JP7240716B2 - Wave power generator real-time control method and wave power generation system - Google Patents

Description

TECHNICAL FIELD The present invention relates to a real-time control method for a wave power generator that generates power using waves and vertical movements of a water surface, and a wave power generation system.

A wave power generator consists of a movable part that moves vertically, a jig mechanism that transmits the movement of the movable part to the generator, and a power generator that generates electricity. The moving part generates a relative displacement with respect to the part, and electric power is generated using the displacement difference.
A wave power generator needs to maximize the amount of power generation while satisfying constraint conditions (for example, displacement of movable parts, limits on control force). This problem can be viewed as a constrained optimization problem, and there are various solutions. However, even if a controller based on the conventional solution method is implemented in a model of a wave power generator or in an actual machine, the optimization calculation cannot be completed within the sampling period, and implementation becomes impossible in many cases.

Here, Patent Document 1 describes a power generation device having an electric machine that operates as a linear motor or generator mounted inside a closed casing, and a control system that includes an optimum control algorithm and communicates with the linear motor and the power storage device. and a wave power generator with a sensor for measuring the magnitude of the magnetic field source of the linear motor/generator with respect to the closed casing, and a restoring force device, such as an air spring, in which the position difference is synchronized with the wave. is disclosed to sense and generate a restoring force.
Further, Patent Document 2 discloses a device for converting wave energy into electric power, which includes a closed floating element, a mass forming element disposed inside the element, and locking means for locking the movement of the mass forming element. It is disclosed that the locking means is controlled by a control means to selectively lock or unlock the element to generate a resonance phenomenon with water surface displacement.
In addition, Patent Document 3 discloses a model adaptation network including a model format change block that changes the form of a process model, a process model block that has a multivariable process model, and an MPC that generates manipulated variable signals used to control the process. An adaptive multivariable process controller comprising an MPC control system including a controller is disclosed using multi-step control for adaptation of single-input, single-output multi-process models.
Further, Patent Document 4 discloses that, in a positioning system including a feedforward control device and a feedback control device, high-speed positioning is performed by combining feedback and feedforward control for the purpose of track access of a magnetic disk. It is

Japanese translation of PCT publication No. 2015-524034 Japanese Patent Publication No. 2008-517197 JP 2005-202934 A JP-A-54-12082

In previous research on wave power generators, there are many examples of applying the quadratic programming method to the above-mentioned optimization problem with constraints. However, in those examples, the problem is simplified by limiting the degree of freedom of the wave power generator to one degree of freedom, reducing the number of constraints, and the like. Therefore, the prior art cannot be applied to the floating wave power generator as it is. In addition, it is difficult to perform real-time control (real-time control) even if the problem is simplified as in previous research, etc., and performance and safety verification at the time of implementation are insufficient, which hinders practical application. there is

In addition, Patent Document 1 focuses on generating a resonance phenomenon with water surface displacement as a control parameter, and does not discuss the degree of freedom of the movable part and the power generation main part and the method of deriving the optimum parameter under that environment. Not disclosed.
Moreover, Patent Literature 2 does not disclose a multistage algorithm for maximizing power generation output.
Further, Patent Documents 3 and 4 do not disclose that the allowable time for optimal control is obtained in a previous step and the configuration of control algorithms with different roles is changed.

Therefore, the present invention is a wave power generator that can solve in real time an optimization problem with constraints applicable to a wave power generator, and that contributes to the practical use of a wave power generator that ensures high efficiency and safety. An object of the present invention is to provide a time control method and a wave power generation system.

A real-time control method for a wave power generator corresponding to claim 1 is a real-time control method for a wave power generator having multiple degrees of freedom having a movable part and a power generation main part for waves, A time series of the first control parameter and the first state variable with multiple degrees of freedom in the first model is obtained by the first model predictive control calculation using the first model modeled on a wave power generator having a degree of freedom, and the wave Apply the first control parameter and the first state variable limited to the degree of freedom that has a large influence on the power generation amount of the power generator to the second model of the wave power generator, and apply the degree of freedom more than the first model predictive control calculation Obtaining the time series of the second control parameter and the second state variable by the reduced second model predictive control calculation, deriving the final control parameter, and controlling the wave power generator based on the final control parameter Characterized by
According to the present invention of claim 1, the first model predictive control calculation (first stage) corresponding to a multi-degree-of-freedom system and the second degree-of-freedom system corresponding to only those that have a large influence on the power generation amount By combining the model predictive control calculation (second stage), it is possible to solve the constrained optimization problem in real time in a short time and obtain the final control parameters. Real-time control of a wave power generator with multiple degrees of freedom can be performed with high accuracy, and power generation efficiency can be improved. In addition, since it is possible to sufficiently perform performance and safety verification at the time of mounting, safety can be ensured.

According to the second aspect of the present invention, wave prediction is performed based on the displacement of the water surface due to the waves around the wave power generator, and the result of the wave prediction is reflected in the first model predictive control calculation.
According to the second aspect of the present invention, it is possible to improve the accuracy of the first model predictive control calculation.

The present invention according to claim 3 is characterized in that wave prediction is performed using a time-series prediction method based on time-series data of water surface displacement.
According to the third aspect of the present invention, it is possible to improve the accuracy of wave prediction and further improve the accuracy of the first model predictive control calculation.

According to the fourth aspect of the present invention, the first control parameter and the first control parameter that take into account the velocity effect and/or the relative displacement effect on the change in the relative displacement between the movable portion and the power generation main portion are determined by the first model predictive control calculation. It is characterized by obtaining state variables.
According to the fourth aspect of the present invention, it is possible to improve the accuracy of the first model predictive control calculation.

The present invention according to claim 5 is characterized by repeating the first model predictive control calculation using the obtained first control parameter to obtain an approximate solution of the optimum values of the first control parameter and the first state variable. and
According to the fifth aspect of the present invention, the obtained approximate solution of the optimum values of the first control parameter and the first state variable can be used as the initial value of the second model predictive control calculation. The initial values for the two-model predictive control calculations can be better.

The present invention according to claim 6 is characterized in that, in the first model predictive control calculation, mechanical constraints on the power generation main part of the movable part and/or the internal energy consumption of the wave power generator are taken into consideration.
According to the sixth aspect of the present invention, the real-time control of the wave power generator can take mechanical constraints and internal energy consumption into consideration.

According to the seventh aspect of the present invention, in the second model predictive control calculation, the calculation and/or the search range of the solution to be obtained is limited to the degree of freedom that greatly affects the power generation amount of the wave power generator between the movable part and the power generation main part. It is characterized by performing a calculation to limit .
According to the seventh aspect of the present invention, the second model predictive control calculation can be performed accurately and speedily.

The present invention according to claim 8 is characterized by repeating the second model predictive control calculation using the obtained second control parameter to obtain the optimum values of the second control parameter and the second state variable.
According to the eighth aspect of the present invention, the wave power generator can be controlled based on the final control parameter derived from the obtained second control parameter and the optimum value of the second state variable. , the real-time control of the wave power generator can be made more accurate.

A ninth aspect of the present invention is characterized in that, in the second model predictive control calculation, mechanical constraints on the power generating main part of the movable part and/or internal energy consumption of the wave power generator are taken into consideration.
According to the ninth aspect of the present invention, the real-time control of the wave power generator can take mechanical constraints and internal energy consumption into consideration.

The present invention according to claim 10 is characterized in that the first model predictive control calculation and the second model predictive control calculation are completed in less than one cycle of the wave.
According to the tenth aspect of the present invention, the continuity of real-time control of the wave power generator can be made more robust.

According to the eleventh aspect of the present invention, the control of the wave power generator controls the relative motion between the movable part and the main power generation part based on the final control parameters, and the control for maximizing the power generation amount of the wave power generator. It is characterized by
According to the eleventh aspect of the present invention, it is possible to realize real-time control of a wave power generator that maximizes the amount of power generation.

In a wave power generation system corresponding to claim 12, a wave power generator having multiple degrees of freedom having a movable part for waves and a power generation main part is controlled using the real-time control method for a wave power generator of the present invention. It is characterized by controlling by
According to the present invention of claim 12, the first model predictive control calculation (first stage) corresponding to a multi-degree-of-freedom system and the second degree-of-freedom system corresponding to only those having a large influence on the power generation amount By combining the model predictive control calculation (second stage), it is possible to solve the optimization problem with constraints in real time and obtain the final control parameters, so it has multiple degrees of freedom based on the obtained final control parameters. Real-time control of the wave power generator can be accurately performed to improve power generation efficiency. In addition, since it is possible to sufficiently perform performance and safety verification at the time of mounting, safety can be ensured.

A thirteenth aspect of the present invention is characterized by comprising wave height measuring means for measuring the height of waves in the vicinity of the movable portion.
According to the thirteenth aspect of the present invention, the wave height measured by the wave height measuring means can be used for wave prediction or the like.

The present invention according to claim 14 is characterized in that it comprises a force-measuring means for measuring the influence of wave force caused by waves.
According to the fourteenth aspect of the present invention, the influence of the wave force measured by the power calibration means can be used for wave prediction, model predictive control calculation, and the like.

A fifteenth aspect of the present invention is characterized by comprising relative displacement measuring means for measuring relative displacement between the movable portion and the power generation main portion.
According to the fifteenth aspect of the present invention, the relative displacement between the movable portion and the main power generation portion measured by the relative displacement measuring means can be used for wave prediction, model predictive control calculation, and the like.

The present invention according to claim 16 is characterized by comprising attitude measuring means for measuring the attitude of the wave power generator.
According to the sixteenth aspect of the present invention, the attitude of the wave power generator measured by the attitude measuring means can be used for wave prediction, model predictive control calculation, and the like.

According to the seventeenth aspect of the present invention, the wave power generator is a linear PTO (power take-off) having a shaft on which armatures or permanent magnets are laminated as a movable part and capable of regenerative operation and power running operation. It is characterized by
According to the seventeenth aspect of the present invention, real-time control of a wave power generator, which is a linear PTO, can be accurately performed.

The present invention according to claim 18 is characterized in that the wave power generator is a hydraulic linear PTO (power take-off) that converts the movement of the movable part into hydraulic pressure and utilizes it.
According to the eighteenth aspect of the present invention, real-time control of a wave power generator, which is a hydraulic linear PTO, can be accurately performed.

The present invention according to claim 19 is characterized in that the wave power generator is a PTO (power take off) having a floating body capable of moving around an axis as a movable part.
According to the nineteenth aspect of the present invention, real-time control of a wave power generator, which is a PTO having a floating body capable of moving about an axis as a movable part, can be performed accurately.

According to a twentieth aspect of the present invention, an edge safety means is provided for limiting displacement of the edge of the movable portion in the relative motion between the movable portion and the power generating main portion with respect to waves.
According to the twentieth aspect of the present invention, even if the displacement of the movable portion exceeds expectations, it is possible to prevent damage due to contact between the end portion of the movable portion and the power generation main portion or the like.

According to the real-time control method of the wave power generator of the present invention, the first model predictive control calculation (first stage) corresponding to the multi-degree-of-freedom system and the degree-of-freedom system limited to those that greatly affect the power generation amount By combining the corresponding second model predictive control calculation (second stage), it is possible to solve the constraint optimization problem in real time in a short time and obtain the final control parameters, so the obtained final control Based on the parameters, the real-time control of the wave power generator with multiple degrees of freedom can be accurately performed to improve the power generation efficiency. In addition, since it is possible to sufficiently perform performance and safety verification at the time of mounting, safety can be ensured.

In addition, when wave prediction is performed based on the displacement of the water surface due to the waves around the wave power generator and the result of wave prediction is reflected in the first model predictive control calculation, the accuracy of the first model predictive control calculation is improved. can be made

In addition, when wave prediction is performed using a time-series prediction method based on time-series data of water surface displacement, the accuracy of wave prediction is increased to further improve the accuracy of the first model predictive control calculation. be able to.

Further, when obtaining the first control parameter and the first state variable considering the velocity effect and/or the relative displacement effect on the change in the relative displacement between the movable part and the power generation main part by the first model predictive control calculation, , the accuracy of the first model predictive control calculation can be improved.

Further, when the first model predictive control calculation is repeated using the obtained first control parameter to obtain an approximate solution of the optimum values of the first control parameter and the first state variable, the obtained first Since the approximate solution of the optimum values of the control parameter and the first state variable can be used as the initial value of the second model predictive control calculation, the initial value of the second model predictive control calculation can be improved. .

In addition, in the first model predictive control calculation, when considering the mechanical constraints on the power generation main part of the moving part and / or the internal energy consumption of the wave power generator, the real-time control of the wave power generator It is possible to take into account physical constraints and internal energy consumption.

In addition, in the second model predictive control calculation, when performing calculation limited to degrees of freedom that greatly affect the power generation amount of the wave power generator between the moving part and the power generation main part and/or calculation limiting the search range of the solution to be obtained , the second model predictive control calculation can be performed accurately and speedily.

Further, when the second model predictive control calculation is repeated using the obtained second control parameter to obtain the optimum values of the second control parameter and the second state variable, the obtained second control parameter and Since the wave power generator can be controlled based on the final control parameter derived from the optimum value of the second state variable, real-time control of the wave power generator can be made more accurate.

In addition, in the second model predictive control calculation, when considering the mechanical constraints on the power generation main part of the moving part and / or the internal energy consumption of the wave power generator, the real-time control of the wave power generator It is possible to take into account physical constraints and internal energy consumption.

In addition, when the time limit is set so that the first model predictive control calculation and the second model predictive control calculation are completed in less than one cycle of the wave, the continuity of the real-time control of the wave power generator can be further strengthened. can be

In addition, the control of the wave power generator controls the relative motion of the movable part and the power generation main part based on the final control parameter, and if the control is to maximize the power generation amount of the wave power generator, the power generation Real-time control of wave power generator with maximizing quantity can be realized.

According to the wave power generation system of the present invention, the first model predictive control calculation (first stage) corresponding to the multi-degree-of-freedom system and the second model corresponding to the degree-of-freedom system limited to those that greatly affect the power generation amount By combining the predictive control calculation (second stage), it is possible to solve the constrained optimization problem in real time and obtain the final control parameters. Real-time control of the power generator can be performed with high accuracy, and power generation efficiency can be improved. In addition, since it is possible to sufficiently perform performance and safety verification at the time of mounting, safety can be ensured.

Moreover, when the wave height measuring means for measuring the wave height of the wave near the movable portion is provided, the wave height measured by the wave height measuring means can be used for wave prediction or the like.

In addition, when a power calibration means for measuring the influence of wave force due to waves is provided, the influence of wave power measured by the power calibration means can be used for wave prediction, model predictive control calculation, and the like.

Further, when a relative displacement measuring means for measuring the relative displacement between the movable part and the power generating main part is provided, the relative displacement between the movable part and the power generating main part measured by the relative displacement measuring means can be used for wave prediction, model predictive control calculation, etc. can be used for

Further, when an attitude measuring means for measuring the attitude of the wave power generator is provided, the attitude of the wave power generator measured by the attitude measuring means can be used for wave prediction, model predictive control calculation, and the like.

In addition, if the wave power generator is a linear PTO (power take-off) that has a shaft on which armatures or permanent magnets are laminated as a moving part and is capable of regenerative operation and power running operation, it is a linear PTO. Real-time control of a certain wave power generator can be performed with high accuracy.

In addition, if the wave power generator is a hydraulic linear PTO (power take-off) that converts the movement of the moving part into hydraulic pressure and utilizes it, the real time of the wave power generator that is the hydraulic linear PTO Control can be performed with high precision.

In addition, when the wave power generator is a PTO (power take-off) having a floating body that can move around an axis as a movable part, it has a floating body that can move around an axis as a moving part. Real-time control of a wave power generator, which is a PTO, can be performed with high accuracy.

In addition, when an end safety means is provided to limit the displacement of the end of the movable part in the relative movement of the movable part and the power generation main part against waves, even if the displacement of the movable part exceeds expectations, It is possible to prevent damage due to contact between the end of the power generation main body and the like.

Schematic diagram of a wave power generation system according to an embodiment of the present invention Control block diagram of the wave power generation system Control flow diagram of the wave power generation system Explanatory diagram for estimation of co-wave forcing Schematic diagram of end safety measures applied to the same wave power generation system Schematic diagram of a wave power generation system according to an embodiment of the present invention. Control flow diagram of the wave power generation system Schematic diagram of a wave power generation system according to another embodiment of the present invention Control flow diagram of the wave power generation system Schematic diagram of a wave power generation system according to yet another embodiment of the present invention

A real-time control method for a wave power generator and a wave power generation system according to embodiments of the present invention will be described below.

FIG. 1 is a schematic diagram of a wave power generation system according to one embodiment of the present invention.
A wave power generation system 1 installed in the ocean or the like includes a floating wave power generator 10 having six degrees of freedom, a control unit 20 that controls the wave power generator 10 in real time (real-time control), and measures wave height. wave height measurement means 30, force measurement means 40 for measuring the influence of wave force due to waves, relative displacement measurement means 50 for measuring the amount of relative displacement, and attitude measurement means 60 for measuring the attitude of the wave power generator 10. Prepare.
The wave power generation system 1 is moored to the bottom of the water X with a plurality of mooring cables 70 . One end of the mooring cable 70 is connected to the lower part of the wave power generator 10 and the other end is connected to the anchor 80 .
FIG. 1 also shows the motion directions of the wave power generator 10 with six degrees of freedom (longitudinal sway, lateral sway, vertical sway, roll, pitch, and bow sway).

The wave power generator 10 has a movable part 11 for waves and a main body part 12 for generating electricity.
The movable portion 11 has a float 11A located at one end, a rod-shaped member 11B located at the other end, and a frame 11C located between the float 11A and the rod-shaped member 11B.
The power generation body 12 has a PTO (power take-off) portion 12A, a columnar spar 12B with its upper part exposed above the water surface, and a disc-shaped heave plate 12C projecting from the lower part of the spar 12B. . The float 11A actually has an annular shape and is arranged symmetrically on the left and right sides of the spar 12B, although not shown in FIG.
The float 11A floating on the water surface Y is arranged so as to be adjacent to the outer surface of the spar 12B, and reciprocates vertically as the water moves up and down. The rod-shaped member 11B is, for example, a shaft, a piston, or the like, and is arranged inside the spar 12B. The frame 11C is arranged to cover the upper part of the spar 12B, and has a float connecting member 11Ca connected to the float 11A and a rod-shaped member connecting member 11Cb connected to the rod-shaped member 11B. tell to
The PTO portion 12A is arranged inside the spar 12B. A rod member 11B is passed through the interior of the PTO portion 12A. The wave power generator 10 causes the movable portion 11 to generate relative displacement with respect to the power generation main portion 12 due to the vertical movement of the waves or the water surface Y, and utilizes the displacement difference to generate electric power. The power generation main part 12 can generate power according to the movement of the water surface displacement, and at the same time, forcibly vibrate the movable part 11 using the accumulated energy etc., causing a phenomenon of synchronization with the water surface displacement. It is a device that enables absorption of larger water surface displacement energy.

The control section 20 is arranged below the PTO section 12A inside the spar 12B.
The wave height measuring means 30 is arranged above the float 11A and measures the wave height in the vicinity of the float 11A. As the wave height measuring means 30, for example, various wave height meters using ultrasonic waves, acceleration, water pressure, radar, etc. can be used.
The force measuring means 40 is arranged at the connecting portion between the rod-shaped member 11B and the rod-shaped member connecting member 11Cb, and measures the influence of the wave force on the wave power generator 10 . As the force-calibrating means 40, for example, column-type, diaphragm-type, S-shaped, strain-gauge-type, capacitance-type, magnetostriction-type, and other various types of force-checking meters can be used.
The relative displacement measuring means 50 is arranged close to the rod-shaped member 11B and measures the relative displacement between the movable portion 11 and the power generation main portion 12 . A linear encoder, a rotary encoder, or the like, for example, can be used as the relative displacement measuring means 50 .
The attitude measurement means 60 is arranged above the spar 12B and measures the attitude of the wave power generator 10 . As the attitude measurement means 60, for example, a gyro, an inclinometer, or the like can be used.

Next, a real-time control method for the wave power generator 10 by the control unit 20 will be described.
FIG. 2 is a control block diagram of the wave power generation system according to one embodiment of the present invention, and FIG. 3 is a control flow chart of the same wave power generation system.
The control unit 20 includes a water surface displacement predictor 21 , a wave forcing force predictor 22 , a first model predictive controller 23 and a second model predictive controller 24 .

First, the wave power generation system 1 measures water surface displacement such as waves. The water surface displacement of waves or the like is calculated from the relative displacement with respect to the power generation main unit 12 by the wave height measuring means 30 attached to the wave power generator 10, and stored in the data storage unit (not shown) (S1: Step 1 ).

The water surface displacement predictor 21 performs wave prediction for calculating the time series of water surface displacement from the present to the future based on the water surface displacement measurement result η which is the time series data of the water surface displacement stored in the data storage unit. The water surface displacement predictor 21 outputs the water surface displacement prediction result η'(k+i|k) (' is above η), which is the wave prediction result, to the wave forcing force predictor 22 (S2: Step 2 ).
As a method for calculating the time series of water surface displacement from the present to the future, there are parametric models such as AR (autoregressive), ARMA (autoregressive moving average), ARIMA (autoregressive integrated moving average), and Kalman filter. be.

The wave forcing force predictor 22 predicts the wave forcing force acting on the movable part 11 and the power generation main part 12 based on the input water surface displacement prediction result η'(k+i|k) (' is above η). calculate. The wave forcing predictor 22 outputs the calculated wave forcing prediction result F _e '(k+i|k) (' is above F _e ) to the first model predictive controller 23 (S3: step 3). A calculation example of the wave forcing will be described later.

In steps 4 to 6, the first model predictive controller 23 performs a first model predictive control calculation using a first model modeled on the wave power generator 10 having multiple degrees of freedom, and calculates a first control parameter and the time series of the first state variable.
First, the first model predictive controller 23, based on the input wave forcing force prediction result F _e '(k+i|k) (' is above F _e ), 6 degrees of freedom (movable part 11, power generation Behavior calculation of the movable portion 11 and the power generation main portion 12 is performed with the maximum degree of freedom of the main portion 12 (S4: Step 4).
After step 4, the first model predictive controller 23 optimizes the first control parameter based on the behavior calculation result in step 4 (S5: step 5). The first control parameters are parameters of velocity effect and relative displacement effect (positional difference effect) with respect to changes in the mutual positional difference between the movable portion 11 and the power generation main portion 12 . Note that there is also a case of either velocity effect or relative displacement effect. In addition, one option is to configure each term with polynomials up to a higher degree.
After step 5, the first model predictive controller 23 performs behavior calculations of the movable part 11 and the main power generation part 12 associated with the optimum control, and calculates multiple floating body six degrees of freedom system (two or more floating bodies) including multiple and polynomial control parameters. ), the approximate solution (control parameter optimization calculation result u _{sub_opt} , state variable optimization calculation result x _{sub_opt} ) is obtained by the iterative convergence method or the algebraic equation method for the time series of the first control parameter and the first state variable in is obtained (S6: step 6). In this embodiment, the floating body consists of two floating bodies, a movable portion 11 and a power generating body portion 12 . The present invention applies to the case where the movable portion 11 and the power generation main portion 12 are accommodated in one container, and the movable portion 11 is displaced relative to the power generation main portion 12 due to displacement of the water surface caused by waves or the like to generate power. It is applicable even when the number of floating bodies is regarded as one. Further, in step 6, the internal energy consumption of the wave power generator 10 and the moving part 11 In addition to the evaluation parameters, the mechanical constraints on the power generation main body 12 are also added to the calculation so that the energy obtained is maximized. Mechanical constraints include, for example, displacement limits and thrust limits.
For the first model predictive control calculation using the first model in the first model predictive controller 23, a model predictive control method capable of performing real-time control of a mechanical system having multiple degrees of freedom can be applied. can. An approximate solution between prediction horizons (evaluation interval) is obtained by a model control method corresponding to a multi-degree-of-freedom system. The obtained approximate solution is a sequence close to the optimum, although not the optimum.
After step 6, return to step 5, optimize the first control parameter again based on the first control parameter obtained in the previous first model predictive control calculation, and proceed to step 6 to obtain an approximate solution. obtain. Thus, the first model predictive controller 23 repeats the optimization calculation of the first control parameter and the first state variable.
The first model predictive controller 23 outputs the calculated approximate solution (control parameter optimization calculation result u _{sub_opt} , state variable optimization calculation result x _{sub_opt} ) to the second model predictive controller 24 .

In steps 7 to 8, the second model predictive controller 24 sets the first control parameters and the first state variables limited to the degrees of freedom that greatly affect the power generation amount of the wave power generator 10 to the wave power generator 10. is applied to the second model, and the time series of the second control parameter and the second state variable are obtained by the second model predictive control calculation to derive the final control parameter.
First, the second model predictive controller 24 sets the input approximate solution (optimization calculation result u _{sub_opt} of the control parameter, optimization calculation result x _{sub_opt} of the state variable) as an initial value, is reduced to one degree of freedom, and the second control parameter is optimized by limiting the search range of the solution to be obtained, such as by linearizing the polynomial control parameter as necessary (S7: Step 7 ). The degree of freedom is not limited to 1 degree of freedom, and may be, for example, 2 degrees of freedom. If the degree of freedom is less than 6 degrees of freedom, the second model predictive controller 24 obtains a solution in a shorter time. be able to.
After step 7, the second model predictive controller 24 performs behavior calculations of the movable part 11 and the main power generation part 12 accompanying the optimum control, and obtains a necessary solution close to the exact solution (second control parameter and second state variable time series) are obtained to derive the final control parameters (S8: step 8). In step 8, similar to step 6 of the first model predictive controller 23, based on the generator characteristics and constraint conditions input to the second model predictive controller 24 from the external input means, the wave power generator 10 In addition to the evaluation parameters, the internal energy consumption and the mechanical constraints on the power generation main body 12 of the movable part 11 are also added to the calculation so as to maximize the energy obtained.
For the second model predictive control calculation using the second model in the second model predictive controller 24, a model predictive control method for achieving optimal power generation control for a one-degree-of-freedom wave power generator is applied. can be done. Optimization calculation is performed for the motion mode to be controlled by the one-degree-of-freedom system model predictive control method. At this time, the approximate solution obtained by the first model predictive controller 23 (control parameter optimization calculation result u _{sub_opt} , state variable optimization calculation result x _{sub_opt} ) is a good initial value of the optimization calculation and an optimal solution used to narrow down the search range. Also, under the assumption that the time series of the state quantity between prediction horizons is known, and that the state variable optimization calculation result x _{sub_opt} obtained by the first model predictive controller 23 is close to the time series that gives the optimum value, other Among the motion modes, the motion modes that have a large effect on the control mode are also considered for the coupled effect. By using a model predictive control method limited to a one-degree-of-freedom system or the like within a range close to the optimum solution, an algorithm that quickly completes the optimization calculation can be achieved.
After step 8, return to step 7, optimize the second control parameters again based on the second control parameters obtained in the previous second model predictive control calculation, and proceed to step 8 to find the necessary solution. to derive the final control parameters. Thus, the second model predictive controller 24 repeats optimization calculations for the second control parameters and the second state variables.
The second model predictive controller 24 outputs the final calculated control parameter (control parameter optimization calculation result u _opt ). Using this final control parameter as a command value, the controller 20 controls the relative motion between the movable part 11 and the main power generation part 12 of the wave power generator 10 to be controlled, and maximizes the power generation amount of the wave power generator 10. to make it better.
Based on the final control parameter derived from the time series of the second control parameter and the second state variable, which are the necessary solutions, the generator starting part in the power generation main unit 12 is forcibly controlled and excited to synchronize with the water surface displacement. By causing the phenomenon, larger water surface displacement energy can be absorbed.

The number of repetitions of the optimization calculations in the first model predictive controller 23 and the number of repetitions of the optimization calculations in the second model predictive controller 24 are optimized in less than one cycle based on wave prediction (prediction of water surface change). The configuration is made so that the calculation ends, and the configuration is changed according to the water surface displacement time. Control errors due to the length of time required for optimization calculation, delay elements of the mechanism system and control circuit system, hysteresis, etc. are corrected by appropriately correcting the timing, control parameters, etc., and the wave power generator 10 can be operated. Relative motion between the portion 11 and the power generating body portion 12 can be controlled.
Also, the state variables of the wave power generator 10 are measured by the force measurement means 40, the relative displacement measurement means 50, and the attitude measurement means 60, and the state variable measurement result (y=x) of the wave power generator 10 is used to predict the water surface displacement. It is fed back to the device 21 (S10: step 10). This improves the accuracy of the future water surface displacement time series calculated by the water surface displacement predictor 21 .
In addition, as shown in FIG. 2, the state variable measurement result (y=x) of the wave power generator 10 can be fed back to the first model predictive controller 23 . This improves the accuracy of the approximate solution calculated by the first model predictive controller 23 .

Next, an example of wave forcing calculation in the wave forcing predictor 22 will be described with reference to FIG. FIG. 4 is an explanatory diagram relating to estimation of wave forcing.
The amount to be measured in calculating the wave forcing is the time series of the water surface displacement (water surface displacement measurement result η) measured by the wave height measurement means 30 and the wave forcing (F _e ) measured by the force measurement means 40. Series.
The wave forcing force predictor 21, for example, models and predicts the time series of the water surface displacement measurement result η as an AR model. Assuming that the AR model can model the water surface displacement measurement result η, the problem of estimating the parameters of the AR model is solved by the least squares estimation method, the AR model is realized in the state space, and long-term prediction is performed using the Kalman filter.

Next, an example of an algorithm of optimization calculation with constraint conditions in the first model predictive controller and the second model predictive controller will be described.
First, an example of formulation in the multi-degree-of-freedom system of the first model predictive controller 23 will be described.
As shown in FIG. 1, a wave power generation system 1 in which a wave power generator 10 has a movable portion 11 and a power generation main portion 12 and uses a linear PTO as a power generation device is taken as an example. The movable part 11 is capable of motion with six degrees of freedom on the water surface. In the case of this example, power is generated by the movable part 11 swaying up and down in waves.
A state vector representing the motion of the movable portion 11 is represented by the following equation (1). x, y, z, φ, θ, and ψ in the formula (1) respectively represent the sway, sway, sway, pitch, roll, pitch, and yaw of the movable part 11, and superscript dots indicate these. shows the time derivative of Hereafter, they are denoted by symbols x ₁ to x ₁₂ .

The control force (F _g ) consists of two control parameters: C _g (control parameter proportional to the motion speed of the movable part 11: u ₁ ) and K _g (control parameter proportional to the displacement of the movable part 11: u ₂ ). is modeled by the following formula (2). It is assumed that F _g acts only as a vertical force on the movable portion 11 .

In random waves, the memory effect on the radiation hydrodynamic force (F _rij ) is given by equation (3) below. Note that m _ij (∞), K _rij , and x _j (dots above x) in equation (3) are respectively the added mass coefficient at infinite frequency and the memory influence function for the radiation fluid force (motion impulse response function with respect to velocity), and the motion velocity of the movable part 11 in the j mode. Also, in equation (3), it is assumed that the movable part 11 is not moving before t=0. K _rij is obtained by the following formula (4). Note that B _ij in Equation (4) is a wave-making damping force coefficient.

In this example, it is assumed that pitch alone cannot be ignored in coupled motion with up-and-down motion. That is, among K _r3j related to the up-and-down motion of the movable part 11, the formulation is carried out considering two of K _r33 and K _r35 . It depends on the shape of the movable part 11 below the waterline.
The integral term in equation (3) means that the entire motion history from t=0 to the present is affected. Therefore, the integral term in equation (3) is modeled by the linear time-invariant system shown in equation (5) below (see Taghipour, R., Perez, T. and Moan, T.: Hybrid frequency-time domain models for dynamic response analysis of marine structures, Ocean Engineering, Vol. 35, Issue 7, 2008, pp. 685-705.). A _r ^ij , B _r ^ij , and C _r ^ij in equation (5) are the coefficient matrices of the linear time-invariant system, x _r ^ij is the state vector of this linear time-invariant system, and u _r ^j is the input to the system. , x ₉ for K _r33 and x ₁₁ for K _r35 .

There are several techniques for estimating the orders of linear time-invariant systems, A _r ^ij , B _r ^ij , and C _r ^ij (see Taghipour, R., Perez, T. and Moan, T.: Hybrid frequency- time domain models for dynamic response analysis of marine structures, Ocean Engineering, Vol. 35, Issue 7, 2008, pp. 685-705.). Here, the order of the integral term in F _r33 can be estimated as n _r33 , and the coefficient matrix is Ar ₃₃ ^, B _r ³³ _, and C _r ³³ _. could be estimated as A _r ³⁵ , B _r ³⁵ , and C _r ³⁵ .
The state vector (x _r ³³ and x _r ³⁵ ) used when modeling the integral terms of F _r33 and F _r35 with a linear time-invariant model is added to the state vector of formula (1) to expand the following formula (6) When the state vector (X) is defined, the equation of motion of the wave power generation system 1 using the extended state vector is given by the following equation (7). In equation (7), M is an inertia matrix corresponding to X, and F is a vector of external force. First, the inertia matrix is defined by the following equation (8). I in equation (8) is the identity matrix of appropriate dimension and 0 is the zero matrix of appropriate dimension. Also, m _{6 ×6} is given by the following formula (9). m _f , I ₄₄ , I ₅₅ , and I ₆₆ in equation (9) are the mass of the movable part 11, the moment of inertia of rolling, the moment of inertia of pitching, and the moment of inertia of bowing, respectively. Also, a _ij represents the added mass or the added moment of inertia.
An external force vector F corresponding to the expanded state vector X is shown in the following equation (10). F ₁ , F ₂ , F ₃ , M ₄ , M ₅ , and M ₆ in the equation (10) are the longitudinal force, lateral force, vertical force, rolling moment, pitching moment, and yaw motion acting on the movable part 11, respectively. indicate a moment. Also, F _r ³³ and F _r ³⁵ are derived from the linear time-invariant system modeling the integral term in equation (3) and correspond to the right hand side of the first equation of equation (5).
Equation (7) is a dynamical system in the MPC-C/GMRES method (reference: Toshiyuki Ohtsuka: A continuation/GMRES method for fast computation of nonlinear receding horizon control, Automatica, Vol. 40, 2004, pp. 563-574.) is the same form as the formulation of

In addition, mechanical constraints such as the movable range of the wave power generation system 1, constraints related to the control force of the power generator, and the like are set. In this example, the inequality constraint condition of the following equation (11) is set for the vertical displacement. Note that the vertical displacement is represented by x ₃ according to the definition of equation (1).

Convert the inequality constraint to an equality constraint using a dummy input (u ₃ ). The equality constraint condition for the vertical displacement shown in the following formula (12) is that the center is ((x _3min +x _3max )/2,0) and the radius is (x _3min +x _3max )/2 in _the x 3 -u ₃ plane. , and the point (x ₃ , u ₂ ) on the circumference always satisfies equation (11).

Also, the dummy input introduced by the constraint condition is added to the control vector, and the control input vector is expanded to the following formula (13).

The evaluation function (J) is defined by the following equation (14) including the generated power in consideration of the copper loss caused by the linear PTO. Q and R in Equation (14) are diagonal matrices. The diagonal elements of Q and R are q ₁ to q _{12 +nr33+nr35} and r ₁ to r ₃ , all of which are non-negative values. Also, t _f , R _s , and K _t indicate the evaluation section length, the winding resistance of the linear PTO, and the thrust coefficient, respectively. The reason why the first-order term of the dummy input is added to the evaluation function is that the formula (12) includes only the second-order term for the dummy input, and the sign of the dummy input cannot be determined from the stationary condition, and the solution tracking is not possible. This is because it may fail. The diagonal elements of Q, R and the dummy input coefficient r _d1 are the tuning parameters of the model predictive control method (MPC).

As described above, given the formulation of the dynamical system, the constraints, and the evaluation function, the optimal control problem can be solved using methods such as the MPC-C/GMRES method.

Next, an example of formulation in the one-degree-of-freedom system of the second model predictive controller 24 will be described.
It is assumed that the first model predictive control calculation identifies the vertical swing of the movable portion 11 as the motion mode that has the greatest effect on the amount of power generation, and that the vertical swing also has a non-negligible effect on the vertical force.
In the second model predictive control calculation, the state vector representing the motion of the movable part 11 is given by the following equation (15). Note that the symbols follow the symbols used in the description of the first model predictive controller 23 .

The control force is modeled by Equation (2) as in the first model predictive control calculation, and assumed to act only as vertical force on the movable portion 11 .
In random waves, the memory effect on the vertical radiation hydrodynamic force (F _r33 ) is obtained by modeling the integral term of K _r33 with a linear time-invariant system using Eqs. (16).

By the first model predictive control calculation, the time series of the state vector used when modeling the pitch, pitch velocity, pitch acceleration, and K _r35 in the evaluation interval (t _f ) with a linear time-invariant model is These are represented by the following equation (17). Note that k in Equation (17) indicates the current time.

The time series of the fluid force acting in the up-and-down direction due to pitching, which is assumed to have a large effect on up-and-down motion, is represented by the following equation (18) using equation (17). The first and second terms in the equation (18) are the coupling terms of the radiation hydrodynamic force, and the third term is the coupling term of the restoring force. In addition, _C35 in the formula (18) is the combined stability coefficient of up-and-down motion and pitch motion.

The expanded state vector (X') obtained by combining the expression (15) and x _r ³³ is defined as the following expression (19), and using this, the equation of motion of the movable floating body can be expressed as the following expression (20). F _e33 and C ₃₃ in the equation (20) are the vertical wave forcing force acting on the movable portion 11 and the vertical restoring force coefficient, respectively.

Although Equation (20) partially includes nonlinear terms, it can be used as a formulation of a dynamical system in quadratic programming.
Constraints and evaluation functions are defined in the same manner as in the first model predictive control calculation, so their description is omitted.

Next, the edge safety means will be described. FIG. 5 is a schematic diagram of end safety means applied to the wave power generation system according to this embodiment.
The end safety means 90 has a disc spring 90A with an inner diameter guide and a resin shock absorbing material 90B, and is arranged on the other end side of the movable portion 11 . The frame 11C has an upper beam 11Cc arranged above the spar 12B and a lower beam 11Cd arranged inside the spar 12B.
The disc springs 90A are arranged around the rod-shaped member connecting member 11Cb as a pair of upper and lower disc springs so as to sandwich the upper surface of the spar 12B. Shock absorbers 90B are attached to the upper and lower ends of the disc spring 90A, respectively.
By providing the end safety means 90, when the float 11C descends significantly and exceeds the lower limit and enters the descent restricted area, the upper beam 11Cc and the upper shock absorbing material 90B come into contact with each other, preventing the movement of the float 11C. Attenuate. Further, when the float 11C rises significantly and exceeds the upper limit value to enter the limited rise range, the lower beam 11Cd and the lower impact absorbing member 90B come into contact with each other, damping the movement of the float 11C.
Thus, when the displacement of the movable part 11 exceeds the upper or lower limit, the end safety means 90 contact the upper beam 11Cc or the lower beam 11Cd and dissipate the kinetic energy of the movable part 11 . The final control parameter, which is the command value from the controller 20 to the wave power generator 10, is set so that the displacement width of the end portion of the movable portion 11 is within a predetermined range by considering the displacement limit value and the like. However, by providing the end safety means 90, even if the displacement of the movable part 11 exceeds expectations due to the occurrence of unexpected waves, rocking, etc., the ends of the movable part 11 and power generation can be prevented. Damage due to contact with the main body 12 or the like can be prevented.
In addition, as in the present embodiment, the end safety means 90 is designed by combining a disc spring 90A capable of absorbing a large amount of energy with a small displacement and a resin shock absorbing material 90B, so that the movable range of the movable portion 11 is reduced. Sufficient braking force can be given without greatly impairing

As described above, the real-time control method for a wave power generator according to the present embodiment provides a wave power generator 10 having multiple degrees of freedom, which has a movable part 11 for waves and a power generation main part 12, and which has multiple degrees of freedom. A time series of the first control parameter and the first state variable is obtained by the first model predictive control calculation using the first model of the wave power generator 10, and the degree of freedom that greatly affects the power generation amount of the wave power generator 10 The first control parameter and the first state variable limited to are applied to the second model of the wave power generator 10, and the second model predictive control calculation calculates the time series of the second control parameter and the second state variable Then, the final control parameter is derived, and the wave power generator 10 is controlled based on the final control parameter.
Combining the 1st model predictive control calculation (first stage) corresponding to the multi-degree-of-freedom system and the 2nd model predictive control calculation (second stage) corresponding to the degree-of-freedom system that has a large impact on the amount of power generation. Therefore, the final control parameters can be obtained by solving the optimization problem with constraints in real time in a short time. Time control can be performed with high precision, and power generation efficiency can be improved. In addition, since it is possible to sufficiently perform performance and safety verification at the time of mounting, safety can be ensured.

Further, wave prediction is performed based on the displacement of the water surface Y due to the waves around the wave power generator 10, and the result of the wave prediction is reflected in the first model predictive control calculation, thereby increasing the accuracy of the first model predictive control calculation. can be improved.
The wave prediction can be performed based on the actual water surface displacement detected by the wave height measuring means 30, and can also be performed based on the wave forcing detected by the force detection means 40 due to the water surface displacement.

In addition, wave prediction is performed using a time-series prediction method based on the time-series data of the displacement of the water surface Y, thereby increasing the accuracy of wave prediction and further improving the accuracy of the first model predictive control calculation. .

Also, by the first model predictive control calculation, the first control parameter and the first state variable considering the velocity effect and/or the relative displacement effect on the change in the relative displacement between the movable portion 11 and the power generation main portion 12 are obtained. , the accuracy of the first model predictive control calculation can be improved.

Further, by repeating the first model predictive control calculation using the obtained first control parameter and obtaining an approximate solution of the optimum values of the first control parameter and the first state variable, the obtained first control Since the approximate solution of the optimum values of the parameters and the first state variable can be used as the initial values for the second model predictive control calculation, the initial values for the second model predictive control calculation can be made better.

In addition, in the first model predictive control calculation, by considering the mechanical constraints on the power generation main part 12 of the movable part 11 and / or the internal energy consumption of the wave power generator 10, real-time control of the wave power generator 10 can take into account mechanical constraints and internal energy consumption.

In addition, in the second model predictive control calculation, the calculation is limited to the degree of freedom of the movable portion 11 and the power generation main portion 12 that greatly affects the power generation amount of the wave power generator 10 and/or the calculation to limit the search range of the solution to be obtained. , the second model predictive control calculation can be performed accurately and speedily.

Further, by repeating the second model predictive control calculation using the obtained second control parameter and obtaining the optimum values of the second control parameter and the second state variable, the obtained second control parameter and the second Since the wave power generator 10 can be controlled based on the final control parameters derived from the optimum values of the two state variables, the real-time control of the wave power generator 10 can be made more accurate.

In addition, in the second model predictive control calculation, by considering the mechanical constraints on the power generation main unit 12 of the movable part 11 and / or the internal energy consumption of the wave power generator 10, real-time control of the wave power generator 10 can take into account mechanical constraints and internal energy consumption.

Further, by setting a time limit so that the first model predictive control calculation and the second model predictive control calculation are completed in less than one cycle of the wave, the continuity of the real-time control of the wave power generator 10 can be further strengthened. can be

Further, the control of the wave power generator 10 is to control the relative motion between the movable part 11 and the main power generation part 12 based on the final control parameter, and to maximize the power generation amount of the wave power generator 10. , it is possible to realize real-time control of the wave power generator 10 that maximizes the amount of power generation.

In addition, by providing a wave height measuring means 30 for measuring the wave height in the vicinity of the movable portion 11,
The wave height measured by the wave height measuring means 30 can be used for wave prediction or the like.

In addition, by providing the force-measuring means 40 for measuring the influence of the wave force due to waves, the influence of the wave force measured by the force-measuring means 40 can be used for wave prediction, model predictive control calculation, and the like.

Further, by providing the relative displacement measuring means 50 for measuring the relative displacement between the movable portion 11 and the power generating main portion 12, the relative displacement between the movable portion 11 and the power generating main portion 12 measured by the relative displacement measuring means 50 can be used for wave prediction and modeling. It can be used for predictive control calculation and the like.

Further, by providing the attitude measurement means 60 for measuring the attitude of the wave power generator 10, the attitude of the wave power generator 10 measured by the attitude measurement means 60 can be used for wave prediction, model predictive control calculation, and the like. .

In addition, by providing an end safety means 90 for limiting the displacement of the end of the movable part 11 in the relative motion of the movable part 11 and the power generation main part 12 against the wave, even if the displacement of the movable part 11 exceeds expectations, Also, it is possible to prevent damage due to contact between the end portion of the movable portion 11 and the power generation main portion 12 and the like.

FIG. 6 is a schematic diagram of a wave power generation system according to an embodiment of the present invention, FIG. 6(a) is an overall schematic diagram, and FIG. 6(b) is an operation explanatory diagram of a linear PTO. In addition, the same code|symbol is attached|subjected to the same function member as above-described embodiment, and description is abbreviate|omitted.
The wave power generation system 1 of the present embodiment includes a wave power generator 10 having six degrees of freedom, a control unit 20 that controls the wave power generator 10 in real time, a wave height measurement means 30, a power measurement means 40, Equipped with relative displacement measuring means 50 and attitude measuring means 60 , it is moored to the bottom of the water X with a plurality of mooring cables 70 .
FIG. 6 also shows the motion directions of the six degrees of freedom of the wave power generator 10 (longitudinal sway, lateral sway, vertical sway, roll, pitch, and bow).

The wave power generator 10 in this embodiment is a linear PTO (power take-off) having a shaft (bar-shaped member) 11B on which armatures or permanent magnets are laminated as a movable part 11 and capable of regenerative operation and power running operation. and An armature (coil) is incorporated in the PTO portion 12A of the power generation main portion 12 . The relative displacement measuring means 50 measures the amount of displacement of the shaft 11B.
A controller 20 and a battery control system 100 are located inside the spar 12B. Controller 20 controls the current supplied from battery control system 100 to the linear PTO. The battery control system 100 stores electric power generated by the linear PTO, and supplies current to the linear PTO according to commands from the controller 20 .

The upper diagram in FIG. 6B shows the state of regenerative operation (at the time of power generation). A change in the magnetic field caused by the movement of the shaft 11B generates an induced electromotive force, and electric power is regenerated. Also, the power generation load acts as a damping force. The generated regenerated electric power is stored in the power storage device of the battery control system 100 . In regenerative operation, the direction of movement of the shaft 11B and the control force are opposite directions.
The lower diagram in FIG. 6B shows the state of power running (when no power is generated). It is used as a linear motor in power running. A portion of the power generation is lost due to winding resistance. In the power running operation, the direction of movement of the shaft 11B and the control force are the same.

FIG. 7 is a control flow diagram of the wave power generation system 1 in this embodiment.
When the controller 20 starts control, the water surface displacement measured by the wave height measuring means 30, the influence of the wave force measured by the force measuring means 40, the displacement of the shaft 11B measured by the relative displacement measuring means 50, and the attitude The attitude of the wave power generator 10 measured by the measuring means 60 is obtained (S20: step 20).
After step 20, the controller 20 causes the wave forcing force predictor 22 to calculate the wave forcing force based on the water surface displacement prediction result η'(k+i|k) (' is above η) by the water surface displacement predictor 21. Estimate (S21: step 21). A linear Kalman filter, for example, can be used to estimate the wave forcing.
After step 21, the controller 20 performs a constrained optimization operation by the first model predictive controller 23 and the second model predictive controller 24, and obtains the second control parameter and the second state variable which are the necessary solutions. to derive the final control parameter (S22: step 22).
After step 22, the controller 20 issues a current supply command to the battery control system 100 so as to output the necessary control force based on the obtained final control parameters (S23: step 23).
After step 23, the process returns to step 20 to start control over the next prediction horizon.
Thus, according to this embodiment, real-time control of the wave power generator 10, which is a linear PTO, can be performed with high accuracy.

8A and 8B are schematic diagrams of a wave power generation system according to another embodiment of the present invention, FIG. 8A is an overall schematic diagram, and FIG. 8B is an operation explanatory diagram of a hydraulic linear PTO. In addition, the same code|symbol is attached|subjected to the same functional member as the above-mentioned embodiment or Example, and description is abbreviate|omitted.
The wave power generation system 1 of the present embodiment includes a wave power generator 10 having six degrees of freedom, a control unit 20 that controls the wave power generator 10 in real time, a wave height measurement means 30, a power measurement means 40, Equipped with relative displacement measuring means 50 and attitude measuring means 60 , it is moored to the bottom of the water X with a plurality of mooring cables 70 .
FIG. 8 also shows the motion directions of the wave power generator 10 with six degrees of freedom (longitudinal sway, lateral sway, vertical sway, roll, pitch, and bow sway).

The wave power generator 10 in this embodiment is a hydraulic linear PTO (power take-off) that converts the movement of the movable part 11 into hydraulic pressure and utilizes it.
The wave power generator 10 generates a control force by means of a piston (rod-shaped member) 11B and a cylinder arranged in the PTO portion 12A of the power generation body portion 12 . The relative displacement measuring means 50 measures the amount of displacement of the piston 11B.
A controller 20, a hydraulic control system 110 and a hydraulic motor 120 are arranged inside the spar 12B. The controller 20 controls the pressure oil supplied from the hydraulic control system 110 to the hydraulic linear PTO. Hydraulic control system 110 delivers pressurized oil produced by the hydraulic linear PTO to pressure accumulator 130 . In addition, it controls the opening of the variable throttle valve 111 according to the command from the controller 20, and sends pressurized oil to the hydraulic linear PTO. The pressure accumulator 130 consists of a high pressure side pressure accumulator 130A and a low pressure side pressure accumulator 130B. The hydraulic control system 110 is also provided with a plurality of check valves 112 .

FIG. 9 is a control flow diagram of the wave power generation system 1 in this embodiment.
When the controller 20 starts control, the water surface displacement measured by the wave height measuring means 30, the influence of the wave force measured by the force measuring means 40, the displacement of the piston 11B measured by the relative displacement measuring means 50, and the attitude measurement. The attitude of the wave power generator 10 measured by the means 60 and the pressure value of the pressure accumulator 130 are acquired (S30: step 30).
After step 30, the controller 20 causes the wave forcing force predictor 22 to calculate the wave forcing force based on the water surface displacement prediction result η'(k+i|k) (' is above η) by the water surface displacement predictor 21. Estimate (S31: step 31). A linear Kalman filter, for example, can be used to estimate the wave forcing.
After step 31, the controller 20 performs a constrained optimization operation by the first model predictive controller 23 and the second model predictive controller 24, and obtains the second control parameter and the second state variable which are the necessary solutions. to derive the final control parameter (S32: step 32).
After step 32, the controller 20 issues a pressurized oil supply command to the hydraulic control system 110 so as to output the necessary control force based on the obtained final control parameters (S33: step 33).
After step 33, the process returns to step 30 to start control over the next prediction horizon.
As described above, according to this embodiment, real-time control of the wave power generator 10, which is a hydraulic linear PTO, can be accurately performed.

FIG. 10 is a schematic diagram of a wave power generation system according to still another embodiment of the present invention. In addition, the same code|symbol is attached|subjected to the same functional member as the above-mentioned embodiment or Example, and description is abbreviate|omitted.
The wave power generation system 1 of the present embodiment includes a wave power generator 10 having six degrees of freedom, a control unit 20 that controls the wave power generator 10 in real time, a wave height measurement means 30, a power measurement means 40, Equipped with relative displacement measuring means 50 and attitude measuring means 60 , it is moored to the bottom of the water X with a plurality of mooring cables 70 .
FIG. 10 also shows the motion directions of the wave power generator 10 with six degrees of freedom (forward-backward sway, left-right sway, up-down sway, roll, pitch, and bow sway).

This embodiment is a so-called bending-swing type wave power generation system, and the wave power generator 10 is a PTO (power take off) having a floating body capable of moving around an axis 17 as a movable part 11. Yes, and the PTO is a hydraulic linear PTO.
The movable part 11 has a float 11Aa on one side and a floating body 11Ab on the other side. The floating body 11Aa on one side and the floating body 11Ab on the other side are flat plate-like or rod-like. One floating body 11Aa and the other floating body 11Ab are arranged to face each other, and opposite ends are connected via a shaft 17 . Wave height measuring means 30, attitude measuring means 60, hydraulic control system 110, hydraulic motor 120, and pressure accumulator 130 are arranged on the upper surface of one floating body 11Aa. A controller 20 and attitude measuring means 60 are arranged on the upper surface of the other floating body 11Ab. A force measuring means 40 and a relative displacement measuring means 50 are arranged on the shaft 17 . The PTO portion 12A and the piston (rod-shaped member) 11B are arranged sideways and supported by frames 11C erected on the one floating body 11Aa and the other floating body 11Ab. are placed across the The piston 11B reciprocates horizontally.
Note that the control flow of the wave power generation system 1 in this embodiment is the same as that described with reference to FIG. 9 above, so description thereof will be omitted.
In this embodiment, the PTO is a hydraulic linear PTO, but a linear PTO having a shaft (bar-shaped member) 11B on which armatures or permanent magnets are laminated as the movable part 11 and which is capable of regenerative operation and power running is also available. You can also The control flow of the wave power generation system 1 in this case is the same as that described using FIG. 7 above.
Also, the PTO may be of a type other than the linear type, for example, a type that generates power by relative motion of a rotating shaft.
In addition, the wave power generator 10 of this embodiment can be provided with an end safety means 90 for limiting the displacement of the end of the movable portion 11 in the relative motion between the movable portion 11 and the power generation main portion 12 with respect to waves.
Thus, according to the present embodiment, real-time control of the wave power generator 10, which is a PTO having the floating bodies 11Aa and 11Ab that can move around the axis as the movable part 11, can be performed with high accuracy.

INDUSTRIAL APPLICABILITY The present invention can be applied to a floating wave power generator and contribute to the practical use of a wave power generator that ensures high efficiency and safety. In addition to wave power, it is also possible to develop power generators that use the force associated with water surface fluctuations caused by tidal currents and ocean currents.

1 wave power generation system 10 wave power generator 11 movable part 12 power generation main part 30 wave height measurement means 40 power measurement means 50 relative displacement measurement means 60 attitude measurement means 90 end safety means

Claims (20)

A real-time control method for a multi-degree-of-freedom wave power generator having a movable part for waves and a main power-generating part, the first model using a first model modeled on the multi-degree-of-freedom wave power generator. A time series of the first control parameter and the first state variable with multiple degrees of freedom in the first model is obtained by one-model predictive control calculation, and the degree of freedom that has a large influence on the power generation amount of the wave power generator is limited to the The first control parameter and the first state variable are applied to the second model of the wave power generator, and the second model predictive control calculation is performed with a reduced degree of freedom compared to the first model predictive control calculation. A real-time wave power generator characterized by obtaining a time series of a control parameter and a second state variable, deriving a final control parameter, and controlling the wave power generator based on the final control parameter. control method. 2. The wave according to claim 1, wherein wave prediction is performed based on the displacement of the water surface caused by the wave around the wave power generator, and the result of the wave prediction is reflected in the first model predictive control calculation. A real-time control method for a power generator. 3. The real-time control method for a wave power generator according to claim 2, wherein the wave prediction is performed using a time-series prediction method based on the time-series data of displacement of the water surface. By the first model predictive control calculation, the first control parameter and the first state variable considering the velocity effect and/or the relative displacement effect on the change in the relative displacement between the movable part and the power generation main part are obtained. A real-time control method for a wave power generator according to any one of claims 1 to 3, characterized in that: 5. Repeating the first model predictive control calculation using the obtained first control parameter to obtain an approximate solution of optimum values of the first control parameter and the first state variable. The real-time control method for the wave power generator according to 1. 6. The method according to claim 4 or 5, wherein the first model predictive control calculation is performed in consideration of mechanical constraints on the power generating main part of the movable part and/or internal energy consumption of the wave power generator. A real-time control method for the described wave power generator. In the second model predictive control calculation, calculation limited to degrees of freedom between the moving part and the power generation main part that greatly affect the power generation amount of the wave power generator and/or calculation limiting a search range for a solution to be obtained A real-time control method for a wave power generator according to any one of claims 4 to 6, characterized in that: 8. The method according to claim 7, wherein the obtained second control parameter is used to repeat the second model predictive control calculation to obtain the optimum values of the second control parameter and the second state variable. A real-time control method for a wave power generator. 9. The method according to claim 7 or 8, wherein the second model predictive control calculation is performed in consideration of mechanical constraints on the power generation main part of the movable part and/or internal energy consumption of the wave power generator. A real-time control method for the described wave power generator. 10. The method according to any one of claims 1 to 9, wherein the first model predictive control calculation and the second model predictive control calculation are completed in less than one cycle of the wave. A real-time control method for the described wave power generator. The control of the wave power generator is a control for maximizing the power generation amount of the wave power generator by controlling the relative motion between the movable part and the power generation main part based on the final control parameter. A real-time control method for a wave power generator according to any one of claims 1 to 10, characterized by: Controlling a wave power generator having multiple degrees of freedom having a movable part for waves and a power generation main part using the real-time control method for a wave power generator according to any one of claims 1 to 11. A wave power generation system characterized by: 13. The wave power generation system according to claim 12, further comprising wave height measuring means for measuring the height of the wave in the vicinity of the movable portion. 14. A wave power generation system according to claim 12, further comprising calibration means for measuring the influence of wave power by said waves. 15. A wave power generation system according to any one of claims 12 to 14, further comprising relative displacement measuring means for measuring relative displacement between said movable portion and said main power generation portion. 16. A wave power generation system according to any one of claims 12 to 15, further comprising attitude measuring means for measuring the attitude of said wave power generator. 3. The wave power generator is a linear PTO (power take-off) capable of regenerative operation and power running operation having a shaft on which armatures or permanent magnets are laminated as the movable part. A wave power system according to any one of claims 12 to 16. 17. Any one of claims 12 to 16, wherein the wave power generator is a hydraulic linear PTO (power take-off) that converts the movement of the movable part into hydraulic pressure and utilizes it. The wave power generation system described in . 17. The wave power generator is a PTO (power take off) having a floating body capable of moving around an axis as the movable part. A wave power system as described. 20. The apparatus according to any one of claims 12 to 19, further comprising an end safety means for limiting displacement of an end of the movable portion in relative motion between the movable portion and the power generation main body with respect to the wave. The wave power generation system described in .

Images