated variable of the disturbed output state (x), which is calculated relative to the operated variable of the undisturbed nominal output state (ZNP) of the gyroscope structure (2).

Description

WEC controllers, methods and systems Technical Field

[0001] This invention relates to a controller for a WEC (Wave Energy Converter) energy conversion system, which includes a generator or WEC capable of generating electrical energy from ocean waves.

[0002] The present invention also relates to a method for controlling an electrical converter of a gyroscope structure associated with a floating hull and a related WEC system. Background Technology

[0003] As is well known, wave energy is one of the main sources of renewable energy, and many large power plants have recently been developed and built to convert wave energy into electricity. Some power plants (such as inertial WEC or ISWEC power plants) use reactive bodies or PTOs, which utilize the inertia of large masses to generate a reaction and extract its electricity.

[0004] Inertial conversion systems, including those for floating hulls, are known. The floating hull is anchored to the seabed and equipped with directional gyroscope converters, each of which is connected to an electric generator. The generator converts the rotational energy caused by the hull's swaying and wave energy into electrical energy through the motion of a flywheel.

[0005] In this configuration, the gyroscope structure comprises a gyroscope and flywheel associated with the hull via a suspension system, and an electric converter connected to a rotational axis substantially orthogonal to the flywheel's principal axis of inertia. This converter includes an electric motor controlled by a driver / inverter, which is coupled to the rotational axis via suitable joints and gears. Thus, by applying a reaction resistance driving torque that primarily acts as a damper, and by operating in the even quadrant of the driver / inverter's VI diagram, electricity can be generated by the electric motor, rather than being supplied to it.

[0006] To maximize the extracted power and increase the efficiency of the structure, the opposing resistance driving torque must be properly adjusted.

[0007] Classical control systems such as PID (Proportional-Integral-Derivative) controllers are known and widely used in industry, typically operating as controllers for SISO (Single-Input Single-Output) systems. While satisfactory in many respects, systems with PID controllers also have drawbacks. In fact, wave motion is described in the literature as a stochastic process with statistically distributed characteristics, known as the "JONSWAP distribution," whose parameters depend on the sea area of ​​interest. In the case of a PID controller, the control parameters are updated by a preset gain schedule based on ocean state forecasts. Therefore, the tabulated values ​​used as control parameters may differ from those required by the actual wave motion, inevitably leading to associated losses in energy extraction.

[0008] The use of controllers with state evolution models for dynamic systems is known. Such dynamic systems employ Model Predictive Control (MPC). The MPC controller is described in the paper "A comparison of WEC control strategies" by D. Wilson et al., Sandia National Labs, Albuquerque, New Mexico, Tech. Rep. SAND2016-4293, April 2016.

[0009] In its most general form, the MPC controller is based on feedback from the state and a control law that is dynamically calculated by minimizing an appropriate cost function used to optimize the system state.

[0010] MPC controllers differ from PID controllers in the following essential aspects:

[0011] a) The control law or control function is based on the solution of the Euler-Lagrange equation. In classical optimization theory, this equation produces a time function as the solution, and the result is a fixed part of the cost function (as the upper extremum).

[0012] b) The cost function internally includes terms related to the state, input signals, and kinetic energy associated with the state and control signals, and it is typically a convex function;

[0013] c) The cost function typically includes terms that become null when the desired state is reached, and terms that become null when the energy of the control action is minimized.

[0014] Furthermore, the control law of the MPC controller can define dynamic constraints to minimize the cost function. The control law of the MPC controller is based on the solution of the Euler-Lagrange equations and is essentially a time function, while minimizing the error relative to the desired state and relative to the energy involved in obtaining that result.

[0015] It is known to control WEC systems using an MPC controller with a gyroscope-based state evolution model. This allows for a single set of control parameters relevant to all wave conditions and the location of the power plant to be installed. The accuracy of the state evolution model and the estimation of the relevant parameter set affect the performance of the control system.

[0016] In other words, these MPC controllers are optimized for systems with fixed and predefined parameters, and therefore are less efficient for systems affected by variations in these parameters or by random disturbances such as wave conditions. Similarly, ocean conditions not considered during the calibration of MPC controller parameters can also lead to suboptimal energy extraction conditions, i.e., poor performance.

[0017] The known solution is described in the article “Optimizing energy production of an Inertial Sea Wave Energy Converter via Model Predictive Control” by BRACOO G et al., Control Engineering Practice, Pergamon Press, Oxford, GB- vol. 96, 17 January 2020 - XP086048062.

[0018] The fundamental technical problem of this application is to design a control device with functional and structural features of a gyroscope structure in order to allow for the reduction of errors caused by wave motion modeling and gyroscope structure, thereby allowing for the maximization of extracted energy and thus overcoming the shortcomings mentioned in the prior art. Summary of the Invention

[0019] The fundamental solution of this invention is to drive the future evolution of the operational variables that define the structural state of the gyroscope in a constrained manner, thereby improving the robustness and efficiency of the control device. Attached Figure Description

[0020] Further features and advantages of the invention will become apparent from the following description of preferred embodiments of the system and its variations, provided for illustrative purposes with reference to the accompanying drawings, in which:

[0021] - Figures 1 and 2 schematically show the structure of the inertial type WEC system (ISWEC) and the gyroscope, respectively;

[0022] - Figure 3 is a block diagram of the controller manufactured according to the present invention;

[0023] Figure 4 schematically illustrates the ideal and actual output state evolution of a tube convergence model applied to a system with two state variables in a Cartesian diagram.

[0024] - Figure 5 shows a block diagram of a second embodiment of the controller manufactured according to the present invention;

[0025] - Figure 6 schematically shows a simulation of the trend of extracted power over time for an inertial WEC conversion system with a controller according to an embodiment of the present invention;

[0026] Figures 7 through 9 show some details of the graph in Figure 6. Detailed Implementation

[0027] Referring to these figures, Figure 1 shows an overview of an inertial WEC system (i.e., ISWEC), which includes a floating hull 3 and a pair of identical and independent gyroscope structures 2, which are arranged symmetrically to balance the forces relative to the floating hull 3. Each gyroscope structure 2 includes an electro-converter 9 suitable for converting the rotational energy of the floating hull 3 into electrical energy.

[0028] In the schematic form shown in Figure 1, the floating hull 3 is substantially symmetrical with respect to the rolling axis X, and has a first gyroscope structure 2 and a second gyroscope structure 2' in which the gyroscope 6 is highlighted. The second gyroscope structure 2' is shown with a protective cover 7 suitable for covering the underlying flywheel. In the following description, reference will be made to the first gyroscope structure 2, which includes the gyroscope 6.

[0029] The floating hull 3 is shaped to allow rotation about the pitch axis Y, where the pitch angle is... and pitch speed This is caused by the angular momentum forced by wave motion. In Figures 1 and 2, wave motion is determined by direction. The arrow above indicates this.

[0030] The floating hull 3 is anchored to the seabed in a conventional manner, not shown in the diagram, with the roll axis X being substantially parallel to the direction of wave motion. The roll axis X is perpendicular to the pitch axis Y. The floating hull 3 also has a yaw axis Z that is substantially perpendicular to the plane P defined by the roll axis X and the pitch axis Y.

[0031] As schematically shown in Figures 1 and 2, the gyroscope 6 is equipped with a flywheel 8, which has a base fixed to the bottom of the floating hull 3, and the flywheel 8 can rotate around a axis parallel to the rolling axis. First axis and parallel to the yaw axis The second axis Free rotation. Furthermore, flywheel 8 has a free rotation relative to the first axis. First speed and relative to the second axis Second speed The flywheel 8 is properly constrained to the gyroscope structure 2 via bearings, which in turn is constrained to the floating hull 3. Due to wave motion, the entire assembly moves at the pitch speed of the floating hull 3. Around the pitch axis Rotation. Through the gyroscopic effect, the pitch angle of the floating hull 3... It can be transformed to revolve around the first axis Angle of motion or precession angle Precession angle This allows the generation of precession torque supplied to the power converter 9.

[0032] The electric converter 9, schematically shown in the figure, includes a motor associated with a driver / inverter and controlled by a controller 10 via a drive signal. Drive, the drive signal To counteract precession torque and maximize the extracted power.

[0033] Figure 3, through a block diagram, shows a portion of the inertial conversion system WEC 1 associated with the gyroscope structure 2 and controller 10 designed according to the present invention in Figure 1.

[0034] The gyroscope structure 2 is represented by a real power plant block 20, i.e., a nonlinear system, which includes elements suitable for representing the actual states of the gyroscope structure 2. Structure block 21. Structure block 21 receives drive signals. To generate an undisturbed output state including gyroscope structure 2. The vector.

[0035] Disturbances and / or interferences ( ) and the undisturbed output state ( The variables are added together to define the perturbed output state, which includes the operating variables of gyroscope structure 2. Disturbance This includes a collection of external and internal disturbances to the gyroscope structure 2, such as the external driving force of wave motion, which can be filtered by a transfer function and added to the mooring effect and other disturbance elements / forces.

[0036] Controller 10 receives the disturbed output state As input, to generate a drive signal suitable for driving or activating the electrical converter 9. In its most general sense, controller 10 is of the TRMPC class, which is an acronym for Tube-Based Robust Model Predictive Control.

[0037] In the first embodiment, the disturbed output state is defined. The operational variables can be represented in the following vector matrix form:

[0038]

[0039] in:

[0040] It is the pitch angle. It is the pitch speed;

[0041] It is around the first axis The precession angle, It is the precession velocity.

[0042] Controller 10 passes the first signal section With the second signal section Add to determine the drive signal .

[0043] First signal section Determined by predictive control block 13, which includes a predictive control model of gyroscope structure 2, the predictive control model having a disturbed output state received as input. In one embodiment, the predictive control model has a cost function, for example, that employed by a conventional MPC controller. In this case, predictive control block 13 uses a control law whose solution can be based on the solution of the Euler-Lagrange equations, minimizing the error and the time series of energy required to achieve the desired state.

[0044] Second signal section This is determined by the nominal convergence module 18, which has nominal tube convergence. Specifically, the nominal convergence module 18, together with block 13, allows the output state under disturbance control according to the TRMPC. nominal evolution to output state The convergence of evolution.

[0045] The nominal convergence module 18 includes a gain matrix. The gain matrix By taking into account system uncertainties and considering the parameter deviations of gyroscope structure 2 based on tube convergence. The convergence is defined by predefined values ​​(such as zero). Parameter bias It is the disturbed output state of gyroscope structure 2. Compared to the undisturbed output state It is obtained by the difference between the operation variables.

[0046] Then, the second signal section As the gain matrix and parameter deviation or error It is obtained by multiplying the two.

[0047] Figure 4 schematically illustrates the situation based on two state variables. and The tube convergence of TRMPC control performed on the linear system. In the variable... and In the plane, solid lines mark the desired trajectory within time interval T, which includes the nominal state. - Evolution length converging to zero The N subsequent steps. The dashed lines represent the N actual states of the system under disturbance. - The actual trajectory generated by the evolution. Using triangles, the cross-section of the tube is actually highlighted, representing the appropriate bounded space. Used to identify each nominal state - The corresponding acceptable actual disturbance state - Considering from the initial moment Until the final moment (Right now( The time evolution of N subsequent steps in the nominal state. Naturally, it can be assumed that for each nominal state... - In this case, it is a bounded space of triangles. Including all possible disturbances And all errors caused by uncertainties related to model parameters.

[0048] Bounded space The perturbation-affected output state is defined. Relative to nominal condition - Maximum distance, nominal state - Positioned in bounded space The center. Of course, bounded space. It can have a shape different from the periphery of the triangle and dimensions that depend on the strength required for tube convergence.

[0049] Figure 4 shows two state variables. and The system has a two-dimensional representation in space. Clearly, in the case of gyroscope structure 2, in the first embodiment, there are four state variables as indicated in vector matrix 1; therefore, the bounded space... It is four-dimensional. Considering the dynamic characteristics of the system being analyzed, such as wave motion with respect to wave cycles, the time interval T is predefined during the design phase. Of course, the characteristics of the instruments and / or hardware used are also relevant.

[0050] Nominal condition - It is the ideal, undisturbed state, and is determined based on the undisturbed model of gyroscope structure 2.

[0051] According to one embodiment, the gain matrix The process is determined by using the theory of linear matrix inequalities or LMI, as described in the article entitled “Tube-Based Robust Model Predictive Control for Spacecraft Proximity Operations in the Presence of Persistent Disturbance” by M. Mammarella, Capello, Park, Guglieri, Romano et al., published on January 6, 2018, in Aerospace Science and Technology, Vol. 77, pp. 585-594.

[0052] Therefore, the definition of the region (“pipe”) and its width are obtained as the uncertainty of both the model used in the control and any unmodeled external disturbances (e.g., mooring effects).

[0053] According to the embodiment shown in Figure 3, prediction block 13 is implemented by nominal unit 14 and prediction unit 15 arranged in a cascaded manner. Nominal unit 14 includes a nominally undisturbed model of gyroscope structure 2. Nominal unit 14 receives the disturbed output state. As input, it also receives a first signal portion generated by the prediction unit 15. As feedback, the undisturbed nominal output state of gyroscope structure 2 is obtained. .

[0054] Prediction unit 15 includes a cost function The predictive dynamic control model, the cost function This includes quadratic terms related to the state and driving action of the gyroscope structure 2, as well as non-quadratic terms related to the instantaneous power absorbed by the gyroscope structure 2.

[0055] Therefore, the prediction unit 15 receives the undisturbed nominal state. As input, and through calculations using a predictive control model and optimization problem, the cost function to be minimized is determined. drive signal sequence Therefore, the first signal part Drive signal sequence defined as follows The first element Confirmed. Alternative location: First signal section Corresponding to the drive signal At least one element in the identified sequence ,in .

[0056] Furthermore, according to this embodiment, the nominal convergence module 18 receives the undisturbed nominal state generated by the nominal unit 14. As input, to calculate parameter deviation Second signal section By parameter deviation With the gain matrix Obtained by multiplication.

[0057] Second signal section Enables modification of drive signals At the same time, maintain the actual state of gyroscope structure 2. - The actual trajectory of evolution is more accurately preserved in bounded space. Or within the optimal state. This allows the gyroscope system 2 to function even in the presence of external random disturbances caused by wave motion. Even under certain circumstances, it can be optimally controlled.

[0058] The second embodiment is shown in Figure 5, and the controller 10 will drive the signal. The signal was identified as an augmented drive signal. The differences from the previous solution will be described in detail below.

[0059] Augmented drive signal Including the first part of the augmented signal and the second part of the signal .

[0060] Prediction unit 15 receives the augmented nominal state as input. This includes the undisturbed nominal state. and parameter deviation This parameter deviation The perturbed output state of gyroscope structure 2 is calculated. The operational variables and the undisturbed nominal state The difference between the variables being operated on.

[0061] Augmented state It is a vector like the one shown below, which allows Block 15 to model and predict perturbations. This trend also enhances the prediction of gyroscope unit states.

[0062] In this way, the prediction unit 15 minimizes the cost function. The generated drive signal sequence It will be more accurate, thus allowing for the determination of the augmented drive signal. This makes the control of the electrical converter 9 more efficient.

[0063] The controller obtained in this way is quite robust to the evolution of the disturbed state. This control is determined by the disturbed output state. The dual feedback is obtained through a dynamic tube convergence model.

[0064] Furthermore, the applicant has been able to verify that the controller 10 designed in this way allows the gyroscope system 2 to maintain or return to the desired state, even when there are uncertainties in the predictive control model parameters of the predictive unit 15 and the nominal unit 14.

[0065] Figure 6 schematically shows a simulation of the trend of extracted power over time for an inertial WEC conversion system with controller 10 according to an embodiment of the present invention. This power, as extracted power, is negative. Figures 7 through 9 show some detailed elements related to the graphs in Figure 6. A maximum error of approximately 2% has been found based on the following parameter variations:

[0066] - Gyroscope mass variation (VMG curve): ;

[0067] - Hull mass change (VMS curve): ;

[0068] - Gyroscope inertia variation (VIG curve): ;

[0069] - Flywheel inertia change (VIV curve): .

[0070] Taking into account the nominal parameters of the model undergoing forced wave motion, the charts in Figures 7 through 9, identified by Nom, show the trend of extracted power.

[0071] In addition to the nominal curve Nom, Figure 7 also shows two VMG curves and two VMS curves. In addition to the nominal curve Nom, Figure 8 also shows two VIV curves and two VIG curves. In Figure 9, in addition to the nominal curve Nom, the lower limit representing the extracted power is highlighted by jointly varying the mass and inertia of the floating hull 3, the gyroscope, and the flywheel. and upper limit The curve.

[0072] The present invention also mentions a control method for the gyroscope structure 2 of the WEC system 1 as described above, for which details and cooperating components having the same structure and function as previously described will be indicated by the same numbers and reference numerals.

[0073] Specifically, the gyroscope structure 2 is associated with the floating hull 3 and includes an electric converter 9 to convert the rotational energy of the floating hull 3 into electrical energy. In its most general form, the method includes TRMPC (pipeline-based robust model predictive control) of the electric converter 9.

[0074] Controller 10 receives the disturbed output state, including the operating variables of the gyroscope structure 2. .

[0075] According to the present invention, the method is designed to use a driving signal To drive the converter 9, the drive signal By passing the first signal part ( ) and the second signal section ( The result is obtained by adding them together.

[0076] This method is designed to determine the first signal part by using a predictive control model of gyroscope structure 2. The predictive control model is based on the output state under disturbance. Calculated.

[0077] Furthermore, the method is designed to determine the second signal portion by using a nominal convergence module 18 with dynamic tube convergence. Dynamic transistor convergence is based on the output state under disturbance. Parameter deviation of the operated variable These parameters were calculated. Defined as: the disturbed output state of gyroscope structure 2 With the undisturbed output nominal state The difference between the operated variables. Unperturbed output nominal state. It is obtained through the undisturbed nominal model of gyroscope structure 2.

[0078] In a first embodiment, the method is designed to use a predictive control model implemented by a conventional MPC controller to determine the first signal portion. .

[0079] Furthermore, this method is designed to use a gain matrix. It is defined as "dynamic tube convergence", that is, convergence to the parameter deviation of gyroscope structure 2. A predefined value (preferably zero). For a system with two state variables. and The system, which typically deals with the dynamic convergence of linear systems, is shown in Figure 4 in its more general implementation. In this way, the gain matrix in state space... Perturbed output state of gyroscope structure 2 - Towards the desired undisturbed output state - evolution.

[0080] Required undisturbed output state - It is determined based on the nominal model prior of gyroscope structure 2, that is, considering an undisturbed linear system.

[0081] During the design phase, this method addresses each desired undisturbed state. - The definition of the design time interval T, the number of subsequent steps N, and the bounded space are also important considerations. Size and shape.

[0082] In this way, in the state space, for each disturbance, the output state is... Parameter deviation Multiply by the gain matrix The parameters keep the actual trajectory of gyroscope structure 2 within a bounded space defined for each desired state. The desired state of being undisturbed. - The evolution converges to the predefined final desired state. Naturally, we can assume each desired undisturbed state. - The surrounding bounded space Including interference or disturbance All possible reasons.

[0083] According to one embodiment, the tube gain matrix It is determined offline using the theory of linear matrix inequalities. In one embodiment, the matrix used in the classical representation of a linear system in state-space discretization... and With in the cost function The weight matrices Q, R, and P used are considered together, as described in detail in the next chapter.

[0084] In this way, the drive signal was determined. Part Two In order to maintain the bounded space The evolution of the perturbed state inside the defined tube is used to converge toward the undisturbed state.

[0085] According to the embodiment shown in Figure 3, the method is designed to determine the first signal portion by implementing a prediction block 13 of a cascaded arrangement of nominal unit 14 and prediction unit 15. In nominal unit 14, the disturbed output state is used. and the first signal part Based on the feedback, the undisturbed nominal model of gyroscope structure 2 is used to generate the undisturbed nominal state. .

[0086] In prediction unit 15, the predictive dynamic control model of gyroscope structure 2 is derived from the undisturbed nominal state received as input. Driven to generate the first signal portion .

[0087] The predictive dynamic control model designs a cost function. The cost function It includes quadratic terms related to the state and driving action of the gyroscope structure 2, as well as non-quadratic terms related to the instantaneous power absorbed by the gyroscope structure 2.

[0088] This method is designed to use computation based on an optimization problem, where the driving signal sequence is determined. So that the cost function minimize.

[0089] Therefore, this method is designed to use a sequence of driving signals. The first element Determine the first signal part Alternative location, first signal section It is achieved by using the identified drive signal sequence at least one element in To obtain, among which .

[0090] Furthermore, the method is designed to use the undisturbed nominal state generated by the nominal unit 14. To calculate parameter deviation And define the second signal part .

[0091] In the alternative form shown in the more general aspect of Figure 5, the method is designed to augment the state The signal is provided as input to the prediction unit 15 to generate a drive signal. As an augmentation driving signal. Augmentation state. Including the undisturbed nominal state and parameter deviation Parameter deviation The output state of gyroscope structure 2 is calculated. The operands and the undisturbed nominal output state The difference between the variables being operated on.

[0092] Augmented state It is a vector like the one shown below, which allows Block 15 to model and predict perturbations. This trend also enhances the prediction of gyroscope unit states.

[0093] In this way, the drive signal is obtained It is more accurate and provides more efficient control over the electrical converter 9.

[0094] A so-called robust drive signal is generated by allowing internal and external disturbances relative to the gyroscope structure of the WEC system. This design method has already achieved the preset goals and objectives.

[0095] Furthermore, the corrections generated by the controller and obtained by tube convergence according to the present invention allow the gyroscope structure to return to operating conditions close to the desired state, even in the presence of uncertainties in the parameters of the nominal model in the nominal cell and prediction block, or in the presence of external disturbances that were not previously considered and / or modeled.

[0096] Cost function

[0097] Considering the state of the simplified model indicated above, predict the dynamic control model and the cost function associated with prediction unit 15. The following formula provides a detailed explanation:

[0098]

[0099] in:

[0100] Indicates the first The power extracted at the step, which is determined by the precession speed of the gyroscope. (It is included in the actual output state containing the gyroscope structure 2) (in the vector) and control torque or drive signal The product is given;

[0101] Indicates the first The energy of the undisturbed state at step is calculated as the sum of squares of the vector elements of the actual output state z of gyroscope structure 2, which is weighted by coefficients contained in the diagonal matrix Q.

[0102] Indicates the first The energy of the control variables at each step is calculated as the square of its value and determined by the weight matrix ( Weighted. In one embodiment, the weight matrix ( ) is a scalar, such as .

[0103] The energy representing the state at step N (i.e., at the end of the prediction time range consisting of N steps) is represented by a matrix containing... The coefficients within are weighted.

[0104] Cost function The cost function includes quadratic terms related to the state and control or actuation actions of gyroscope structure 2, as well as non-quadratic terms related to instantaneous power absorption. Since the instantaneous power term is by definition a mixed-term term, the cost function... It is not a convex function; its minimization is achieved by determining the driving signal. Come and seek.

[0105] In the The power extracted in the first step, the first step Energy of the step state and the energy of control variables All of them are added together, and they are calculated within a time interval T consisting of N steps.

[0106] Through the state-related matrix and the matrix related to control variables The contribution of each term is adjusted in the calculation of the total cost function. As the weighting coefficients increase, the energy of the relevant terms decreases.

[0107] According to one embodiment, the matrix It is based on the theory of linear matrix inequalities or LMI, along with the gain matrix. Calculated.

Claims

1. A controller (10) for a gyroscope structure (2), the gyroscope structure (2) being associated with a floating hull (3) and equipped with an electric converter (9) for converting the rotational energy of the floating hull (3) into electrical energy, the controller (10) receiving, in its input, a disturbed output state including the operating variables of the gyroscope structure (2). ), characterized in Used to determine the drive signal for driving the electrical converter (9) The driving signal ( ) includes: the first signal section ( ), which is used in the disturbed output state ( The predictive control model of the gyroscope structure (2) calculated based on the second signal part is used to determine the second signal part. ), which uses the parameter deviation of the operating variable of the gyroscope structure (2) The parameter deviation is determined by the convergence of the tube calculated on the ) and the tube convergence. ) is calculated as the disturbed output state ( ) relative to the undisturbed nominal output state ( The difference between the operational variables.

2. The controller according to claim 1, characterized in that, Including those equipped with a gain matrix ( The nominal convergence module (18) is used to take into account the required undisturbed state of the gyroscope structure (2). - The time evolution of the parameter deviation is defined to the time evolution of the parameter deviation. The convergence of the predetermined value of the gain matrix () This takes into account the need for each desired undisturbed state. - Bounded space () It is defined by ).

3. The controller according to claim 2, characterized in that, The system includes a prediction block (13) equipped with a cascaded nominal unit (14) and a prediction unit (15), the nominal unit (14) comprising an undisturbed nominal model of the gyroscope structure (2), and the nominal unit (14) receiving the disturbed output state. ) as input, and receive the first signal portion generated by the prediction unit (15) ) as feedback to produce the undisturbed nominal output state ( The prediction unit (15) includes a predictive dynamic control model, and the prediction unit (15) receives the undisturbed output nominal state ( ) as input to generate the first signal portion ( ), or the prediction unit (15) receives the undisturbed output nominal state ( ) and the parameter deviation ( ) as input to generate the first signal portion ( ）。 4. The controller according to claim 3, characterized in that, The nominal convergence module (18) receives the undisturbed output nominal state generated by the nominal unit (14). () as input.

5. The controller according to claim 3, characterized in that, The predictive dynamic control model of the predictive unit (15) includes a cost function ( The cost function ( The gyroscope structure (2) has a quadratic term related to the state of the gyroscope structure (2) and the drive signal, and also includes a non-quadratic term related to the instantaneous power absorbed by the gyroscope structure (2), and it uses the cost function to minimize the cost. The calculation determines the driving signal sequence () ), the first signal part ( ) by the driving signal sequence ( at least one element in ) )Sure.

6. A method for controlling an electrical converter (9) of a gyroscope structure (2) associated with a floating hull (3), the electrical converter (9) being configured to convert rotational energy of the floating hull (3) into electrical energy, the method providing for receiving a disturbed output state including operating variables of the gyroscope structure (2). ), characterized by: - By means of including the first signal part ( ) and the second signal section ( The driving signal of ) - To drive the electrical converter (9); - Using the output state under disturbance ( The predictive control model of the gyroscope structure (2) calculated based on the first signal part is used to determine the first signal part. ); and - using the output state under the disturbance ( The parameter deviation of the operated variable () The second signal portion is determined by the convergence of the tube calculated on the tube. The parameter deviation ( ) is the undisturbed nominal output state of the gyroscope structure (2). It is calculated based on the operation of ).

7. The method according to claim 6, characterized in that: - Define the desired undisturbed state of the gyroscope structure (2) using the nominal model of the gyroscope structure (2). - The evolution of ) to reach the predefined final desired state ( The convergence of the state, wherein the evolution of the state occurs within a time interval of N subsequent steps ( Determined in ) ; - Define the output state for each disturbance ( - ) and the corresponding required undisturbed state ( - The parameter deviation compared to ) ); - The parameter deviation ( Multiply by the gain matrix ( The parameters of ) are used to determine the output state affected by the disturbance. - The true trajectory defined by the evolution of ) remains in a bounded space ( Within this bounded space, the desired undisturbed state is defined as follows: - The predefined enclosure of ).

8. The method according to claim 6, characterized in that, - Using the undisturbed nominal model of the gyroscope structure (2), from the disturbed output state of the gyroscope structure (2) ( ) and the first signal portion ( Feedback begins, generating the undisturbed nominal output state ( ), and - in the undisturbed nominal output state received as input ( Based on the predictive dynamic control model of the gyroscope structure (2), the first signal part is generated. ), or in the undisturbed output nominal state ( ) and further received as input the parameter deviation ( Based on the predictive dynamic control model of the gyroscope structure (2), the first signal part is generated. ）。 9. The method according to claim 8, characterized in that, The predictive dynamic control model is designed with a cost function ( The cost function ( ) has the state of the gyroscope structure (2) and the drive signal ( The quadratic terms related to the gyroscope structure (2) and the non-quadratic terms related to the instantaneous power absorbed by the gyroscope structure (2) are used according to the cost function ( The computation is performed to minimize the optimization problem, where the driving signal sequence ( ) is determined, and - based on the driving signal sequence ( at least one element in ) ) to determine the first signal part ( ）。 10. The method according to claim 6, characterized in that: - By receiving the disturbed output state ( ) as input and receive the first signal portion ( ) as feedback to provide for defining the ideal, undisturbed nominal output state ( The nominal unit (14) of the nominal unit (14), and - using the ideal undisturbed output nominal state generated by the nominal unit (14) To calculate the parameter deviation () ）。 11. A WEC system, comprising: - Floating hull (3); - At least one gyroscope structure (2) associated with the floating hull (3) and equipped with an electric converter (9) for converting the rotational energy of the floating hull (3) into electrical energy; - Controller (10) receiving a disturbed output state including the operating variables of the at least one gyroscope structure (2). As input, the controller (10) is characterized as a controller according to any one of claims 1 to 5.

Images