Method for controlling motion stabilized platform for doppler lidar

Abstract

Provided is a method for controlling a motion stabilized platform for Doppler Lidar with which it is possible to reduce the shaking of the Doppler Lidar itself and to achieve high-precision wind condition observations. A control method for reducing the motion of a motion stabilized platform that is installed on an offshore platform and has mounted thereto a Doppler Lidar for observing wind conditions, wherein the motion stabilized platform is installed on an offshore platform so as to be capable of movement, and the motion of the motion stabilized platform is controlled so that control torque is applied to the motion stabilized platform by using a variable damper.

Description

Control method of motion absorption stand for Doppler LIDAR

[0001] The present invention relates to a method for controlling a motion absorbing stand for a Doppler LIDAR for observing offshore wind conditions.

[0002] The installation of a wind power generation facility requires one year of observation data to thoroughly assess wind conditions at the proposed installation site, and much of this measurement is done using wind observation towers. As wind observation projects expand, plans for offshore wind power generation facilities are progressing rapidly along Japan's coast, but as the installation areas expand further in the future, there are concerns that the water depths will increase, making the foundation work for installing wind observation facilities more difficult.

[0003] A method of installing a laser-type wind speed and direction meter (Doppler LIDAR) on a floating platform has been considered as an observation method that is easy to install and has little environmental impact. However, this method requires reducing the motion of the floating platform (sway and tilt) caused by tides, wind, and waves, which affects the measurement accuracy of the Doppler LIDAR.

[0004] However, in the above-mentioned method, the motion (sway and tilt) of the floating platform due to currents, wind, and waves affects the measurement accuracy of the Doppler LIDAR, so it is necessary to reduce these motions.

[0005] The present invention has been made in response to such demands, and its object is to provide a control method for a motion absorption stand for a Doppler LIDAR that can reduce the vibration of the Doppler LIDAR itself and enable highly accurate wind observation.

[0006] The object of the present invention is achieved by a control method for reducing the vibration of a vibration absorption base that is installed on an offshore platform and on which a Doppler LIDAR for observing wind conditions is mounted, wherein the vibration absorption base is installed so as to be able to vibrate relative to the offshore platform, and the method controls the vibration of the vibration absorption base by applying a control torque to the vibration absorption base using a variable damper.

[0007] According to the present invention, it is possible to provide a control method for a motion absorbing stand for a Doppler LIDAR that can reduce the motion of the Doppler LIDAR itself and achieve highly accurate wind observation.

[0008] 1 shows an image of an actual spar buoy with attitude stabilization equipment on which a motion absorber according to the present invention is mounted, along with its schematic diagram. Left: A schematic diagram of the spar buoy and motion absorber, and right: A model of a single pendulum motion absorber (including Doppler LIDAR) simplified to a single pendulum. 2 shows a model related to the skyhook damper theory. 3 shows a simplified model of a single pendulum motion absorber equipped with a skyhook damper (left) and an active damper controlled by the skyhook damper theory (right). 4 shows time series data of the behavior of a single pendulum motion absorber equipped with a linear damper caused by the sinusoidal motion of the spar buoy. 5 shows time series data of the behavior of a single pendulum motion absorber equipped with an active damper controlled by the skyhook damper theory caused by the sinusoidal motion of the spar buoy. 6 shows a graph showing the motion amplitude magnification of a single pendulum motion absorber. 7 shows time series data of the behavior of a single pendulum motion absorber equipped with a linear damper or an active damper controlled by the skyhook damper theory caused by the irregular motion of the spar buoy. A comparison diagram of simplified models of a pendulum-type motion absorber (left) and a counterweight-type motion absorber (right). Time series data of the behavior caused by sinusoidal motion of a spar buoy with a counterweight-type motion absorber equipped with a linear damper or active damper controlled by skyhook damper theory. A graph showing the motion amplitude magnification of the counterweight-type motion absorber. A graph showing the motion amplitude of the counterweight-type motion absorber. Time series data of the behavior caused by irregular motion of a spar buoy with a counterweight-type motion absorber equipped with an active damper controlled by skyhook damper theory. Time series data of the behavior caused by irregular motion of a spar buoy with a non-resetting counterweight-type motion absorber equipped with a linear damper or active damper according to skyhook damper theory and an active spring according to skyhook spring theory. A simplified model of a counterweight-type motion absorber mounted on a discus-type buoy. Time series data of the behavior of a discus-type buoy in irregular waves. This is time series data of the behavior of a discus-type buoy due to irregular motion, using a non-resetting counterweight type motion absorber with a linear damper or active damper based on the skyhook damper theory and an active spring based on the skyhook spring theory.17 shows time series data of the behavior of a non-rebounding counterweight type motion absorber equipped with an active damper and a spring based on the skyhook theory, at a control frequency different from that of FIG. 17, due to the irregular motion of the discus-type buoy shown in FIG. 16. 18 shows time series data of the behavior of a non-rebounding counterweight type motion absorber equipped with an active damper and a spring based on the skyhook theory, at a control frequency of 1 Hz and a control delay of one step, as shown in FIG. 18. 19 shows time series data of the damping torque and restoring torque of a non-rebounding counterweight type motion absorber equipped with an active damper and a spring based on the skyhook theory, for the time series data of FIG. 17. 20 shows time series data of the behavior of a non-rebounding counterweight type motion absorber equipped with an active damper and a spring based on the skyhook theory, in rough weather (significant wave height 9 m, significant wave period 12 sec).

[0009] The control method for the Doppler LIDAR motion absorber mount according to the present invention will be described below. The Doppler LIDAR motion absorber mount according to the present invention can be broadly divided into two types: a single pendulum type motion absorber mount and a counterweight type motion absorber mount, which is an improved version of the single pendulum type.

[0010] First, we will explain the single pendulum motion absorber. Figure 1 shows an image of the actual spar buoy with attitude stabilization equipment and its schematic diagram. A Doppler LIDAR is mounted on the upper structure of the actual spar buoy, and in order to reduce the vibration of the Doppler LIDAR itself, the Doppler LIDAR is placed on a single pendulum motion absorber and then mounted on the upper structure.

[0011] The left side of Figure 2 shows a schematic diagram of the spar buoy and motion absorber platform, and the right side shows a model of the simple pendulum motion absorber platform (including Doppler LIDAR) simplified to a single pendulum. For simplicity, we assume unidirectional long-crested waves, and consider only motion in the vertical plane including the wave direction. In the figure, M is the mass of the motion absorber platform including the Doppler LIDAR, l1 is the distance from the mooring point at the bottom of the spar buoy to the fulcrum of the simple pendulum, l2 is the distance from the fulcrum of the simple pendulum to the center of gravity of the motion absorber platform, and Tc is the control torque. θ1 is the pitching motion of the spar buoy, and θ2 is the sway of the motion absorber platform (angle from above vertically). Note that a state in which the Doppler LIDAR is not swaying means that θ2 = π continues to hold.

[0012] The specifications of the simple pendulum type motion absorber are shown in Table 1. To avoid synchronization with the external force period, the natural period satisfies the shape constraints and is outside the spar buoy's motion period of 3 to 12 seconds (determined based on wave data for the sea area where the spar buoy is installed).

[0013]

[0014] The equation of motion for the simple pendulum motion absorber is explained below. Assuming that the floating platform is sufficiently heavy compared to the motion absorber that includes the Doppler LIDAR, and that the effect of the Doppler LIDAR motion on the floating platform's motion can be ignored, the equation of motion for the simple pendulum is expressed as Equation 1 below, where the magnitude of the gravitational acceleration is g. Here, θ1 and its time derivative (θ1') are input values. By rearranging Equation 1 with respect to θ2" to obtain Equation 2, The equation of motion was solved using the Euler method. The initial values ​​of θ2 and θ2' were set to 0.97π (approximately 5 degrees from the vertical downward direction) and 0.

[0015] The control torque was a linear damper according to Equation 3 and a variable damper according to the skyhook damper theory. Here, c is the linear damping coefficient. Skyhook damper theory involves suspending an object subjected to displacement excitation via a linear spring (as in the model on the left of Figure 3) from an imaginary ceiling with a linear damper (skyhook), and reproducing a similar damping force with the variable damper in the model on the right of Figure 3. Note that ZA and ZB in Figure 3 correspond to θ1 and θ2 in Figure 2, respectively.

[0016] Furthermore, if the mass M, spring constant k, linear damping coefficient csky, variable damping coefficient csemi, vertical displacement ZB of the mass point, and vertical displacement ZA due to displacement excitation in FIG. 3 are used, the equation of motion for the left model in FIG. 3 is Equation 4, and the equation of motion for the right model is Equation 5. Therefore, in order to simulate the damping force of Equation 4 using the variable damping coefficient csemi, csemi should be set to Equation 6. When (Z'B-Z'A) → 0, csemi → ∞, so an upper limit must be set for csemi, which is csky. Also, when csemi < 0, csemi = 0.

[0017] Figure 4 shows the application of skyhook damper theory to a simple pendulum. In this figure, instead of the imaginary ceiling in Figure 3, an imaginary vertical wall containing the axis of rotation of the simple pendulum is assumed. The equation of motion for a simple pendulum is Equation 1, which is obtained by substituting the variable damping coefficient csemi in Equation 7 for the control torque Tc. As can be seen from this equation, measurement of θ2" is required to apply the skyhook damper.

[0018] First, assuming that the pitching motion of the spar buoy is a sinusoidal motion, the motion reduction effect of the motion absorber was confirmed. The pitching amplitude θ1a of the spar buoy was set to 7.5 degrees, and its motion period 2π / ω1 was set to 3 to 12 seconds (determined based on wave data for the sea area where the spar buoy is installed) in 1-second increments. The calculation time step was set to 0.05 seconds, and the calculation time was set to 50 seconds. The skyhook damper was also applied without control delay at each step.

[0019] Figure 5 shows the calculation results when the spar buoy's motion period is 5 seconds and the linear damping coefficient c is set to 0, 10, and 50 Nms. When c = 0, the amplitude of the motion absorber is larger than that of the spar buoy, and the waveform shows components of the spar buoy's motion period and the motion absorber's natural period overlapping. Increasing the linear damping coefficient (c = 50) almost completely eliminates the natural period component of the motion absorber, but the motion period component of the spar buoy remains almost unchanged. Therefore, instead of the linear damping coefficient, the linear damping coefficient csky based on skyhook damper theory is set to 0, 10, and 50 Nms, and the upper limit of each variable damping coefficient csemi is set to 0, 10, and 50 Nms. Figure 6 shows the results. However, the results are almost identical to those of the linear damping coefficient.

[0020] In order to evaluate the validity of the above calculation results, the tilt angle θ2QS of the motion absorber caused by the acceleration acting on the motion absorber at the moment of maximum displacement of the spar buoy under the quasi-static assumption was calculated using Equation 8, and the amplitude magnification was calculated by dividing this by the amplitude θ2a of the motion absorber.

[0021] Figure 7 shows the amplitude magnification, with ω1 / ω2n on the horizontal axis. ω2n is the natural circular frequency of the motion absorber base shown in Table 1, and the amplitude θ2a on the vertical axis is half the maximum and minimum width of the waveforms in Figures 5 and 6. The natural period of the motion absorber base is shorter than the motion period of the spar buoy in question, and since the amplitude magnification is roughly 1, a sufficient damping effect is being achieved, and in order to aim for further motion reduction, it is necessary to lengthen the natural period of the motion absorber base.

[0022] Next, we confirmed the motion-reduction effect of the motion absorber when the spar buoy's pitching motion was irregular. The pitching motion θ1 of the spar buoy was calculated by converting the results of a tank test using a small-scale similarity model to the actual size according to the Froude law. The tank test was conducted in the Osaka Municipal University Ship Test Tank, using a spar buoy equipped with a shallow-water attitude stabilization device and a single-point tension mooring. The irregular wave spectrum was Bretschneider-Mitsuyasu type, and the significant wave height and significant wave period were converted to the actual size of 7.11 m and 10.2 s, respectively, to measure motion in relatively rough sea conditions. Since the tank test results were converted to the actual size in the numerical simulation, the calculation time step was 0.120 s, and the calculation time was 290 s. The skyhook damper was applied without control delay to each calculation step.

[0023] The calculation results using the linear damping coefficient and the variable damping coefficient are shown in Figure 8. Although the vibration of the vibration absorber base was reduced by applying a damping force compared to the result when c = 0, it was still larger than the motion of the spar buoy.

[0024] Based on the results of the above study on the simple pendulum type motion absorber, we considered lengthening the natural period of the motion absorber. In the case of the simple pendulum type, to lengthen the natural period, it is necessary to lengthen l2 according to the formula in Table 1. However, even if l2 is set to 30 m, the same as the total length of the spar buoy, the natural period would be about 11 seconds, making this difficult to achieve. Therefore, we considered the counterweight type motion absorber shown in Figure 9. As with the simple pendulum type, if we assume that the impact of the motion of the motion absorber, including the Doppler LIDAR, on the motion of the floating platform can be ignored, the equation of motion for the counterweight type is expressed as Equation 9, using the letters in Figure 9. where: Therefore, if we consider the counterweight type to be stable when M2 is at the bottom, the natural period is This becomes:

[0025] Table 2 shows the specifications of the counterweight type. First, we assumed that the spar buoy's motion was sinusoidal, and confirmed the motion-reduction effect of the motion absorber. The calculation conditions were the same as those for the sinusoidal pitch motion of the spar section described above. Figure 10 shows the calculation results (time series data) for a linear damping coefficient c of 10 Nms, a linear damping coefficient csky based on skyhook damper theory of 0 and 10 Nms, and an upper limit for the variable damping coefficient csemi of 0 and 10 Nms. Figure 11 shows the amplitude magnification, and Figure 12 shows the motion amplitude. In Figure 10, when csky = 0, the waveform shows the overlap of the spar buoy's motion period and the motion absorber's natural periodic component, but the half amplitude is approximately 3°. When c = 10 and csky = 10, the natural periodic component of the motion absorber disappears. For c = 10, the half amplitude is approximately 1.5°, and for csky = 10, the half amplitude is approximately 0.5°. Furthermore, from Figures 11 and 12, it can be seen that when the motion period is longer at c = 10, the oscillation becomes larger than when c = 0, while at c = 10, the amplitude magnification is 0.1 or less and the motion amplitude is 0.5 deg or less, which is sufficient effect.

[0026] Furthermore, the motion reduction effect of the motion absorber was also confirmed when the pitching motion of the spar buoy was irregular. The calculation conditions were the same as when the pitching motion of the spar section was irregular as described above, and the calculation results are shown in Figure 13. As with the sinusoidal motion of the spar buoy, when csky = 0, the waveform is one in which the motion period of the spar buoy and the natural period component of the motion absorber overlap, but the half amplitude is about 0.6 deg. At csky = 10, the natural period component of the motion absorber disappears and the half amplitude is about 0.4 deg, confirming a sufficient motion reduction effect.

[0027]

[0028] The effectiveness of the counterweight type motion absorber was confirmed in the above study. However, it was also found that it is difficult to completely stop motion. The reason for this is that the counterweight type motion absorber has a restoring force as part of its mechanism, and so the forcing force of the second term in Equation 9 acts on it. Therefore, we considered setting this restoring force to zero, thereby zeroing the forcing force of the motion absorber caused by the movement of the spar buoy. The main details are shown in Table 3.

[0029]

[0030] In this case, if the motion absorber platform is displaced from the vertical downward direction by some kind of disturbance, it will not be able to return to the vertical downward direction due to the lack of restoring force. To solve this problem, we decided to measure θ2 at each instant and actively apply a restoring force. The equation of motion in this case is Equation 12. Here, Tc is the damping torque of the linear damper or skyhook damper, and Tk is the restoring torque of the skyhook spring. The skyhook spring provides a restoring force by replacing the variable damping coefficient of the skyhook damper with a variable restoring coefficient and multiplying it by θ2, which can be expressed as Equation 13. The coefficient of stability was set to the same as above (ksky = 1.31), and the upper limit of ksemi was set to ksky. Since only 290 seconds of irregular motion data for θ1 was available for the calculation, the irregular motion data was repeatedly combined. Other calculation conditions were the same as when the spar buoy's pitching motion was sinusoidal, and the skyhook damper and skyhook spring were applied without control delay at each step.

[0031] The upper diagram in Figure 14 shows the pitching motion of the spar buoy in irregular waves, the middle diagram shows the calculated results of the motion of the motion absorber base when a linear damper is used, and the lower diagram shows the calculated results of the motion of the motion absorber base when a skyhook damper is used. When Tc is a linear damper, the motion of the motion absorber base does not subside even if the coefficient is increased, but by using a skyhook damper and increasing the coefficient, the motion can be reduced to almost zero.

[0032] Next, we will explain the application of the counterweight-type motion absorber with no restoring force that uses the above-mentioned skyhook spring to a discus-type buoy, which is widely used in wave observation buoys, etc. Figure 15 shows a schematic diagram and coordinate system of a discus-type buoy and a counterweight-type buoy, and their main specifications are shown in Table 4. Unlike the above-mentioned spar buoy, a discus-type buoy is loosely moored, and even if the motion of the discus-type buoy is limited to a vertical plane, it still has three degrees of freedom: fore-and-aft sway, heave, and pitch.

[0033]

[0034] As mentioned above, if we assume that the influence of the motion of the motion absorber platform including the Doppler LIDAR on the motion of the floating platform can be ignored, the equation of motion of the counterweight type mounted on the discus-type buoy is expressed as Equation 14 using the characters in Figure 15. If there is no restoring force, the above equation 12 is obtained. Here, Tc is the linear damper or the skyhook damper of equation 7, and Tk is the skyhook spring of equation 13. Note that in order to apply the skyhook damper or skyhook spring, it is necessary to measure θ2 and θ2', respectively.

[0035] Figure 16 shows the results of a tank test of a discus-type buoy in irregular waves, converted to full-scale size according to the Froude law. The test was conducted in the Osaka Municipal University Ship Test Tank using a catenary-moored discus-type buoy. The irregular wave spectrum was Bretschneider-Mitsuyasu type, with a significant wave height of 4.5 m and a significant wave period of 12 seconds. Because the motion data was short, the data was repeatedly stitched together, and other calculation conditions were the same (see below for the influence of control frequency, control delay, and wave height on motion reduction, as well as the possibility of control torque generation). Figure 17 (top) shows the calculated motion of the motion absorber with a linear damper, and (bottom) with a skyhook damper. When Tc is a linear damper, increasing the coefficient does not reduce the motion of the motion absorber. However, when using a skyhook damper, adjusting the coefficient can reduce the motion of the motion absorber to almost zero.

[0036] In this paper, in order to improve the continuous measurement capability and measurement accuracy of the offshore wind observation system, we considered placing the Doppler LIDAR on a motion-absorbing platform to reduce the vibration of the sensor itself, and developed a model for this, confirming its validity through numerical simulations.

[0037] First, we considered a single-pendulum type motion absorber and evaluated its motion reduction performance when it was installed on a shallow-water spar buoy equipped with an attitude stabilizer. Linear and skyhook dampers were used to reduce motion, but the motion of the motion absorber in regular and irregular waves was not smaller than that of the spar buoy. However, an evaluation based on amplitude magnification revealed that the skyhook damper was effective in reducing motion. The reason for this result is thought to be that the natural period of the motion absorber cannot be made sufficiently long compared to the period of the external force due to the single-pendulum type mechanism.

[0038] Next, a counterweight-type vibration absorber was considered to lengthen the natural period of the vibration absorber. This counterweight-type vibration absorber was more effective at reducing vibration than a simple pendulum-type without applying damping force, but the motion amplitude increased in proportion to the vibration frequency. Therefore, when a linear damper was applied, vibration was significantly reduced in the high vibration frequency range, but the results showed that the vibration was inversely proportional to the vibration frequency. Furthermore, by applying a skyhook damper, the motion amplitude was reduced regardless of the vibration frequency, and a sufficient vibration reduction effect was confirmed even during irregular motion, but it was difficult to completely stop the vibration. Finally, by using a skyhook damper with zero restoring force and applying a skyhook spring to maintain a vertically downward position, it was confirmed that vibration could be reduced to almost zero.

[0039] By applying a skyhook spring and skyhook damper to a counterweight-type motion absorption base that has no restoring force, it was confirmed that a sufficient motion reduction effect could be achieved even for a discus-type buoy.

[0040] Here, the control frequency of the control torque applied to the motion absorber is 0.071 seconds (approximately 14 Hz) converted to an actual system, which may be too high for actual implementation. Therefore, we thinned out the data from the motion measurement results of the discus-type buoy in Figure 16 and converted the motion sampling frequency to 4 Hz and 1 Hz, and confirmed the motion reduction effect. Note that other calculation conditions were the same as when a counterweight-type motion absorber was applied to a discus-type buoy (ksky = 1.31, csky = 2). The results shown in Figure 18 confirm that there is no difference in the motion reduction effect within the range of motion sampling frequencies under consideration.

[0041] Furthermore, the control delay of the skyhook damper and skyhook spring is not taken into account. However, in reality, calculations are performed after motion sampling, and then the actuator is activated, resulting in a control delay. Therefore, data was thinned from the motion measurement results of the discus buoy in Figure 16, and the motion sampling data was converted to a motion sampling frequency equivalent to 1 Hz in the actual vessel. Using this data, calculations were performed by applying a one-step delay to the control torque of the skyhook damper and skyhook spring. Figure 19 confirms that, although control delays affect the motion reduction effect, sufficient motion reduction can be achieved by increasing the damping coefficient of the skyhook damper from csky = 2 to 3.

[0042] Furthermore, in order to confirm the feasibility of the motion absorption stand, time series data of the control torque (damping torque and righting torque) for a discus-type buoy without the above-mentioned righting force is shown in Figure 20. From this figure, it can be confirmed that the maximum control torque is less than 0.1 Nm even when both torques are added together, and that this can be achieved with a commercially available DC motor.

[0043] Furthermore, in the measurement of the discus buoy's motion in irregular waves, the significant wave height of the incident waves was set to 4.5 m (significant wave period: 12 seconds, Bretschneider-Mitsuyasu spectrum) due to the wave-making performance of the Osaka Municipal University Ship Test Basin. However, because this wave height is low for rough seas, we made the bold assumption that the motion of the discus buoy in irregular waves is proportional to the significant wave height. Then, we calculated the motion at a significant wave height of 9 m and applied the same control (ksky = 1.31, csky = 2) to confirm the motion reduction effect. The results shown in Figure 21 demonstrate that the motion damping time of the motion absorber is longer than that of Figure 17, but the motion reduction effect can still be confirmed.

Claims

1. A control method for reducing the vibration of a vibration absorption base that is installed on an offshore platform and equipped with a Doppler LIDAR for observing wind conditions, the vibration absorption base being installed so as to be able to swing relative to the offshore platform, and the method controlling the vibration of the vibration absorption base by applying a control torque to the vibration absorption base using a variable damper.

Images