A golf shot control method based on variable loft parameters

Abstract

The application discloses a golf ball hitting control method based on variable loft parameters, comprising the following steps: S1: setting a target ball course parameter set T; S2: obtaining a loft control quantity set U through reverse solving based on a mapping model F; S3: executing ball hitting according to the control quantity U; S4: measuring actual ball launching parameters T_meas; and S5: adaptively updating the control quantity U based on an error e. The application has the beneficial effects that: the application realizes high-precision collaborative control of multiple parameters such as ball speed, take-off angle and side spin by inversely solving the loft control quantity from the target ball course parameters, thereby having excellent ball course reproduction capability; meanwhile, the system can output stable and repeatable standard ball hitting samples, provides data benchmarks for AI teaching, and has clear physical meaning in the control process, which is convenient for teaching explanation and error analysis; and the application can be widely applied to various scenes such as simulators, teaching robots and ball tool testing, and has good scalability and industrialization prospects.

Description

Technical Field

[0001] This invention relates to the field of golf ball control technology, specifically a golf shot control method based on variable clubface parameters. Background Technology

[0002] In golf instruction, simulator testing, equipment development, and AI-assisted training, there is often a need to generate standard shot results that are "reproducible, calibrable, and interpretable," such as fixed ball speed, fixed launch angle, and fixed backspin and sidespin, for use in training sample generation, user action comparison analysis, or equipment performance calibration. Common implementation methods currently include: Traditional dual-wheel squeeze or pneumatic catapult ball machines can control the ball speed, but have limited precision in controlling spin and launch angle; mechanical hitting devices can simulate the swing action, but clubface posture and contact process are difficult to control precisely; most systems only achieve a rough ball trajectory simulation by adjusting open-loop parameters such as ball angle or wheel speed, which cannot meet the needs of high-precision teaching and data calibration.

[0003] However, existing technologies generally suffer from the following defects and shortcomings: Incomplete clubface parameter system: Most solutions can only control ball speed or launch angle, and cannot simultaneously and accurately adjust key variables such as clubface loft, opening and closing angle, incident angle, and contact time; Unstable mapping between ball trajectory and control variables: There is a strong coupling relationship between ball speed, launch angle, backspin, and sidespin, making it difficult for existing methods to achieve stable reproduction of the target ball trajectory; Open-loop control is susceptible to interference: Environmental changes, ball condition, and mechanical wear can cause ball trajectory drift, and there is a lack of closed-loop correction or adaptive update mechanisms; Lack of standard teaching balls: There is a lack of a method to output "standard ball trajectory samples," making it difficult for AI teaching systems to establish a unified comparison dataset and evaluation scale; Uncontrollable contact process: In actual ball striking, the contact time between the clubface and the ball and the force pulse pattern have a key impact on spin generation, and existing technologies often ignore this dimension or cannot adjust it.

[0004] Therefore, there is an urgent need for a golf shot control method that can solve the problem by reverse engineering the clubface parameters, dynamically control the contact process, and has closed-loop adaptive capabilities, in order to improve the accuracy of shot reproduction and support the construction of high-standard teaching and training systems. Summary of the Invention

[0005] The purpose of this invention is to provide a golf shot control method based on variable clubface parameters to solve the problems of uncontrollable clubface parameters, low accuracy of ball path reproduction, and lack of closed-loop correction capability in existing golf shot control systems.

[0006] This invention achieves the above objective through the following technical solution: a golf shot control method based on variable clubface parameters, comprising the following steps: S1: Set the target ball trajectory parameter set T, wherein T includes at least the ball's initial velocity V, launch angle θ, backspin ω_back, and sidespin ω_side; S2: Based on the mapping model F, the set of control quantities U of the pole face is obtained by inverse solution. U includes at least the pole face tilt angle α_loft, the pole face opening and closing angle Ψ_face, the incident angle β_path, and the contact time τ_contact. S3: Execute the shot according to the control quantity U, and dynamically coordinate the adjustment of τ_contact and β_path during the contact process; S4: Obtain the actual ball-launching parameters T_meas through the measuring device, and calculate the error e between them and the target parameters T; S5: Adaptively update the control quantity U based on the error e to achieve closed-loop optimization and long-term stability.

[0007] Furthermore, the set of clubface control quantities U may also include one or more of the following: the offset Δc of the striking point relative to the center of the ball, the effective coefficient of friction μ_eff, and the waveform w(t) of the striking force changing over time.

[0008] Furthermore, the inverse solution is achieved through the following optimization formula: U=argmin_U‖F(U)-T_target‖+λ·R(U) R(U) is a constraint term used to limit the range of motion, energy limit, and safe operating range of the mechanism.

[0009] Furthermore, the mapping model F is one of a physics collision model, a data-driven regression model, or a physics and data fusion model.

[0010] Furthermore, the dynamic adjustment during the contact process includes: The control of the contact time τ_contact can be achieved by one of the following methods: flexible contact surface, pre-compression mechanism, or drive waveform; The incident angle β_path is adjusted by controlling the impact path and the end effector attitude. In the initial contact phase, the opening and closing angle Ψ_face is finely adjusted in real time.

[0011] Furthermore, the adaptive update employs one of the following methods: gradient descent, Newton's method, Bayesian optimization, or Kalman filtering.

[0012] Furthermore, the method also includes constructing a standard ball path data structure: Round={T_target,U*,T_meas,e,env,device_state} Used for calibration of teaching systems, comparative training, or equipment self-testing.

[0013] Furthermore, the golf shot control method is applicable to golf simulators, teaching robots, or intelligent shot control devices.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. High-precision ball path reproduction capability: By reverse-engineering the target ball path parameters, the clubface control parameters are solved, enabling coordinated control of multiple parameters such as ball speed, launch angle, sidespin, and backspin, significantly improving reproduction accuracy; 2. Supports the construction of teaching standard balls: It can output stable and repeatable standard shot samples, providing a unified data scale and reference for AI teaching systems; 3. The control process is highly interpretable: there is a clear physical causal relationship between control quantities such as loft angle, opening and closing angle, incident angle, and contact time and ball trajectory parameters, which facilitates teaching explanation and error analysis; 4. Possesses anti-drift and self-adaptive capabilities: Through ball output measurement and closed-loop feedback, it achieves long-term stability and resists the impact of environmental changes and equipment wear; 5. Wide system applicability: It is suitable for various scenarios such as simulators, teaching robots, and ball equipment testing, and has good scalability and industrialization prospects. Attached Figure Description

[0015] Fig. 1 This is a flowchart of the control method of the present invention; Fig. 2 This is a schematic diagram illustrating the mapping between the variable parameters of the clubface and the ball path parameters in this invention; Fig. 3 This is a schematic diagram of the dynamic adjustment of the contact window in this invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below 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.

[0017] Combination Figs. 1 to 3 The golf shot control method based on variable clubface parameters, as shown, includes the following steps: S1: Set the target ball trajectory parameter set T, wherein T includes at least the ball's initial velocity V, launch angle θ, backspin ω_back, and sidespin ω_side; S2: Based on the mapping model F, the set of control quantities U of the pole face is obtained by inverse solution. U includes at least the pole face tilt angle α_loft, the pole face opening and closing angle Ψ_face, the incident angle β_path, and the contact time τ_contact. S3: Execute the shot according to the control quantity U, and dynamically coordinate the adjustment of τ_contact and β_path during the contact process; S4: Obtain the actual ball-launching parameters T_meas through the measuring device, and calculate the error e between them and the target parameters T; S5: Adaptively update the control quantity U based on the error e to achieve closed-loop optimization and long-term stability.

[0018] Example 1:

[0019] Step S1: Preset standard ball path set The system has built-in or user-defined sets of standard ball path target parameters T_target: Straight ball: V=70m / s, θ=12°, ω_back=3000rpm, ω_side=0rpm Left curveball: V=68m / s, θ=14°, ω_back=2800rpm, ω_side=+500rpm Right curveball: V=68m / s, θ=14°, ω_back=2800rpm, ω_side=-500rpm High-arc ball: V=50m / s, θ=30°, ω_back=6000rpm, ω_side=0rpm Low-flying ball: V=75m / s, θ=8°, ω_back=2000rpm, ω_side=0rpm Step S2: Inversely solve for the control quantities of the rod surface The system calls a pre-established mapping model F:U→T, which can be a physical model based on collision mechanics and aerodynamics, and has been experimentally calibrated and corrected. For each target ball path, solve the following optimization problem: U*=argmin_U||F(U)-T_target||+λ·R(U) R(U) represents constraints, such as limiting the rod surface tilt angle α_loft to between 8° and 30° and the contact time τ_contact to between 0.4 and 1.2 ms, to ensure the safety of the mechanism and the accessibility of the motion.

[0020] The solution algorithm can be the Levenberg-Marquardt method or a gradient-based optimization algorithm, and finally output the optimal control variable set U∗, including α_loft, Ψ_face, β_path, τ_contact, Δc, μ_eff, w(t).

[0021] Step S3: Perform the shot and make dynamic adjustments The ball-hitting device (such as a six-degree-of-freedom robotic arm with a variable stiffness rod face mechanism) executes the shot according to U*. During contact: The rod surface preload is adjusted in real time by piezoelectric actuator or voice coil motor to stabilize the actual contact time near the target value of τ_contact; During the initial contact phase (0-0.2ms), the incident angle β_path is finely adjusted using PID control based on the real-time force measurement signal. Maintain the opening and closing angle Ψ_face of the rod before contact, or perform dynamic compensation within ±2° in the early stage of contact to suppress sidespin deviation.

[0022] Step S4: Measurement and Error Calculation A high-speed stereo vision system (frame rate ≥ 2000fps) combined with Doppler radar was used to measure the actual ball launch parameter T_meas. The errors of each parameter were calculated. e_V = V_target - V_meas, e_θ=θ_target-θ_meas, e_ω_back=ω_back,target-ω_back,meas, e_ω_side=ω_side,target-ω_side,meas.

[0023] Step S5: Adaptive Update and Library Generation If any error exceeds the threshold (e.g., |e_V|>0.5m / s or |e_ω_side|>50rpm), the update mechanism is activated: Calculate the control update quantity based on the error vector e and the model Jacobian matrix J = ∂F / ∂U. ; Update control variable: U_new = U* + η·ΔU (η is the learning rate); Repeat steps S3-S4 until the error meets the requirements or the maximum number of iterations (e.g., 10 times) is reached.

[0024] Finally, the successfully calibrated ball trajectory data is saved as a standard configuration file, in the following format: { "shot_type":"draw", "T_target":{"V":68,"theta":14,"omega_back":2800,"omega_side":500}, "U_optimal":{"α_loft":16.2,"Ψ_face":-3.5,"β_path":-2.1,...}, "T_measured":{"V":67.9,"theta":14.1,...}, "calibration_date":"2025-10-01", "device_id":"robot_arm_01" } This configuration file can be directly called by simulator software to generate standard comparison ball paths or system self-calibration.

[0025] Example 2:

[0026] Step S1: Analysis of Teaching Tasks The AI ​​teaching system generates specific hitting tasks based on the student's current training goals (such as "reducing right-hand curve") and specifies the target ball trajectory parameter T_target. For example, the task is to "generate a standard straight ball with a launch angle of 13°, backspin of 2600 rpm, and sidespin close to 0".

[0027] Step S2: Real-time inverse solution and control After receiving T_target, the system uses a lightweight neural network model (already trained offline) for real-time inference, outputting the control quantity U within <10ms. This neural network model takes T_target and current environmental parameters (temperature, humidity) as input and outputs U.

[0028] Step S3: Continuously output standard balls The striking device (such as a compact linear motor-driven striker) performs N=5 consecutive strikes based on U*. During the continuous striking process: A contact force sensor is used to monitor the force pulse waveform w(t) of each shot in real time; If the standard deviation of waveform consistency exceeds 5%, the drive current waveform will be automatically fine-tuned to ensure stable output.

[0029] Step S4: Data Collection and Comparative Analysis The average parameter T_meas of N standard shots is collected using radar and vision systems as a benchmark. Simultaneously, the student's shot data T_user is collected. The AI ​​system performs comparative analysis. Visualizing the differences: Overlaying the student's ball trajectory with the standard ball trajectory on a 3D ballistic graph; Cause tracing: Based on the differences, infer possible technical issues. For example, if the student's ball has too much sidespin, the system can prompt: "Your clubface is too open relative to the path at impact; we suggest checking your grip strength and release timing." Generate training suggestions: Based on the comparison results, recommend targeted training courses or auxiliary equipment (such as pole face angle trainers).

[0030] Step S5: Closed-loop optimization teaching model The "standard ball - student ball" comparison data pairs (T_meas, T_user, error correction suggestions) generated in this teaching task will be stored in the teaching database for continuous optimization of the AI's diagnostic model and recommendation algorithm.

[0031] Example 3:

[0032] Step S1: Set the objective of the comparative experiment Set two sets of target ball trajectories, requiring the ball speed V and launch angle θ to be exactly the same, with only the sidespin directions being opposite: Target A: V = 65 m / s, θ = 15°, ω_side = +400 rpm Target B: V=65m / s, θ=15°, ω_side=-400rpm The backspin ω_back was set to 2500±100rpm.

[0033] Step S2: Primarily control the clubface opening / closing angle and impact point offset. Since the ball speed and takeoff angle are mainly affected by the loft angle α_loft and the incident angle β_path, the system first fixes these two and then solves for α_loft* and β_path that can satisfy V and θ through the inverse model.

[0034] Next, using the clubface opening angle Ψ_face and the horizontal striking point offset Δc as the main control variables for adjusting the sidespin, a local optimization problem is established, and the Ψ_face and Δc* that satisfy the target sidespin are solved.

[0035] Step S3: Fine-tune the pulse waveform within the contact window During the shot, the system finely adjusts the drive waveform w(t) during the main pulse phase (P3) of the contact window: For target A (left-handed), the tangential force component is enhanced in the middle of the pulse, and the contact time τ_contact is slightly extended (by about 0.05 ms) to promote the generation of left-handed lateral rotation; For target B (right-handed), the direction and time distribution of the tangential force are adjusted accordingly.

[0036] Step S4: Real-time monitoring and iterative learning After each shot, measure the sidespin value ω_side,meas. If the target range is not reached, update the control parameters according to the following rules of thumb: ΔΨ_face=k_ψ·(ω_side,target-ω_side,meas) Δ(Δc)=k_c·(ω_side,target-ω_side,meas) Where k_ψ and k_c are the gain coefficients learned from historical data.

[0037] Step S5: Create a "Rotation Control Mapping Table" After multiple iterations of learning, the system is able to establish a high-precision mapping relationship between (Ψ_face, Δc, τ_contact) and ω_side under a specific ball speed and takeoff angle. This mapping table can serve as the core knowledge base of the device for rapid and accurate sidespin control under subsequent conditions.

[0038] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0039] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A golf shot control method based on variable clubface parameters, characterized in that, Includes the following steps: S1: Set the target ball trajectory parameter set T, including the ball's initial velocity V, launch angle θ, backspin ω_back and sidespin ω_side; S2: Based on the mapping model F, the set of control variables U of the pole face is obtained by inverse solution. U includes the pole face tilt angle α_loft, the pole face opening and closing angle Ψ_face, the incident angle β_path, and the contact time τ_contact. S3: Execute the shot according to the control quantity U, and dynamically adjust τ_contact and β_path during the contact process; S4: Measure the actual ball launch parameters T_meas and calculate the error e between the actual parameters and the target parameters; S5: Adaptively update the control quantity U based on the error e to achieve closed-loop optimization.

2. The golf shot control method based on variable clubface parameters according to claim 1, characterized in that: The set of control parameters U for the rod face also includes at least one of the following: impact point offset Δc, effective friction parameter μ_eff, and impact force waveform w(t).

3. The golf shot control method based on variable clubface parameters according to claim 1, characterized in that: The inverse solution is achieved through the following optimization problem: U=argmin_U‖F(U)-T_target‖+λ·R(U) R(U) is a constraint term used to limit the range of motion, energy limit, and safe operating range of the mechanism.

4. The golf shot control method based on variable clubface parameters according to claim 1, characterized in that: The mapping model F is one of the following: a physics collision model, a data-driven regression model, or a physics and data fusion model.

5. The golf shot control method based on variable clubface parameters according to claim 1, characterized in that: The dynamic adjustment during the contact process includes: The control of the contact time τ_contact can be achieved by one of the following methods: flexible contact surface, pre-compression mechanism, or drive waveform; The incident angle β_path is adjusted by controlling the impact path and the end effector attitude. In the initial contact phase, the opening and closing angle Ψ_face is finely adjusted in real time.

6. The golf shot control method based on variable clubface parameters according to claim 1, characterized in that: The adaptive update employs one of the following methods: gradient descent, Newton's method, Bayesian optimization, or Kalman filtering.

7. The golf shot control method based on variable clubface parameters according to claim 1, characterized in that: This also includes constructing a standard ball path data structure: Round={T_target,U*,T_meas,e,env,device_state} Used for calibration of teaching systems, comparative training, or equipment self-testing.

8. The golf shot control method based on variable clubface parameters according to claim 1, characterized in that: The golf shot control method described herein is applicable to golf simulators, teaching robots, or intelligent shot control devices.

Images