A control algorithm of an active intelligent gyro stabilizer

Abstract

This invention discloses an active intelligent multi-directional gyro stabilizer and its intelligent control algorithm for improving the flutter stability of long-span bridges. The gyro stabilizer structure includes a motor, an attitude angle sensor, a power supply box, a control box, a gyro stabilizer body, and a chain drive device. The gyro stabilizer obtains the aerodynamic torque at a unit torsional angle of the main beam by identifying the flutter derivative through wind tunnel testing. When the main beam undergoes torsion, the system combines the aerodynamic torque at the corresponding torsional angle of the main beam with the instantaneous optimal mapping relationship between the torsional angle of the main beam and the spin angular velocity of the gyro stabilizer established based on artificial intelligence neural network technology. It then outputs the spin angular velocity of the gyro stabilizer at the corresponding torsional angle of the main beam, as well as the gyro precession angular velocity calculated according to gyro precession theory. The chain drive device then controls the gyro precession to generate gyro torque, intelligently controlling the motion state and attitude behavior of the stabilizer, thereby improving the flutter stability of the bridge cross-section.

Description

Technical Field

[0001] This invention relates to the field of civil engineering bridge engineering technology, and in particular to a control algorithm for an active intelligent multi-directional gyro stabilizer used to improve the flutter stability of long-span bridges. Background Technology

[0002] Flutter is a divergent aerodynamic instability phenomenon, typically manifesting as coupled torsional and bending vibrations, with torsion being the dominant mode. It can potentially lead to bridge collapse. Therefore, flutter stability, as a key indicator of the wind resistance performance of long-span bridges, plays a crucial role in bridge design and construction.

[0003] As bridge spans increase, long-span bridges often exhibit characteristics such as high flexibility and low damping, making them susceptible to dynamic environmental loads like traffic and wind, resulting in relatively limited flutter stability. Currently, two common methods are used to improve the flutter stability of bridge sections: 1) Modifying the aerodynamic shape: By altering the aerodynamic shape of the bridge section, such as adding a central stabilizing plate or horizontal guide vanes, the aerodynamic characteristics of the section can be changed, thereby improving flutter stability within a certain range; 2) Installing damper devices: By driving additional mass blocks to dissipate the energy acquired from the outside by the system (TMD system), the flutter stability can be improved.

[0004] In recent years, with the rapid development of bridge spans, the span of large bridges has exceeded 2000 meters. As the span increases, the width-to-span ratio of large bridges decreases, and the torsional frequency ratio drops rapidly. This reduces the economic viability of further improving bridge flutter stability through aerodynamic measures, increases the engineering workload, and shows a clear marginal effect in improving flutter stability through aerodynamic means, with declining effectiveness and performance benefits. While bridge damper (TMD) systems can further improve bridge flutter stability to some extent compared to the former, their control robustness is poor, and the flutter suppression effect is highly sensitive to the frequency of the TMD. Furthermore, within a small amplitude vibration range, the response speed is limited due to the small stroke of the mass block. For example, patent CN111172860A discloses a bridge flutter suppression device and its usage method, but this method may have limited effectiveness in improving the flutter stability of large-span bridges with complex cross-sectional shapes. Patent CN112031194A discloses a TMD device with an eddy current damper and its usage method. This method has a good effect on vertical vibration control, but its performance improvement may be relatively limited for the flutter stability of long-span bridges, which are dominated by bending and torsional modes.

[0005] Current methods for improving bridge flutter stability commonly employ traditional approaches, typically relying on pneumatic measures and TMD (Transient Damping Device) vibration control systems. However, these methods suffer from drawbacks such as diminishing returns, poor control robustness, and low damping efficiency for long-span bridges. Therefore, a vibration control device with high energy efficiency, fast response, and good economic performance is urgently needed for long-span bridges. Summary of the Invention

[0006] To address the problem that existing bridge flutter suppression methods are not applicable to long-span bridges and have poor flutter suppression effects on long-span bridges, this invention provides an active intelligent multi-directional gyroscope stabilizer for improving the flutter stability of long-span bridges.

[0007] As the span of a bridge increases, its flexibility gradually increases while its damping decreases, and the flutter stability of the structural system also decreases with increasing span. Therefore, based on the bridge flutter mechanism, this invention utilizes the gyroscopic torque generated during the precession of a gyro stabilizer to counteract the aerodynamic torque on the main beam section. Furthermore, by combining the flutter derivative measured in wind tunnel tests and the attitude angle sensors installed on the actual bridge equipment, the invention actively controls the spin angular velocity, precession direction, and precession angular velocity of the gyro stabilizer in real time through an embedded algorithm, thereby improving the bridge's flutter stability.

[0008] The present invention relates to an active intelligent multi-directional gyro stabilizer for improving the flutter stability of long-span bridges. Its structure includes a motor, an attitude angle sensor, a power supply box, a control box, a gyro stabilizing body, a chain drive device, and a slide rail fixing device.

[0009] The gyro stabilizing body includes a spherical frame and a gyro rotor housed within the spherical frame. The gyro rotor's central axis is vertically positioned, with its upper and lower ends movably connected to the top and bottom of the spherical frame, respectively. The top of the axis is connected to a motor, which is located above the spherical frame. Driven by the motor, the gyro rotor rotates within the spherical frame around the axis. An attitude angle sensor is also fixedly mounted on the motor. The attitude angle sensor monitors the lateral deflection angle of the main beam in real time and transmits the deflection angle to the control box via Bluetooth for parameter input. The spherical frame is mounted on a support frame, which includes four columns and a ring fixed to the top of each column. The inner diameter of the ring is larger than the diameter of the spherical frame. The spherical frame is located within the ring and connected to the ring via two symmetrically arranged pins. The spherical frame can rotate up and down within the ring using the two pins as fixed points.

[0010] The preferred structure is that the spherical frame consists of multiple warp plates and an equatorial plate, with the equatorial plate fixedly connected to the ring of the support by a pin.

[0011] The chain drive device includes an embedded motor and a chain. One end of the chain is connected to the power output end of the motor, and the other end of the chain is connected to the spherical frame through a gear. Power is provided by the motor, and the chain drive drives the spherical frame to rotate up and down in the ring with two pins as fixed points, thus realizing the precession of the gyroscope rotor.

[0012] The control box is connected to the attitude angle sensor, motor one and motor two respectively. The control box analyzes the data transmitted by the attitude angle sensor to give the optimal gyroscope precession angular velocity, and drives the gyroscope rotor to precess through the chain drive device to suppress the bending and torsional mode vibration of the bridge.

[0013] The power supply box provides power to the entire device.

[0014] The chute fixing device is a mounting plate made by splicing two rectangular plates. The mounting plate is used to fix the gyroscope stabilizing body to the bridge, that is, the bottom ends of the four columns are fixed to the mounting plate. The mounting plate is equipped with a stretchable telescopic rod, and the telescopic rod is provided with mounting holes for fixed connection with the bridge. The length of the telescopic rod can be adjusted to adapt to different bridge widths.

[0015] The algorithm for intelligent control implemented by the intelligent multi-directional gyroscope stabilizer of this invention comprises the following steps:

[0016] S1. Establish the motion control equations for the coupled motion of the gyro stabilizer and the main beam, as follows:

[0017]

[0018] in, Represents the torsional vibration of the main beam. c and k represent the moment of inertia, torsional damping, and torsional stiffness of the main beam, respectively. U is Incoming flow velocity ，B Characteristic width of the entire bridge ，K To reduce the frequency, ， h is the vertical vibration of the main beam section under the action of the flow field; dimensionless constant. , i =1,2,3,4; air density, It is the torsional vibration velocity. Represents vertical vibration velocity. It is torsional vibration acceleration.

[0019] S2. A simulation calculation model of the main beam and gyro stabilizer is established using simulation technology. This model is used to simulate the prototype bridge. The aerodynamic self-excited force per unit angle of the main beam, calculated using flutter derivatives, is input into the main beam model to simulate the motion state of the main beam. At the same time, the rotational speed of the gyro stabilizer is continuously changed, and the motion state of the main beam is recorded to simulate different working conditions. The relationship between the attitude of the main beam and the rotational angular velocity and precession velocity of the gyro stabilizer under multiple working conditions is established. Then, through BP neural network training, the motion state and attitude behavior of the gyro stabilizer are finally intelligently controlled. The gyro torque generated during the precession of the gyro is used to balance the self-excited aerodynamic torque on the bridge, thereby improving the flutter stability of the bridge over a long span.

[0020] In step S2, the BP neural network training specifically involves: using a BP neural network to construct a prediction model for the main beam torsion angle and the gyro stabilizer's rotation angular velocity. A three-layer neuron system is selected, consisting of an input layer, a hidden layer, and an output layer. From the input layer to the hidden layer, the Sigmoid function is used as the activation function to preprocess the input data, transforming the unbounded input into a predictable form. Finally, from the hidden layer to the output layer, under the approximation of the MSE loss function, the predicted value of the gyro stabilizer's rotation angular velocity increasingly approaches the true value.

[0021] The input data is preprocessed to transform the unbounded input into a predictable form. The method and formula are as follows:

[0022]

[0023] x This represents the variables in the input layer, and the formula converts the data to the range [0,1].

[0024] The MSE loss function is as follows:

[0025]

[0026] in, n Indicates the number of samples. Represents the true value of the variable. This represents the predicted value of the variable, i.e., the output value of the grid.

[0027] The working principle of the gyro stabilizer of this invention is as follows: The attitude angle sensor (WT9011DCL-BT50) transmits the tilt angle of the main beam section back to the control box in real time via Bluetooth. Based on prior wind tunnel tests and according to the flutter derivative principle, the self-excited aerodynamic torque on the main beam per unit angle is effectively given. At this time, the control box converts the calculated precession angular velocity into the rotational speed of the chain drive device. Simultaneously, the gyro rotor rotates from top to bottom at the predicted rotational speed. Finally, through the chain drive device, it drives the gyro stabilizer body to rotate along the precession direction, generating a gyro torque opposite to the precession direction and balancing the external torque (self-excited aerodynamic torque), improving the flutter stability of the main beam section, and thus enhancing the flutter stability of long-span bridges.

[0028] Compared with the prior art, the advantages of the present invention are:

[0029] This invention utilizes the principles of flutter derivatives and employs an artificial intelligence algorithm embedded in the device to establish a mapping relationship between the unit torsional angle of the main beam and the spin angular velocity of the gyro stabilizer, thereby calculating the precession angular velocity and obtaining the required gyro torque. The device of this invention features high efficiency, strong operability, and a clear algorithm structure, making it applicable to various bridge cross-sections. The device also boasts high robustness, good sensor sensitivity, and high energy efficiency.

[0030] This invention differs significantly from existing technologies in its vibration suppression mechanism, offering advantages such as high energy efficiency, fast response speed, and broad applicability compared to existing bridge flutter suppression methods. Furthermore, it is not limited by regional wind conditions and can improve the flutter stability of beams without altering their aerodynamic shape, particularly for aerodynamically sensitive sections. Compared to TMD vibration control equipment, it not only exhibits good vibration suppression in vertical bending modes but also demonstrates good suppression performance in bending-torsional modes. Simultaneously, it avoids problems such as insufficient frequency tuning accuracy and poor control robustness.

[0031] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0032] Figure 1 A three-dimensional view of the active intelligent multi-directional gyroscope stabilizer device for improving the flutter stability of long-span bridges provided by this invention.

[0033] Figure 2 A schematic diagram of the gyroscope's stabilizing body.

[0034] Figure 3 A schematic diagram of the slide rail fixing device.

[0035] Figure 4Vibration diagram of a bridge cross section in a flow field.

[0036] Figure 5 Simulation model of the coupling between the main beam and the gyro stabilizer.

[0037] Figure 6 Neural network training flowchart.

[0038] Figure 7 BP neural network prediction structure diagram.

[0039] Figure 8 The present invention provides a flowchart of the operation of an active intelligent multi-directional gyroscope stabilizer device for improving the flutter stability of long-span bridges.

[0040] Figure 9 The working principle diagram of the active intelligent multi-directional gyroscope stabilizer device for improving the flutter stability of long-span bridges provided by this invention. Detailed Implementation

[0041] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0042] like Figure 1 As shown, the active intelligent multi-directional gyro stabilizer for improving the flutter stability of long-span bridges provided by the present invention includes a motor 1, an attitude angle sensor 2, a power supply box 3, a control box 4, a gyro stabilizing body 5, a chain drive device 6, and a slide rail fixing device 7.

[0043] The gyroscope stabilizing body 5 includes a spherical frame 51 and a gyroscope rotor 52 disposed within the spherical frame. The central shaft 53 of the gyroscope rotor 52 is vertically arranged, and its upper and lower ends are movably connected to the top and bottom ends of the spherical frame 51, respectively. The top end of the shaft 53 is connected to a motor 1, which is located above the spherical frame 51. Driven by the motor 1, the gyroscope rotor 52 rotates within the spherical frame around the shaft 53. An attitude angle sensor 2 is also fixedly mounted on the motor 1. The spherical frame 51 is mounted on a support, which includes four columns 54 and a ring 55 fixed to the top of the columns. The inner diameter of the ring 55 is larger than the outer diameter of the spherical frame 51. The spherical frame 51 is located within the ring 55 and is connected to the ring by two symmetrically arranged pins 56. The spherical frame 51 can rotate up and down within the ring 55 using the two pins 56 as fixed points.

[0044] The spherical frame consists of multiple meridian plates 57 and an equatorial plate 58. The equatorial plate 58 is fixedly connected to the ring 55 of the support by a pin 56, thereby enabling the spherical frame 51 to be installed on the support.

[0045] The motor 1 is used to drive the gyroscope rotor 52 to rotate around the central shaft 53, with a speed adjustment range of 10,000-100,000 RPM. The power supply box 3 is a DC power supply box with a rated power of 70 kWh.

[0046] The control box 4 contains an integrated circuit board with a CPU: R3 4100 4-core 8-thread 3.8GHz. The intelligent control algorithm of this invention is embedded in the control box 4, which intelligently controls the attitude of the gyro stabilizer by changing the rotational speed and precession angular velocity, thereby improving the flutter stability of the main beam.

[0047] The main function of the gyro stabilizing body 5 is as the core equipment of the entire stabilizing device. While rotating, it precesses along the direction of the main beam's torsion, providing the required gyro torque to the target object and resisting external torque.

[0048] The function of the chain drive device 6 is to combine the flutter derivative of the corresponding section obtained from the wind tunnel test with the actual torsional angle of the main beam through the control box to obtain the torsional torque on the main beam, calculate the angular velocity required for the precession of the gyro stabilizer, and apply it through the chain drive device. The chain drive device 6 includes an embedded motor 2 (not shown) and a chain 61. One end of the chain 61 is connected to the power output end of the motor 2 through a gear, and the other end of the chain is connected to the spherical frame 51 through a gear 62. Power is provided by the motor 2, and the chain drive drives the spherical frame to rotate up and down in the ring with two pins as fixed points, thus realizing the precession of the gyro rotor. The control box 4 is connected to the attitude angle sensor 2, motor 1, and motor 2 respectively. The control box 4 analyzes the data transmitted by the attitude angle sensor 2 to give the optimal gyro precession angular velocity, and drives the gyro rotor 52 to precess through the chain drive device 6 to suppress the bending and torsional modal vibration of the bridge.

[0049] The power supply box 3 provides power to the entire device.

[0050] The sliding track fixing device 7 is an installation plate formed by splicing two rectangular plates 71. Each of the two rectangular plates has a narrow extension section 72 at its left and right ends, and each extension section has a mounting hole for installing the bottom end of the column 54. An extendable telescopic rod 73 is installed along the long side of the rectangle inside the splice of the two rectangular plates. The telescopic rod has evenly distributed mounting holes for fixing the installation plate to the bridge. The length of the telescopic rod can be adjusted to accommodate installations on bridges of different widths.

[0051] The flutter control principle of the gyro stabilizer device of the present invention is as follows: based on the flutter derivative of the main beam section measured by wind tunnel test, the aerodynamic external torque of the main beam section under a unit deflection angle is obtained. At the same time, combined with the WT9011DCL-BT50 real-time equipment attitude sensor built into the actual invention device, the mapping relationship between the main beam deflection angle and the gyro stabilizer rotation angular velocity is obtained through Python neural network program training. This overcomes the error caused by the system equipment itself, realizes intelligent control of the overall equipment attitude, and thus improves the overall flutter stability of the bridge.

[0052] The intelligent control algorithm embedded in the control box 4 is as follows:

[0053] (1) Establish the motion control equations of the gyro stabilizer-bridge coupling.

[0054] Flutter behavior can cause structural damage and even wind-induced collapse of long-span bridge systems. Therefore, the flutter performance of long-span bridges is a key indicator in bridge wind-resistant design, and flutter derivative tests are usually conducted on standard cross-sections. In this embodiment, wind tunnel tests are first used to capture the free decay time history of a segmental model under initial excitation. Based on Scanlan flutter theory, such as... Figure 4 As shown, the main beam cross-section experiences vertical vibration h and torsional vibration under the influence of the flow field. .

[0055] like Figure 4 The motion state of the main beam section shown can be obtained through the state vector. Therefore, according to Scanlan's flutter theory, the aerodynamic self-excited force at this point is the incoming flow velocity. U Vibration frequency The function relating the state vector is expressed as shown in equation (1):

[0056] (1)

[0057] In the formula, Represents vertical vibration velocity, It is the torsional vibration velocity.

[0058] Ultimately, Scanlan introduced eight dimensionless flutter derivatives. , , i =1,2,3,4, and finally, the self-excited force of the bridge cross-section in the flow field is approximately expressed as a linear function of the state vector, i.e.:

[0059] (2)

[0060] (3)

[0061] In the formula B The characteristic width of the entire bridge, K To reduce the frequency ( ).

[0062] Based on the flutter theory described above, the torsional component of the obtained free vibration time history information was separated using Python and subjected to Hilbert transformation to obtain the time-varying amplitude. The total damping was obtained by fitting the time-varying amplitude and then subtracting the experimental mechanical damping using equation (4). C S Identify respectively Four dimensionless constants were used to finally obtain the self-excited aerodynamic torque on the main beam section by substituting them into the measured data.

[0063] (4)

[0064] in This represents the aerodynamic self-excited force generated on the main girder section under the action of vertical velocity, torsional angular velocity, vertical displacement, and torsional displacement, respectively. Based on this, for the torsional degree of freedom of the main girder, [the following is considered]... As the self-excited force on the right-hand side of the motion equation of the gyro stabilizer-main beam, as shown in equation (5), the coupled motion control equation between the stabilizer and the main beam is established.

[0065] (5)

[0066] in a Represents the torsional displacement of the main beam. I m , c , k These represent the moment of inertia, torsional damping, and torsional stiffness of the main beam, respectively.

[0067] (2) Intelligent control of gyro stabilizers

[0068] To achieve intelligent control of the gyro stabilizer, simulation technology is used to establish a simulation calculation model of the main beam and the gyro stabilizer, such as... Figure 5 As shown.

[0069] This model is used to simulate a prototype bridge. The aerodynamic self-excited force per unit angle of the main beam, calculated using flutter derivatives, is input into the main beam model to simulate the motion state of the main beam. At the same time, the rotation speed of the gyro stabilizer is continuously changed to record the motion state of the main beam, thereby simulating different working conditions. The relationship between the attitude of the main beam and the rotational angular velocity and precession velocity of the gyro stabilizer under multiple working conditions is established. Through BP neural network training, the motion state and attitude behavior of the gyro stabilizer are finally intelligently controlled.

[0070] The self-excited aerodynamic torque corresponding to the unit torsion angle is obtained by combining the flutter derivative. According to the principle of gyro precession, the magnitude of the gyro torque is calculated according to formula (6). Its direction is opposite to the precession direction according to Newton's third law and the principle of action and reaction.

[0071] (6)

[0072] in The moment of inertia represents the rotation of the gyroscope rotor; The angular velocity of the gyroscope rotor's spin; It represents the precession angular velocity.

[0073] As shown in the above formula, the magnitude of the gyroscope torque is related to the precession angular velocity, moment of inertia, and spin angular velocity. For practical bridge applications, the mass of the gyroscope rotor in this device is set to 1000. kg The servo brushless motor (motor one) can provide a speed of 10,000-100,000 RPM, while the default initial speed of the gyroscope rotor is 50,000 RPM. Through simulation calculation, the different speeds corresponding to the gyroscope stabilizer under different torsional angles of the main beam are obtained. Therefore, the main beam torsion angle-gyro stabilizer spin angular velocity is used as the training sample. According to the main beam torsion angle in increments of 0.1º, a total of 301 samples are obtained from the main beam torsion angle from -15º to +15º. These samples are divided into two parts in a 9:1 ratio, namely 271 samples as the training set and the remaining 30 sets of data as the test set. A BP neural network is established using Python to train and test the training set (main beam torsion-gyro stabilizer spin angular velocity). Artificial neural network learning technology is used to perform forward calculation and backward propagation to optimize the impact of existing recognition errors and obtain the instantaneous optimal mapping relationship between the main beam torsion angle and the gyro stabilizer spin angular velocity. Then, the precession angular velocity of the gyro stabilizer is calculated by combining equations (5) and (6), and finally the intelligent control of the device attitude of the present invention is realized.

[0074] Figure 6 This is a flowchart of the BP neural network training process used by the device of this invention to establish the relationship between different angles of the main beam and the rotational angular velocity of the gyro stabilizer.

[0075] A back propagation (BP) neural network was used to construct a prediction model for the torsion angle of the main beam and the rotational angular velocity of the gyro stabilizer. After multiple parameter adjustments, the algorithm used three layers of neurons, namely the input layer, hidden layer, and output layer. The Sigmoid function was used as the activation function from the input layer to the hidden layer to preprocess the input data, transforming the unbounded input into a predictable form, as shown in Equation (7). Finally, from the hidden layer to the output layer, under the approximation of the MSE loss function (mean squared error function), the predicted value of the gyro stabilizer's rotational angular velocity will increasingly approach the true value, as shown in Equation (8). The model optimization algorithm adopted the stochastic gradient descent method to optimize the weights and intercept terms of the grid.

[0076] (7)

[0077] in, x The variables representing the input layer are converted to the range [0,1] using equation (7).

[0078] The MSE loss function used in this model is:

[0079] (8)

[0080] in, n Indicates the number of samples. Represents the true value of the variable. This represents the predicted value of the variable, i.e., the output value of the grid.

[0081] like Figure 7 As shown, the input first passes through the Sigmoid activation function and then reaches the hidden layer. The hidden layer then calculates the output value through weights, calculates the loss value through the MSE function, and uses stochastic gradient descent to perform backpropagation within the BP network framework to redistribute the error value for calculation. Finally, the weights and intercept terms are optimized to make the output value continuously approach the true value, improve the accuracy of the output value, and achieve the purpose of intelligently controlling the motion and attitude adjustment of the gyroscope stabilizer through the neural network.

[0082] Under the corresponding main beam torsion angle, the error between the predicted rotational angular velocity and the actual value is within 5%, which meets the engineering requirements. Based on this, the precession angular velocity of the gyro stabilizer is calculated by combining equations (5) to (6), so as to achieve the purpose of using the gyro torque generated by the gyro precession to balance the self-excited aerodynamic torque.

[0083] (3) Improve bridge flutter stability

[0084] As bridge spans continue to increase, simply changing the aerodynamic shape to improve the flutter stability of long-span bridges has certain limitations. This invention utilizes the gyroscopic torque generated during gyro precession to balance the self-excited aerodynamic torque on the bridge, thereby improving the flutter stability of long-span bridges without altering the aerodynamic shape of the bridge cross-section.

[0085] like Figure 8 and Figure 9 As shown, under the continuous action of wind load, the transient angle of attack of a bridge structure in a flow field changes, and the flow field also changes accordingly due to the change in angle of attack. Therefore, under the continuous effect of fluid-structure interaction, the self-excited aerodynamic torque of the cross-section continuously increases, leading to a continuous increase in the torsion of the main beam cross-section. This invention uses a built-in attitude sensor WT9011DCL-BT50 to transmit the inclination angle of the main beam cross-section back to the control box in real time via Bluetooth. Based on prior wind tunnel tests, and according to the flutter derivative principle, the self-excited aerodynamic torque on the main beam per unit angle is effectively given. At this time, the control box converts the calculated precession angular velocity into the rotational speed of the chain drive device. Simultaneously, the gyro stabilizer rotates from top to bottom at the predicted rotational speed. Finally, through the chain drive device, the gyro stabilizer rotates along the precession direction, generating a gyroscopic torque opposite to the precession direction, balancing the external torque and the self-excited aerodynamic torque, thus improving the flutter stability of the main beam cross-section.

[0086] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A control algorithm for an active intelligent multi-directional gyroscope stabilizer, characterized in that, Used to improve the flutter stability of long-span bridges; the gyro stabilizer includes a motor, an attitude angle sensor, a power supply box, a control box, a gyro stabilizer body, a chain drive device, and a slide rail fixing device; The gyroscope stabilizing body includes a spherical frame and a gyroscope rotor set inside the spherical frame. The central axis of the gyroscope rotor is vertically arranged, and the upper and lower ends of the axis are movably connected to the top and bottom ends of the spherical frame, respectively. The top end of the axis is connected to a motor, which is located above the spherical frame. Driven by the motor, the gyroscope rotor rotates inside the spherical frame. An attitude angle sensor is fixed on the motor. The spherical frame is mounted on a support, which includes a column and a ring fixed to the top of the column. The spherical frame is located inside the ring and is connected to the ring by two pins arranged symmetrically on the left and right. The spherical frame can rotate up and down inside the ring with the two pins as fixed points. The chain drive device includes an embedded motor 2 and a chain. One end of the chain is connected to the power output end of the motor 2, and the other end of the chain is connected to the spherical frame through a gear. Power is provided by the motor 2, and the chain drive drives the spherical frame to rotate up and down in the ring with two pins as fixed points, thus realizing the precession of the gyroscope rotor. The control box is connected to the attitude angle sensor, motor one and motor two respectively. The control box analyzes the data transmitted by the attitude angle sensor to give the optimal gyroscope rotation angular velocity and precession angular velocity. It also drives the gyroscope rotor to rotate through motor one and the chain drive device to drive the gyroscope rotor to precession, thereby suppressing the bending and torsional mode vibration of the bridge. The control algorithm has the following steps: S1. Establish the motion control equations for the coupled motion of the gyro stabilizer and the main beam, as follows: in, Represents the torsional vibration of the main beam. c and k represent the moment of inertia, torsional damping, and torsional stiffness of the main beam, respectively. U For the incoming flow velocity ，B Characteristic width of the entire bridge ，K To reduce the frequency, ， h is the vertical vibration of the main beam section under the action of the flow field; a dimensionless constant. , i =1,2,3,4; air density, It is the torsional vibration velocity. Represents vertical vibration velocity. It is torsional acceleration; S2. A simulation calculation model of the main beam and gyro stabilizer is established using simulation technology. This model is used to simulate the prototype bridge. The aerodynamic self-excited force per unit angle of the main beam, calculated using flutter derivatives, is input into the main beam model to simulate the motion state of the main beam. At the same time, the rotational speed of the gyro stabilizer is continuously changed, and the motion state of the main beam is recorded to simulate different working conditions. The relationship between the attitude of the main beam and the rotational angular velocity and precession velocity of the gyro stabilizer under multiple working conditions is established. Then, through BP neural network training, the motion state and attitude behavior of the gyro stabilizer are finally intelligently controlled. The gyro torque generated during the precession of the gyro is used to balance the self-excited aerodynamic torque on the bridge, thereby improving the flutter stability of the bridge over a long span.

2. The control algorithm for the active intelligent multi-directional gyroscope stabilizer as described in claim 1, characterized in that, In step S2, the BP neural network training specifically involves: using a BP neural network to construct a prediction model for the main beam torsion angle and the gyro stabilizer's rotation angular velocity. A three-layer neuron system is selected, consisting of an input layer, a hidden layer, and an output layer. From the input layer to the hidden layer, the Sigmoid function is used as the activation function to preprocess the input data, transforming the unbounded input into a predictable form. Finally, from the hidden layer to the output layer, under the approximation of the MSE loss function, the predicted value of the gyro stabilizer's rotation angular velocity increasingly approaches the true value. The input data is preprocessed to transform the unbounded input into a predictable form. The method and formula are as follows: x This represents the variables in the input layer, and the formula converts the data to the range [0,1]. The MSE loss function is as follows: in, n Indicates the number of samples. Represents the true value of the variable. This represents the predicted value of the variable, i.e., the output value of the grid.

3. The control algorithm for the active intelligent multi-directional gyroscope stabilizer as described in claim 1, characterized in that, The sliding groove fixing device is a mounting plate formed by splicing two rectangular plates. The mounting plate is used to fix the gyroscope stabilizing body to the bridge. The mounting plate is equipped with a stretchable telescopic rod with mounting holes for fixed connection with the bridge. The length of the telescopic rod can be adjusted to adapt to different bridge widths.

4. The control algorithm for the active intelligent multi-directional gyroscope stabilizer as described in claim 1, characterized in that, There are four columns, and the bottom of the columns are fixedly connected to the mounting plate.

5. The control algorithm for the active intelligent multi-directional gyroscope stabilizer as described in claim 1, characterized in that, The spherical frame consists of multiple warp plates and an equatorial plate, with the equatorial plate fixedly connected to the ring of the support by a pin.

Images