Helicopter rotor blade profile design method
By employing intelligent adaptive deformation profile modeling, chaotic perturbation particle swarm optimization, and multidisciplinary collaborative optimization, the problems of local optima and multi-objective conflicts in helicopter rotor blade design using traditional methods have been solved, achieving efficient multi-condition adaptation and performance improvement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JINGDEZHEN CERAMIC UNIV
- Filing Date
- 2024-12-09
- Publication Date
- 2026-04-21
AI Technical Summary
Traditional PSO algorithms are prone to getting trapped in local optima in helicopter rotor blade optimization design. They are difficult to generate highly diverse solution sets in high-dimensional and complex design spaces, have high optimization time costs, and are difficult to respond quickly to the needs of multi-task scenarios. Furthermore, traditional methods are difficult to balance conflicting objectives such as aerodynamic performance, noise control, and structural strength.
We employ intelligent adaptive deformation profile modeling, chaotic perturbation particle swarm optimization, dynamic proxy model-driven optimization, and multidisciplinary collaborative optimization methods. Combining piecewise Bézier curves and smart materials, we adjust particle positions through Logistic mapping, construct a Gaussian process regression model, and dynamically adjust disciplinary weights to achieve multi-objective balance.
It significantly improves the global search capability and solution set diversity of rotor blade design, enhances aerodynamic efficiency, reduces noise, optimizes structural strength, adapts to multiple operating conditions, rapidly responds to multiple mission scenarios, and improves helicopter rotor performance.
Smart Images
Figure CN121902285A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of helicopter technology, specifically to a method for designing the cross-sectional profile of a helicopter rotor blade. Background Technology
[0002] The rotor system provides lift, thrust, and control force for helicopters, but it also introduces vibration and noise. Proper rotor structure design plays a crucial role in improving helicopter performance. Within the rotor system, the blades directly bear aerodynamic forces and transmit them to the fuselage to complete propulsion and other maneuvers; therefore, the blade structure is a key component of rotor system design.
[0003] Because the length of the blade structure is much greater than the dimensions of its cross-section in both directions, beam theory is used to analyze the blade structure. The earliest method for beam section analysis was the Euler-Bernoulli beam, in which the cross-section remains unchanged during deformation and is always perpendicular to the centerline of the cross-section. Later, Timoshenko introduced shear deformation based on the former method, but assumed that the shear deformation was uniformly distributed across the cross-section, thus failing to reflect the actual blade conditions. In actual blade structures, non-classical factors such as shear deformation and warping deformation have a significant impact, rendering traditional beam theory no longer applicable.
[0004] Currently, the main methods for cross-section analysis are analytical methods and finite element methods. Analytical methods use analytical expressions to describe shear and warping deformations, which are difficult to apply to complex curved surfaces. The finite element method, by establishing a finite element model of the cross-section, obtains the cross-sectional properties of complex cross-sectional shapes. Due to its good versatility and high angularity, it is widely used in blade structure design.
[0005] Traditional PSO (Programmable Optimization Search) is prone to getting trapped in local optima during the optimization process, causing the optimization results to deviate from the global optimum, thus reducing the quality of the design and the comprehensiveness of the search. Existing PSO algorithms have limited ability to explore the design space, especially in high-dimensional and complex design spaces, resulting in a small coverage of the solution set and difficulty in generating highly diverse Pareto front solutions. Consequently, when dealing with conflicting objectives such as aerodynamic performance, noise control, and structural strength, traditional methods require a large number of iterations, resulting in high optimization time costs, slow convergence speed, and difficulty in quickly responding to the needs of multi-tasking scenarios. Furthermore, the optimization methods cannot effectively address the multi-condition requirements of helicopters and cannot simultaneously consider the performance optimization of the rotor in different states such as forward flight, hovering, and climb. Summary of the Invention
[0006] To address the shortcomings of existing technologies, this invention provides a helicopter rotor blade profile design method. This method solves the problems of traditional PSO (Programmable Optimization Solution) which is prone to getting trapped in local optima during the optimization process, causing the optimization results to deviate from the global optimal solution, reducing the quality and comprehensiveness of the design scheme, and having a small coverage of the solution set in high-dimensional and complex design spaces, making it difficult to generate highly diverse Pareto front solutions. Furthermore, it suffers from high optimization time costs, slow convergence speed, and difficulty in quickly responding to the needs of multi-task scenarios when dealing with conflicting objectives such as aerodynamic performance, noise control, and structural strength.
[0007] To achieve the above objectives, the present invention provides the following technical solution: A method for designing the profile of a helicopter rotor blade includes the following steps: S1, Intelligent Adaptive Deformable Profile Modeling: First, the geometric parameters of the rotor blades are parametrically modeled, and the profile shape is described by piecewise Bézier curves. At the same time, smart materials are introduced to design a dynamically deformable profile, which can adjust the thickness and camber in real time and optimize the performance under different flight conditions. It reduces the risk of stall in the retreating blade area, improves aerodynamics in the advancing blade area, and reduces tip vortex noise. S2, Chaotic Perturbation Particle Swarm Optimization: Initialize design variables and initial particle swarm, set search space, update particle position and velocity using PSO formula, and introduce chaotic perturbation mechanism to dynamically adjust particle position using Logistic mapping to enhance global search capability and ensure the diversity and globality of the exploration design space until a preliminary Pareto front solution set is generated. S3, Dynamic Proxy Model Driven Optimization: Construct a dynamic proxy model based on Gaussian process regression. The dynamic proxy model is used to reduce the computational burden of CFD simulation. The proxy model is dynamically updated during the optimization process. The model prediction accuracy is improved by sampling new design points, and the next optimization point is selected by combining the maximization probability improvement criterion. S4, Multidisciplinary Collaborative Optimization: Integrates aerodynamic performance, noise level and structural strength objectives into a multidisciplinary collaborative optimization framework, and adopts a decomposition strategy to optimize multiple objectives in parallel, in order to ensure the dynamic balance of performance indicators of different disciplines and output the optimized solution set.
[0008] Preferably, in S1, the geometric parameters include thickness distribution, curvature distribution, and sweep angle, and the piecewise Bézier curve formula is: ,in As control points, The smart material comprises shape memory alloys or piezoelectric materials, and its deformation formula is given by the Bessel basis function. ,in , For amplitude, The value is the deformation frequency.
[0009] Preferably, in S2, the design variables include section thickness and curvature, and the PSO formula is: , The formula for dynamically adjusting particle positions using Logistic mapping is as follows: ,in .
[0010] Preferably, in S3, the surrogate model formula is: ,in It is a mean function. The perturbation term of the Gaussian process, kernel function Used to describe the correlation between design variables.
[0011] Preferably, in S3, the formula for selecting the optimization point using the criterion is MPI. ,in It is a normal distribution function. To predict uncertainties, further accelerate the optimization convergence.
[0012] Preferably, in S4, the optimization framework objective includes maximizing aerodynamic performance. Minimize noise level Structural strength guarantee By dynamically adjusting the subject weights To resolve the objective conflict, the overall optimization function is: ,in .
[0013] This invention provides a method for designing the cross-sectional profile of a helicopter rotor blade. It has the following beneficial effects: 1. This invention introduces a chaotic perturbation mechanism into particle swarm optimization and uses Logistic mapping to dynamically adjust the position and velocity of particles, significantly enhancing the global search capability and avoiding getting trapped in local optima. This method effectively improves the exploration capability and solution set diversity of the design space, and has a significant improvement in optimization efficiency compared with traditional methods. It quickly generates a preliminary Pareto front solution set, improves the aerodynamic efficiency of the rotor, reduces noise, and optimizes structural strength. It is suitable for multi-condition and multi-task scenarios and has the effect of improving the performance of helicopter rotors.
[0014] 2. This invention uses piecewise Bézier curves to parametrically model the geometric parameters of rotor blades. By combining the intelligent deformation capabilities of shape memory alloys or piezoelectric materials, dynamic adjustment of the blade profile is achieved. This method can reduce the risk of stall in the retreating blade region, improve aerodynamic efficiency in the advancing blade region, and significantly reduce noise caused by tip vortices. It effectively adapts to multiple operating conditions and provides support for the efficient and stable operation of helicopters.
[0015] 3. This invention integrates aerodynamic performance, noise level, and structural strength into a unified multidisciplinary collaborative optimization framework. By decomposing strategies and optimizing multiple objectives, and using a dynamic weight adjustment method to resolve objective conflicts, this framework effectively maximizes aerodynamic performance, minimizes noise level, and ensures structural strength, thereby ensuring the dynamic balance of multi-objective design and generating a more comprehensive set of optimized solutions. Attached Figure Description
[0016] Figure 1 This is a flowchart of a helicopter rotor blade profile design method according to the present invention. Detailed Implementation
[0017] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on 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.
[0018] Example: Please see the appendix Figure 1 This invention provides a method for designing the cross-sectional profile of a helicopter rotor blade, comprising the following steps: S1, Intelligent Adaptive Deformable Profile Modeling: First, the geometric parameters of the rotor blades are parametrically modeled, and the profile shape is described by piecewise Bézier curves. At the same time, smart materials are introduced to design a dynamically deformable profile, which can adjust the thickness and camber in real time and optimize the performance under different flight conditions. It reduces the risk of stall in the retreating blade area, improves aerodynamics in the advancing blade area, and reduces tip vortex noise. S2, Chaotic Perturbation Particle Swarm Optimization: Initialize design variables and initial particle swarm, set search space, update particle position and velocity using PSO formula, and introduce chaotic perturbation mechanism to dynamically adjust particle position using Logistic mapping to enhance global search capability and ensure the diversity and globality of the exploration design space until a preliminary Pareto front solution set is generated. S3, Dynamic Proxy Model Driven Optimization: Construct a dynamic proxy model based on Gaussian process regression. The dynamic proxy model is used to reduce the computational burden of CFD simulation. The proxy model is dynamically updated during the optimization process. The model prediction accuracy is improved by sampling new design points, and the next optimization point is selected by combining the maximization probability improvement criterion. S4, Multidisciplinary Collaborative Optimization: Integrates aerodynamic performance, noise level and structural strength objectives into a multidisciplinary collaborative optimization framework, and adopts a decomposition strategy to optimize multiple objectives in parallel, in order to ensure the dynamic balance of performance indicators of different disciplines and output the optimized solution set.
[0019] In S1, the geometric parameters include thickness distribution, curvature distribution, and sweep angle. The piecewise Bézier curve formula is: ,in As control points, For Bessel basis functions, smart materials include shape memory alloys or piezoelectric materials, and their deformation formula is: ,in , For amplitude, The value is the deformation frequency.
[0020] Specifically, the blade's geometric parameters include thickness distribution, camber distribution, and sweep angle. A piecewise Bézier curve is used to parametrically model the blade profile, enabling efficient design space exploration and optimization. The formula for the piecewise Bézier curve is:
[0021] in, These are control points used to define the thickness and curvature characteristics of the profile. For Bessel basis functions, this formula simplifies the complex blade profile shape into a linear combination of a small number of control points, enabling the optimization algorithm to efficiently handle the geometric variables of the profile. By adjusting the position of the control points, the piecewise Bessel curve can flexibly adapt to various profile design requirements, such as optimizing stall characteristics in the retreating propeller region and improving the lift-to-drag ratio in the advancing propeller region. The dynamic adjustment of the profile is achieved through smart materials such as shape memory alloys or piezoelectric materials. These materials can actively adjust the profile shape according to external signals, and the deformation formula is:
[0022] in, This is the initial profile. This is the deformed cross-section. Indicates the amount of dynamic adjustment to the shape. For amplitude, As the deformation frequency, by controlling the amplitude and frequency, the profile can be optimized in real time according to the flight conditions. In the retreating propeller region, the stall is delayed and the lift is increased by increasing the camber amplitude. In the high-speed advancing propeller region, the drag is reduced and the drag ratio is optimized by reducing the camber amplitude. In the hovering or transition state, the vibration and turbulence noise are reduced by optimizing the thickness distribution and fine-tuning the sweep angle.
[0023] Firstly, the intelligent deformable profile can adjust its geometry in real time according to the helicopter's dynamic operating conditions, such as hovering, forward flight, and climb, thereby improving the overall aerodynamic efficiency and stability of the rotor. Secondly, by actively adjusting the sweep angle and camber distribution, it significantly reduces vortex noise and blade tip shock wave noise. Thirdly, under different flight conditions, it achieves a balance between high lift-to-drag ratio and low drag by adjusting the profile geometry parameters, thus meeting the requirements of multi-objective design. The dynamic adjustment of the profile can smooth the aerodynamic load distribution, reduce blade vibration, and improve overall flight comfort.
[0024] In S2, the design variables include section thickness and curvature, and the PSO formula is: , The formula for dynamically adjusting particle positions using Logistic mapping is: ,in .
[0025] Specifically, let's first understand the PSO optimization principle: Particle Swarm Optimization is a swarm intelligence algorithm that simulates particles searching for optimal solutions in the solution space. The update formulas for particle velocity and position are:
[0026]
[0027] in: : No. The particle in the first The speed of each iteration; The position of the particle indicates the values of thickness and curvature; The particle's own historical best position; : Global optimal position; The learning factor controls the convergence speed of particles toward their individual and global optimal positions. Inertial weights control the ability to explore and develop particles.
[0028] Using the above formula, each particle can dynamically adjust the values of thickness and curvature, gradually approaching the global optimum of the multi-objective optimization solution; Then, a chaotic perturbation is introduced: In traditional PSO, particles may converge to a local optimum prematurely. To enhance search capability, a Logistic mapping is introduced as a chaotic perturbation to dynamically adjust the particle's position, as shown in the formula:
[0029] in: The particle's current position; Chaos control parameters are typically: This ensures that the Logistic mapping possesses chaotic properties. The introduction of chaotic perturbations breaks the fixed convergence pattern of the particle swarm optimization (PSO), enhances population diversity, expands the search range, and ensures global exploration capabilities. For blade profile optimization, this means that thickness and camber distributions can explore a wider design space, avoiding getting trapped in local optima. The chaotic properties of the Logistic mapping endow the PSO with the ability to dynamically adjust, enabling it to find superior thickness and camber combinations globally, improving optimization efficiency. In rotor profile design, thickness and camber distributions may be subject to multiple constraints related to structural strength, manufacturing processes, and aerodynamic performance. The introduction of chaotic perturbations allows PSO to flexibly seek high-quality solutions that satisfy constraints under complex conditions.
[0030] In S3, the surrogate model formula is: ,in It is a mean function. The perturbation term of the Gaussian process, kernel function Used to describe the correlation between design variables. In S3, the formula for selecting the optimization point is MPI. ,in It is a normal distribution function. To predict uncertainties, further accelerate the optimization convergence.
[0031] Specifically, to mitigate the limitations of computationally expensive CFD simulations on optimization efficiency, a Gaussian process regression model is employed to construct a dynamic surrogate model for rapidly predicting the impact of profile design variables on the aerodynamic performance, noise level, and structural strength of performance targets. The prediction formula for the surrogate model is as follows:
[0032] in: Mean function, representing design variables The expected value of the forecast; The random perturbation term of the Gaussian process is used to capture prediction errors; Kernel functions describe the correlation between design variables; a common form is the radial basis function (RBF).
[0033] in, For signal variance, This is the kernel width parameter.
[0034] The surrogate model learns the design variables of historical sampling point profiles and their corresponding CFD simulation results, providing fast and accurate performance predictions for subsequent optimization steps, thereby significantly reducing computational costs. Then, the MPI criterion for optimizing point selection is applied: During the optimization process, new sampling points are dynamically selected by introducing the probability maximization improvement criterion, thereby accelerating convergence to the global optimum. The MPI formula is:
[0035] in: The standard normal distribution function represents the improved probability; Current forecast value; The optimal value in the current optimization process; Prediction uncertainty is determined by the kernel function. The decision is made by balancing predictive performance and model uncertainty, ensuring that the selection of optimal points considers both the potential for performance improvement and the ability to explore unexplored areas.
[0036] In S4, the optimization framework objectives include maximizing aerodynamic performance. Minimize noise level Structural strength guarantee By dynamically adjusting the subject weights To resolve the objective conflict, the overall optimization function is: ,in .
[0037] Specifically, a multidisciplinary optimization framework is employed to coordinate multiple performance objectives in helicopter rotor blade design, including maximizing aerodynamic performance, minimizing noise levels, and ensuring structural strength. By introducing a dynamic weight adjustment mechanism, the optimization framework can achieve a balance among different objectives, avoiding performance trade-offs caused by conflicting objectives. Among these: maximizing aerodynamic performance and optimizing profile design to improve lift-to-drag ratio:
[0038] in, The lift coefficient, This is the drag coefficient; Minimize noise levels by reducing noise levels caused by tip vortices and turbulence:
[0039] Where, p This represents the sound pressure level at the noise source. The sound source region; Structural strength is ensured by optimizing the cross-sectional thickness distribution and material parameters to guarantee a structural strength safety margin.
[0040] in, For the material's yield strength, This represents the maximum stress.
[0041] To resolve conflicts among multiple objectives, the optimization framework employs a dynamic weight adjustment method, adjusting the weights of each objective in real time based on the priority requirements of the current design phase. The overall optimization function is .
[0042]
[0043] in: : No. The weights of each objective can be dynamically adjusted. : No. Objective functions for each discipline; Objective functions include aerodynamics, noise, and structural strength. This is achieved through dynamic adjustment. The optimized framework can focus on strengthening a specific objective based on the needs of the design phase; During the pre-flight optimization phase, increase the weight of aerodynamic performance. To ensure the maximum lift-to-drag ratio; During the hovering optimization phase, increase the noise control weight. Reduce tip vortex noise; In high-load scenarios, increase the structural strength weight. This improves the safety margin of the design.
[0044] 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 method for designing the cross-sectional profile of a helicopter rotor blade, characterized in that, Includes the following steps: S1, Intelligent Adaptive Deformable Profile Modeling: First, the geometric parameters of the rotor blades are parametrically modeled, and the profile shape is described by piecewise Bézier curves. At the same time, smart materials are introduced to design a dynamically deformable profile, which can adjust the thickness and camber in real time and optimize the performance under different flight conditions. It reduces the risk of stall in the retreating blade area, improves aerodynamics in the advancing blade area, and reduces tip vortex noise. S2, Chaotic Perturbation Particle Swarm Optimization: Initialize design variables and initial particle swarm, set search space, update particle position and velocity using PSO formula, and introduce chaotic perturbation mechanism to dynamically adjust particle position using Logistic mapping to enhance global search capability and ensure the diversity and globality of the exploration design space until a preliminary Pareto front solution set is generated. S3, Dynamic Proxy Model Driven Optimization: Construct a dynamic proxy model based on Gaussian process regression. The dynamic proxy model is used to reduce the computational burden of CFD simulation. The proxy model is dynamically updated during the optimization process. The model prediction accuracy is improved by sampling new design points, and the next optimization point is selected by combining the maximization probability improvement criterion. S4, Multidisciplinary Collaborative Optimization: Integrates aerodynamic performance, noise level and structural strength objectives into a multidisciplinary collaborative optimization framework, and adopts a decomposition strategy to optimize multiple objectives in parallel, in order to ensure the dynamic balance of performance indicators of different disciplines and output the optimized solution set.
2. The helicopter rotor blade profile design method according to claim 1, characterized in that: In S1, the geometric parameters include thickness distribution, curvature distribution, and sweep angle, and the piecewise Bézier curve formula is: ,in As control points, The smart material is a Bessel basis function, comprising shape memory alloys or piezoelectric materials, and its deformation formula is: ,in , For amplitude, The value is the deformation frequency.
3. The helicopter rotor blade profile design method according to claim 1, characterized in that: In S2, the design variables include section thickness and curvature, and the PSO formula is: , The formula for dynamically adjusting particle positions using Logistic mapping is as follows: ,in .
4. The helicopter rotor blade profile design method according to claim 1, characterized in that: In S3, the surrogate model formula is: ,in It is a mean function. The perturbation term of the Gaussian process, kernel function Used to describe the correlation between design variables.
5. The helicopter rotor blade profile design method according to claim 1, characterized in that: In S3, the formula for selecting the optimization point based on the criterion is MPI. ,in It is a normal distribution function. To predict uncertainties, further accelerate the optimization convergence.
6. The helicopter rotor blade profile design method according to claim 1, characterized in that: In S4, the optimization framework objective includes maximizing aerodynamic performance. Minimize noise level Structural strength guarantee By dynamically adjusting the subject weights To resolve the objective conflict, the overall optimization function is: ,in .