A shafting multi-point support adjusting system based on parallel magnetic bearing

Abstract

The application relates to the technical field of shafting alignment, in particular to a shafting multi-point support adjusting system based on a parallel magnetic bearing, which comprises at least three magnetic bearing units distributed along the axial direction of the shafting, each magnetic bearing unit comprises a group of radial magnetic pole pairs and a group of axial magnetic pole pairs, and each magnetic bearing unit is electrically interconnected with a unified control module through a common DC bus; the beneficial effect is that: by electrically interconnecting the multiple magnetic bearing units distributed along the axial direction of the shafting with the unified control module through the common DC bus, and by internally arranging a shafting finite element mechanics model in the unified control module, an optimal support force distribution scheme meeting the static force balance, the torque balance and the critical speed avoidance constraint is solved in real time by the multi-span elastic beam model, and the fundamental transformation from the single-point independent control to the multi-point collaborative global optimization is realized.

Description

Technical Field

[0001] This invention relates to the field of shaft alignment technology, specifically to a multi-point support adjustment system for shafts based on parallel magnetic bearings. Background Technology

[0002] The ship's shafting system is the core transmission component connecting the main engine and propeller, and its alignment directly affects shafting vibration, bearing life, and ship noise levels. Currently, shafting alignment is mostly completed during initial installation by adjusting the base elevation, which is a one-time static adjustment. During long-term ship operation, the shafting alignment continuously deteriorates due to the accumulation of factors such as bearing wear, shafting deflection, and hull structural deformation. However, current technology no longer performs online alignment adjustments, resulting in the shafting system being in a suboptimal alignment state for a long time, leading to increased vibration, accelerated bearing wear, and increased noise.

[0003] In recent years, magnetic bearings have been increasingly used in ship shafting support due to their advantages such as being contactless, lubrication-free, and actively controllable. However, current applications of magnetic bearings mostly involve independent control modes where each bearing operates independently. The support force, support clearance, and operating status of each bearing lack unified coordination, making it impossible to optimize the overall mechanical state of the shafting system. A single magnetic bearing can only solve local problems at its own support point and is insufficient to address the systemic requirements of the shafting system as a continuous elastic body under multiple operating conditions, such as maintaining overall alignment, vibration reduction, noise reduction, and extended lifespan. Summary of the Invention

[0004] The purpose of this invention is to provide a multi-point support adjustment system for shafts based on parallel magnetic bearings, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a multi-point support adjustment system for shaft systems based on parallel magnetic bearings, comprising:

[0006] At least three magnetic bearing units are distributed along the axial direction of the shaft system. Each magnetic bearing unit contains a set of radial magnetic pole pairs and a set of axial magnetic pole pairs. Each magnetic bearing unit is electrically interconnected with a unified control module through a common DC bus.

[0007] The unified control module has a built-in finite element mechanical model of the shaft system, which discretizes the shaft system into a multi-span elastic beam. Each magnetic bearing unit is an elastic support. Based on this mechanical model, the unified control module solves the optimal support force distribution scheme of each support in real time, so that the shaft system can simultaneously satisfy static balance, torque balance and first-order critical speed avoidance constraint at any operating moment.

[0008] The radial magnetic pole pairs of each magnetic bearing unit adopt a Halbach permanent magnet array structure. The magnetization direction angle of the Halbach array is controlled online by a unified control module through an electrically excited bias coil. By changing the direction and amplitude of the bias coil current, the radial support stiffness of each magnetic bearing unit can be continuously adjusted without changing the main magnetic field of the permanent magnet.

[0009] Preferably, each magnetic bearing unit further includes an eddy current damping ring disposed between the magnetic pole surface and the journal. The eddy current damping ring is a non-magnetic and conductive ring, and the gap between it and the journal is adjusted online by a unified control module through a piezoelectric micro-displacement actuator.

[0010] The unified control module increases the gap of the eddy current damping ring at the corresponding frequency to enhance damping in that frequency band, or decreases the gap to reduce damping in that frequency band, based on the amplitude of each component in the vibration spectrum of the shaft system, for selective suppression of wideband vibration.

[0011] Preferably, the unified control module adopts a master-slave collaborative architecture, wherein the magnetic bearing unit closest to the geometric center of the shaft system is defined as the master control unit, and the rest are slave control units;

[0012] The main control unit collects the alignment deviation data at both ends of the shaft system in real time and calculates the overall bending mode amplitude of the shaft system. Based on this, it generates a global stiffness allocation command and sends it to each slave control unit. Each slave control unit performs local compensation adjustment based on local sensor data after executing the global command. The compensation amount is superimposed and executed after being coordinated by the unified control module.

[0013] Preferably, the system further includes a shaft strain monitoring section disposed between two adjacent magnetic bearing units, wherein at least four fiber Bragg grating sensors are evenly distributed along the circumference of the strain monitoring section;

[0014] The unified control module reconstructs the bending deformation curve of the shaft system in this segment based on the strain data of each fiber Bragg grating sensor, and uses the reconstruction result as the real-time boundary condition input of the mechanical model, so that the solution accuracy of the support force distribution scheme is dynamically corrected according to the actual deformation state of the shaft system.

[0015] Preferably, the unified control module includes a shaft life prediction module, which predicts the remaining life of each bearing position based on the cumulative load spectrum of each magnetic bearing unit, the history of journal surface temperature, and the evolution trend of vibration intensity using the Paris crack propagation model.

[0016] When the remaining life of any bearing position is lower than the preset threshold, the unified control module automatically transfers the load of that position to the adjacent bearing position and simultaneously adjusts the support stiffness of each magnetic bearing unit on the transfer path, so that the shaft system maintains the alignment state and vibration level within the allowable range during the progressive failure of the bearing.

[0017] Preferably, the unified control module is equipped with a ship motion compensation channel, which receives pitch, roll and heave signals from ship motion sensors and uses the bearing seat displacement caused by hull deformation as an external disturbance input to the shaft system mechanical model.

[0018] When solving the support force distribution scheme, the unified control module superimposes the hull deformation displacement as a known boundary displacement onto the shaft system elastic beam model, so that the support force command of each magnetic bearing unit pre-compensates for the hull deformation effect, thereby achieving decoupling between the shaft system alignment state and the hull motion.

[0019] Preferably, the system further includes a rigid coupling compensation mechanism between magnetic bearing units, the compensation mechanism being disposed between the bearing seats of two adjacent magnetic bearing units, and comprising a set of piezoelectric ceramic actuators driven by a unified control module;

[0020] When the unified control module detects that the bending deformation of the shaft system between two adjacent magnetic bearing units exceeds a preset threshold, the piezoelectric ceramic actuator generates a small axial thrust to actively correct the local bending of the shaft system, forming a "force-displacement" dual-channel coordinated correction with the adjustment of the support force of the magnetic bearing.

[0021] Preferably, the unified control module adopts a digital twin architecture, including a virtual axis system model that is synchronized with the physical axis system in real time;

[0022] The virtual shaft system model uses physical shaft system sensor data as driving input and control parameters of each magnetic bearing unit as adjustable variables. The unified control module pre-simulates the effect of the support force distribution scheme at the next moment on the virtual model, and sends the scheme verified by simulation to the physical magnetic bearing unit for execution, thus forming a pre-verification closed loop of control commands.

[0023] The virtual shaft system model also includes a bearing wear evolution sub-model. Based on the historical load and temperature data of each magnetic bearing unit, this sub-model predicts the gap growth trend of each bearing position within a preset time window. The unified control module adjusts the initial gap setting value of each magnetic bearing unit in advance accordingly for wear feedforward compensation.

[0024] Preferably, the system further includes a shaft thermal deformation compensation module for the cross magnetic bearing unit. This module has at least three temperature sensing points arranged along the axial direction on the shaft, and the data from each temperature sensing point is input into a unified control module in real time.

[0025] The unified control module calculates the thermal bending deformation and thermal elongation of the shaft system based on the temperature difference data of each temperature sensing point. The thermal bending deformation is added as an additional deflection to the load term of the mechanical model, and the thermal elongation is input as an axial boundary condition. This allows the support force distribution scheme of each magnetic bearing unit to simultaneously compensate for the thermal deformation effect of the shaft system and eliminate the additional alignment deviation caused by the temperature difference.

[0026] Preferably, the system is applied to a complete ship shafting system including a main engine output shaft section, an intermediate shaft section and a stern shaft section, and a magnetic bearing unit is provided at the coupling between each shaft section as an inter-section coupling support point;

[0027] The unified control module treats each shaft segment as a subsystem. The magnetic bearing unit of each subsystem prioritizes the alignment and vibration reduction requirements of its own shaft segment. When the vibration energy of any shaft segment is transmitted to the adjacent shaft segment through the coupling, the unified control module coordinates the magnetic bearing unit of the coupling support point to generate a support force in the opposite phase, blocking the cross-segment transmission path of the vibration energy.

[0028] The unified control module contains a reinforcement learning strategy network trained on the ship's entire life cycle data. This strategy network takes the current shaft vibration characteristics, ship operating conditions and environmental loads as state inputs and the optimal control parameters of each magnetic bearing unit as action outputs. During operation, it continuously interacts and iterates with the digital twin model, so that the control strategy can adaptively evolve with the entire life cycle of the shaft.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] The proposed multi-point support adjustment system for shaft systems based on parallel magnetic bearings achieves this by electrically interconnecting multiple magnetic bearing units distributed along the shaft axis with a unified control module via a common DC bus. The unified control module incorporates a finite element mechanical model of the shaft system, using a multi-span elastic beam model to solve in real-time for the optimal support force distribution scheme that satisfies static equilibrium, torque equilibrium, and avoids constraints at critical speeds. This represents a fundamental shift from "single-point independent control" to "multi-point collaborative global optimization." Simultaneously, each magnetic bearing unit employs a Halbach permanent magnet array in conjunction with an electrically excited bias coil to achieve continuously adjustable radial support stiffness. Furthermore, online adjustment of the piezoelectric micro-displacement of the eddy current damping ring gap enables wide-frequency vibration selection. The system employs a master-slave collaborative architecture that balances global stiffness distribution with local compensation. The fiber Bragg grating strain monitoring section feeds back the actual deformation of the shaft system to the mechanical model in real time to correct boundary conditions. The progressive load transfer driven by the Paris crack model and the "force-displacement" dual-channel correction of the piezoelectric ceramic actuator ensure alignment maintenance during bearing failure. The digital twin pre-verification closed loop and reinforcement learning strategy network enable the simulation and execution of control commands and adaptive evolution throughout the entire life cycle. The ship motion compensation channel and thermal deformation compensation module decouple the shaft system alignment state from the hull motion and temperature field, ultimately ensuring that the shaft system maintains optimal alignment and vibration reduction throughout its entire life cycle, significantly extending the service life of the shaft system and bearings. Attached Figure Description

[0031] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the present invention clear and complete, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only some, not all, embodiments of the present invention, and are merely illustrative of the embodiments of the present invention. They are not intended to limit the embodiments of the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] Example 1: A dual-channel vibration reduction system based on Halbach array bias adjustment and eddy current damping frequency selection suppression.

[0034] This embodiment is applied to an ocean-going cargo ship with a total length of 180m and a total shaft length of 42m, including a 12m main engine output shaft section, an 18m intermediate shaft section and a 12m stern shaft section. The total mass of the shaft system is about 8500kg and the maximum speed is 1200r / min.

[0035] Four magnetic bearing units, numbered M1 to M4, are arranged along the shaft system axis. M1 is located 1.5m from the main unit flange at the front end of the main unit output shaft section, M2 is located at the coupling between the main unit output shaft section and the intermediate shaft section, M3 is located at the coupling between the intermediate shaft section and the tail shaft section, and M4 is located 1.2m from the tail pipe seal at the rear end of the tail shaft section. The distances between adjacent bearing units are 10.5m, 18m, and 10.8m, respectively.

[0036] Each magnetic bearing unit's radial pole pair adopts a Halbach permanent magnet array structure, composed of 8 NdFeB (N48SH) permanent magnets, each measuring 40mm × 30mm × 20mm, with a main magnetic field strength of 1.42T. Each radial pole pair is surrounded by an electrically excited bias coil with 120 turns, a single coil resistance of 2.8Ω, and a rated current range of -5A to +5A.

[0037] The unified control module adjusts the bias coil current in real time based on the shaft vibration spectrum. When the shaft is running at 800 r / min, a peak amplitude of 0.085 mm appears at 240 Hz (second harmonic). The unified control module adjusts the bias coil current of M1 and M4 to +3.2A, reducing the radial support stiffness at these two locations from the reference value of 1.2 × 10⁻⁶. 7 N / m decreased to 8.5×10 6 N / m, and simultaneously adjust the bias coil current of M2 and M3 to -2.8A, thereby increasing the stiffness at these two locations to 1.6×10 N / m. 7 N / m, thereby shifting the first-order bending natural frequency of the shaft system from the measured 18.5Hz to 22.3Hz, avoiding the 13.3Hz excitation frequency and its second harmonic corresponding to 800r / min.

[0038] Each magnetic bearing unit has an eddy current damping ring between its magnetic pole face and the journal. The ring is made of pure copper and has an electrical conductivity of 5.8 × 10⁻⁶. 7 S / m, ring width 15mm, wall thickness 3mm, outer diameter 280mm. The reference gap between the damping ring and the journal is 0.5mm, adjustable by a piezoelectric micro-displacement actuator, with an adjustment range of 0.2mm to 1.0mm and an adjustment resolution of 0.01mm. When the amplitude of the 120Hz component in the shaft vibration spectrum exceeds 0.06mm, the unified control module increases the damping ring gap at M2 to 0.85mm to enhance eddy current damping in this frequency band, with the damping coefficient increased from the reference value of 3.2×10. 3 The N·s / m ratio has been increased to 5.8 × 10⁻⁶. 3 N·s / m; When the amplitude of the 600Hz high-frequency component exceeds 0.03mm, the damping ring gap at M3 is reduced to 0.3mm to reduce the damping of this frequency band and avoid excessive suppression of high-frequency energy, which would lead to insufficient control bandwidth.

[0039] After the above adjustments, under full load conditions of 800 r / min, the vibration amplitude at M1 decreased from 0.12 mm to 0.04 mm, and at M4 it decreased from 0.09 mm to 0.035 mm, resulting in a 62% reduction in the overall vibration intensity of the shaft system.

[0040] Example 2: Multi-point centering adjustment system based on fiber Bragg grating strain reconstruction and master-slave collaborative control.

[0041] This embodiment applies to a chemical tanker with a total length of 156m, a shaft system length of 38m, a main engine rated power of 6800kW, and a maximum shaft system speed of 960r / min.

[0042] Three magnetic bearing units, numbered M1 to M3, are arranged along the axial direction of the shaft system, with spacing of 14m and 16m respectively. M2 is defined as the master control unit, located at the geometric center of the shaft system, 14m from M1 and 16m from M3. M1 and M3 are slave control units.

[0043] A strain monitoring section is set up between M1 and M2, and between M2 and M3. The length of each section is 1 / 3 of the distance between adjacent bearings, that is, the monitoring section between M1 and M2 is about 4.7m long, and the monitoring section between M2 and M3 is about 5.3m long. Six fiber Bragg grating (FBG) sensors, numbered S1 to S6, are evenly distributed along the circumferential direction of each section and are attached to the outer surface of the shaft system. The center frequency of each sensor is 1550nm, the strain measurement range is -3000με to +3000με, the measurement accuracy is ±1με, and the sampling frequency is 5kHz.

[0044] When the spindle speed is 720 r / min, the measured strain values ​​of each FBG sensor in the M1-M2 section are as follows: S1 (0° direction) +820 με, S2 (60° direction) +610 με, S3 (120° direction) +340 με, S4 (180° direction) -150 με, S5 (240° direction) -480 με, S6 (300° direction) -710 με. The unified control module reconstructs the bending deformation curve of this section using the least squares method based on the six-point strain data, calculating the maximum deflection of this section to be 0.062 mm and the bending modal amplitude to be 0.048 mm.

[0045] Based on this, the main control unit M2 generates a global stiffness allocation command: the support stiffness of M1 is set to 1.05 × 10⁻⁶. 7 N / m, M2 support stiffness is set to 1.38×10 N / m. 7 N / m, the stiffness of the M3 support is set to 1.12×10 N / m. 7 N / m. This scheme makes the first natural frequency of the shaft system 20.1Hz at 720r / min, which is 1.675 times the excitation frequency of 12Hz, satisfying the critical speed avoidance requirement (ratio > 1.3).

[0046] Each slave control unit performs local compensation based on the execution of global commands: the local displacement sensor at M1 detects a radial offset of 0.038mm, exceeding the allowable value of 0.03mm. The M1 slave control unit adds a +15% stiffness correction to the global stiffness command and increases the radial support force by 2.3kN. The local vibration sensor at M3 detects a 180Hz component amplitude of 0.055mm. The M3 slave control unit adds a -8% stiffness correction and adjusts the damping ring gap to 0.6mm. The compensation amounts are combined and executed after coordination by the unified control module. Ultimately, the offset at M1 is reduced to 0.021mm, and the 180Hz vibration amplitude at M3 is reduced to 0.032mm.

[0047] Example 3: Progressive load transfer system based on Paris lifetime prediction and piezoelectric ceramic dual-channel correction.

[0048] This embodiment applies to a VLCC tanker with a total length of 200m, a shaft system length of 48m, a main engine rated power of 14800kW, a maximum shaft system speed of 1020r / min, and a total shaft system mass of approximately 18000kg.

[0049] Five magnetic bearing units, numbered M1 to M5, are arranged along the axial direction of the shaft system, with spacings of 8m, 10m, 12m, and 10m respectively. M1 is located on the main output shaft section, M5 is located at the end of the tail shaft, and M2 to M4 are located on the intermediate shaft section.

[0050] The life prediction module within the unified control module calculates based on the cumulative load spectrum of each bearing position. Taking bearing position M3 as an example, the cumulative load spectrum data for this position over the past 18 months of operation is as follows: maximum radial load 42kN, minimum radial load 8kN, and approximately 2.6 × 10⁻⁶ load cycles. 7 The highest surface temperature of the journal reached 78℃, and the root mean square value of the vibration intensity evolved from an initial 0.04 mm / s to the current 0.11 mm / s. The parameters for the Paris crack propagation model were set to C = 1.8 × 10⁻⁶. -12 m=3.2, threshold stress intensity factor ΔKth=5.5MPa·m 1 / 2 The calculated current crack propagation rate at position M3 is 4.7 × 10⁻⁶. -9 m / cycle, remaining lifetime is approximately 8.2 × 10 6 Based on the current average daily operation of 16 hours, the remaining service life is estimated to be about 4.7 months, which is lower than the preset threshold of 6 months.

[0051] The unified control module initiates a gradual load transfer: During the first week, 15% of the load on M3, approximately 4.5 kN, is transferred to M2 and M4, increasing the load on M2 by 2.8 kN and the load on M4 by 1.7 kN. Simultaneously, the support stiffness of M2 is increased from 1.15 × 10⁻⁶ kN. 7 N / m adjusted to 1.28×107 N / m, M4 support stiffness from 1.10×10 7 N / m adjusted to 1.22×10 7 N / m; 10% was transferred again in the second week and another 10% in the third week, until the load on M3 dropped to 55% of the design value by the end of the third week. During the transfer, the maximum vibration amplitude at shaft M2 increased from 0.065mm to 0.078mm, still within the allowable range of 0.09mm, while the vibration amplitude at M3 decreased from 0.092mm to 0.051mm.

[0052] During the load transfer process, the bending deformation of the shaft system between M2 and M3 reached 0.15 mm in the second cycle, exceeding the preset threshold of 0.12 mm. The unified control module drove the piezoelectric ceramic actuator between the bearing housings of M2 and M3 to generate axial thrust. The thrust range of a single actuator is 0-5 kN. In this example, the output thrust is 2.8 kN, which reduces the local bending deformation of the shaft system from 0.15 mm to 0.09 mm within 30 seconds. This correction is completed in conjunction with the adjustment of the magnetic bearing support force.

[0053] Example 4: A whole-axis cooperative control system based on digital twin pre-verification and reinforcement learning adaptive evolution.

[0054] This embodiment is applied to an LNG carrier with a total length of 220m and a shaft system length of 52m, including a 14m main engine output shaft section, a 22m intermediate shaft section and a 16m stern shaft section. The main engine has a rated power of 19800kW, the maximum shaft system speed is 1180r / min, and the total mass of the shaft system is approximately 22000kg.

[0055] Six magnetic bearing units, numbered M1 to M6, are arranged along the axial direction of the shaft system. M1 is located at the front end of the main output shaft section, M2 is located at the coupling between the main output shaft section and the intermediate shaft section, M3 and M4 are located on the intermediate shaft section with a spacing of 10m, M5 is located at the coupling between the intermediate shaft section and the tail shaft section, and M6 is located at the end of the tail shaft section. M2 and M5 at each coupling are defined as inter-section coupling support points.

[0056] In the digital twin architecture of the unified control module, the virtual shaft system model is synchronized with the physical shaft system in real time, with a data refresh cycle of 2ms. The virtual model contains 126 beam elements and 78 nodes, with each magnetic bearing element corresponding to an elastic support node. Before issuing control commands, the unified control module simulates the shaft system response over the next 500ms on the virtual model. For example, when a ship switches from a fully loaded sailing condition with a speed of 14 knots to a ballast sailing condition with a speed of 19.5 knots, the simulation results show that if the full-load parameters are directly applied, the vibration at point M3 will surge to 0.14 mm after 3.2 seconds, exceeding the 0.12 mm threshold. After simulation optimization, the unified control module reduces the stiffness of the M3 support from 1.20 × 10⁻⁶ mm to 1.20 × 10⁻⁶ mm. 7N / m adjusted to 1.42×10 7 N / m, M4 support stiffness from 1.18×10 7 N / m adjusted to 1.05×10 7 The vibration peak value at M3 was 0.088 mm (N / m) verified by the virtual model. After the scheme was approved, it was sent to the physical system for execution. The actual vibration peak value at M3 was 0.091 mm, which was 3.4% different from the simulation.

[0057] The bearing wear evolution sub-model embedded in the virtual model predicts based on historical data that, according to the current load spectrum, the M4 clearance will increase from 0.35mm to 0.48mm within the next 90 days, and the M2 clearance will increase from 0.32mm to 0.45mm within 120 days. Based on this, the unified control module adjusts the initial clearance setting of M4 from 0.35mm to 0.40mm and the initial clearance setting of M2 from 0.32mm to 0.36mm in advance, achieving wear feedforward compensation and ensuring that the clearance growth trend remains within the allowable range after 60 days.

[0058] Five temperature sensing points (PT1000 platinum resistance thermometers) with an accuracy of ±0.1℃ were set along the axial direction of the shaft system. The measured data were: T1=62℃, T2=58℃, T3=51℃, T4=47℃, and T5=44℃. The unified control module calculated the thermal bending deformation of the shaft system to be 0.073mm, convex towards the T1 side, with a thermal elongation of 4.2mm. The thermal bending deformation was added as an additional deflection to the mechanical model, and the thermal elongation was used as the displacement boundary condition for the M6 ​​axial magnetic pole pair, adjusting the M6 ​​axial constraint force from the reference value of 18kN to 14.6kN to release thermal stress.

[0059] The reinforcement learning policy network within the unified control module was trained based on 5 years of full lifecycle operational data of the vessel. Its state space has 48 dimensions, including 12 dimensions of vibration characteristics, 8 dimensions of operating parameters, 6 dimensions of environmental loads, 5 dimensions of temperature field, 8 dimensions of load spectrum, and 9 dimensions of life state. The action space has 18 dimensions, consisting of 6 bearing units, each with 3 control parameters: stiffness, clearance, and damping. The policy network was trained using the PPO algorithm, with a cumulative interaction iteration count of 1.2 × 10⁻⁶. 6 After 18 months of operation, the M3 stiffness parameter output by the policy network was 1.37 × 10⁻⁶. 7 The N / m value was increased by 14.2% compared to the initial design value. This parameter was implemented after verification by the digital twin, resulting in a 28% reduction in vibration intensity at M3 compared to the initial strategy. The strategy network interacts and iterates with the digital twin model every 72 hours to continuously optimize the control parameters.

[0060] When the 85Hz vibration energy generated by the propeller excitation in the tail shaft section (M5-M6 section) is transmitted to the intermediate shaft section, the unified control module detects that the 85Hz vibration amplitude at M5 reaches 0.068mm. It immediately coordinates M5 to generate a reverse phase support force. The radial support force at M5 is superimposed with an alternating component of -3.2kN on the 85Hz component, blocking the cross-section transmission of vibration energy. The 85Hz vibration amplitude at M4 drops from 0.051mm to 0.022mm, and the transmission attenuation ratio reaches 57%.

[0061] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-point support adjustment system for shafts based on parallel magnetic bearings, characterized in that: include: At least three magnetic bearing units are distributed along the axial direction of the shaft system. Each magnetic bearing unit contains a set of radial magnetic pole pairs and a set of axial magnetic pole pairs. Each magnetic bearing unit is electrically interconnected with a unified control module through a common DC bus. The unified control module has a built-in finite element mechanical model of the shaft system, which discretizes the shaft system into a multi-span elastic beam. Each magnetic bearing unit is an elastic support. Based on this mechanical model, the unified control module solves the optimal support force distribution scheme of each support in real time, so that the shaft system can simultaneously satisfy static balance, torque balance and first-order critical speed avoidance constraint at any operating moment. The radial magnetic pole pairs of each magnetic bearing unit adopt a Halbach permanent magnet array structure. The magnetization direction angle of the Halbach array is controlled online by a unified control module through an electrically excited bias coil. By changing the direction and amplitude of the bias coil current, the radial support stiffness of each magnetic bearing unit can be continuously adjusted without changing the main magnetic field of the permanent magnet.

2. The shaft multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: Each magnetic bearing unit also includes an eddy current damping ring disposed between the magnetic pole surface and the journal. The eddy current damping ring is a non-magnetic and conductive ring, and the gap between it and the journal is adjusted online by a unified control module through a piezoelectric micro-displacement actuator. The unified control module increases the gap of the eddy current damping ring at the corresponding frequency to enhance damping in that frequency band, or decreases the gap to reduce damping in that frequency band, based on the amplitude of each component in the vibration spectrum of the shaft system, for selective suppression of wideband vibration.

3. The shaft system multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: The unified control module adopts a master-slave collaborative architecture, in which the magnetic bearing unit closest to the geometric center of the shaft system is defined as the master control unit, and the rest are slave control units; The main control unit collects the alignment deviation data at both ends of the shaft system in real time and calculates the overall bending mode amplitude of the shaft system. Based on this, it generates a global stiffness allocation command and sends it to each slave control unit. Each slave control unit performs local compensation adjustment based on local sensor data after executing the global command. The compensation amount is superimposed and executed after being coordinated by the unified control module.

4. A shaft system multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: The system also includes a shaft strain monitoring section disposed between two adjacent magnetic bearing units, and at least four fiber Bragg grating sensors are evenly distributed along the circumference of the strain monitoring section. The unified control module reconstructs the bending deformation curve of the shaft system in this segment based on the strain data of each fiber Bragg grating sensor, and uses the reconstruction result as the real-time boundary condition input of the mechanical model, so that the solution accuracy of the support force distribution scheme is dynamically corrected according to the actual deformation state of the shaft system.

5. A shaft system multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: The unified control module includes a shaft life prediction module. This module predicts the remaining life of each bearing position based on the cumulative load spectrum of each magnetic bearing unit, the history of journal surface temperature, and the evolution trend of vibration intensity, using the Paris crack propagation model. When the remaining life of any bearing position is lower than the preset threshold, the unified control module automatically transfers the load of that position to the adjacent bearing position and simultaneously adjusts the support stiffness of each magnetic bearing unit on the transfer path, so that the shaft system maintains the alignment state and vibration level within the allowable range during the progressive failure of the bearing.

6. A shaft system multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: The unified control module is equipped with a ship motion compensation channel, which receives pitch, roll, and heave signals from ship motion sensors and uses the bearing seat displacement caused by hull deformation as an external disturbance input to the shaft system mechanical model. When solving the support force distribution scheme, the unified control module superimposes the hull deformation displacement as a known boundary displacement onto the shaft system elastic beam model, so that the support force command of each magnetic bearing unit pre-compensates for the hull deformation effect, thereby achieving decoupling between the shaft system alignment state and the hull motion.

7. A shaft system multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: The system also includes a rigid coupling compensation mechanism between magnetic bearing units. The compensation mechanism is located between the bearing seats of two adjacent magnetic bearing units and includes a set of piezoelectric ceramic actuators driven by a unified control module. When the unified control module detects that the bending deformation of the shaft system between two adjacent magnetic bearing units exceeds a preset threshold, the piezoelectric ceramic actuator generates a small axial thrust to actively correct the local bending of the shaft system, forming a "force-displacement" dual-channel coordinated correction with the adjustment of the support force of the magnetic bearing.

8. A shaft system multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: The unified control module adopts a digital twin architecture, including a virtual axis system model that is synchronized with the physical axis system in real time; The virtual shaft system model uses physical shaft system sensor data as driving input and control parameters of each magnetic bearing unit as adjustable variables. The unified control module pre-simulates the effect of the support force distribution scheme at the next moment on the virtual model, and sends the scheme verified by simulation to the physical magnetic bearing unit for execution, thus forming a pre-verification closed loop of control commands. The virtual shaft system model also includes a bearing wear evolution sub-model. Based on the historical load and temperature data of each magnetic bearing unit, this sub-model predicts the gap growth trend of each bearing position within a preset time window. The unified control module adjusts the initial gap setting value of each magnetic bearing unit in advance accordingly for wear feedforward compensation.

9. A shaft system multi-point support adjustment system based on parallel magnetic bearings according to claim 1, characterized in that: The system also includes a shaft thermal deformation compensation module for the cross magnetic bearing unit. This module has at least three temperature sensing points set along the axial direction on the shaft. The data from each temperature sensing point is input into a unified control module in real time. The unified control module calculates the thermal bending deformation and thermal elongation of the shaft system based on the temperature difference data of each temperature sensing point. The thermal bending deformation is added as an additional deflection to the load term of the mechanical model, and the thermal elongation is input as an axial boundary condition. This allows the support force distribution scheme of each magnetic bearing unit to simultaneously compensate for the thermal deformation effect of the shaft system and eliminate the additional alignment deviation caused by the temperature difference.

10. A multi-point support adjustment system for shafts based on parallel magnetic bearings according to claim 1, characterized in that: The system is applied to the complete shafting system of a ship, which includes the main engine output shaft section, the intermediate shaft section and the stern shaft section. A magnetic bearing unit is set at the coupling between each shaft section as an inter-section coupling support point. The unified control module treats each shaft segment as a subsystem. The magnetic bearing unit of each subsystem prioritizes the alignment and vibration reduction requirements of its own shaft segment. When the vibration energy of any shaft segment is transmitted to the adjacent shaft segment through the coupling, the unified control module coordinates the magnetic bearing unit of the coupling support point to generate a support force in the opposite phase, blocking the cross-segment transmission path of the vibration energy. The unified control module contains a reinforcement learning strategy network trained on the ship's entire life cycle data. This strategy network takes the current shaft vibration characteristics, ship operating conditions and environmental loads as state inputs and the optimal control parameters of each magnetic bearing unit as action outputs. During operation, it continuously interacts and iterates with the digital twin model, so that the control strategy can adaptively evolve with the entire life cycle of the shaft.

Images