A voice interaction golf smart ball and training control method and system

Abstract

The application discloses a kind of golf intelligent ball and training control method and system of voice interaction, wherein the method comprises: voice wake-up and parse user natural language instruction, generate structured training parameters;Safety interlock check is carried out through multi-source sensor;According to parameter generation and execution, ball, ball and ball position adjustment action sequence;After successfully putting, generate unique round ID and prompt;Collect the result of hitting ball, and training parameters, action log, safety data and the result of hitting ball are bound and stored, and execute according to rhythm cycle.The system includes voice acquisition, semantic understanding, action planning, ball execution, safety interlock and round binding module.The application realizes the whole process automatic control and data closed loop of voice-driven golf training, improves training efficiency, safety and data degree, and supports adaptive intelligent training optimization.

Description

Technical Field

[0001] This invention relates to the field of intelligent golf training technology, specifically to a voice-interactive intelligent golf ball-laying and training control method and system. Background Technology

[0002] In golf practice (including driving ranges, chipping areas, putting areas, and simulators), golfers typically need to repeatedly complete the cycle of "retrieving the ball, placing the ball / on the tee, setting practice goals, hitting the ball, and reviewing the play." Currently, the main technologies and equipment in this field include: passive ball-feeding baskets or fairway ball-feeding devices, which have limited functionality and cannot achieve personalized control; fixed / simple mechanical tees that require manual placement or adjustment, making it difficult to ensure consistent ball height and affecting the training rhythm; and while some indoor simulators have built-in ball-laying or ball-placing functions, they are usually limited to fixed points, and their "ball placement" operation is disconnected from core training processes such as setting training goals and motion analysis. In addition, the application of existing voice interaction technology in golf training is mostly limited to peripheral auxiliary functions such as playing music, timing, or scoring, failing to delve into closed-loop control of core training processes such as "ball placement, ball positioning, and training motion sequence."

[0003] In actual training, golfers often need to quickly switch between various practice modes according to the training plan (such as high lob shot, low chip shot, distance gradient practice, designated landing point training, etc.) and hope to conveniently set a series of combined parameters (such as target distance, landing area, ball position height and offset, number of training sets and rhythm, etc.). However, existing technologies generally suffer from the following defects and shortcomings: First, the process of placing the ball and tee relies on manual labor, which is inefficient and difficult to standardize, affecting the comparability of training data and disrupting the training rhythm; second, there is a lack of an interactive method that can naturally and efficiently parameterize complex training intentions; third, existing devices lack a reliable multi-factor safety interlock mechanism, posing a safety risk of malfunction due to voice misrecognition, interference from others, or personnel approaching; in addition, key data in the training process (such as training intentions, ball position parameters, ball placement actions, and shot results) are isolated from each other and cannot be bound to the same training unit, making training review difficult and hindering personalized analysis and guidance; finally, existing equipment has weak scene adaptability and is difficult to flexibly adapt to different training environments (such as outdoor driving ranges and indoor simulators) and different training subjects (such as full swing and putting). Summary of the Invention

[0004] The purpose of this invention is to solve the problems in the background art mentioned above, and to provide a golf training method and system that can achieve intelligent control of the entire ball-laying process through natural language interaction, ensure operational safety, and form a closed loop of training data.

[0005] This invention achieves the above objective through the following technical solution: a voice-interactive intelligent golf ball-playing and training control method, comprising the following steps: S1. Voice wake-up and command input: The system is activated by a wake word and receives natural language commands from the user that contain training intentions; S2. Semantic parsing and parameter generation: Recognize and semantically understand the received natural language training instructions, and output structured current training parameters P, wherein the current training parameters P include at least pattern parameters, target parameters and ball position parameters; S3. Safety Interlock Judgment: Collects multi-source safety signals through sensors and performs safety verification based on preset interlock judgment rules. If the verification passes, proceed to the next step; if the verification fails, enters a waiting or prompting state. S4. Action Sequence Generation and Execution: Based on the current training parameters P, generate and execute action sequence A to drive the ball feeding and placement mechanism to complete feeding, ball placement, and ball position adjustment, and verify whether the ball is in place using a sensor; if the verification fails, enter the abnormal handling or shutdown state; if the verification succeeds, proceed to the next step; S5. Round ID Generation and Notification: The round binding and data recording module generates a unique round ID, and the system notifies the user that "ready". S6. Shot Result Acquisition and Round Data Binding: Collect the result data after the user hits the ball. The round binding and data recording module binds the structured training parameter P, the execution log of the action sequence A, the safety interlock judgment related data and the shot result data to the round ID for storage. S7. Loop execution: After waiting according to the rhythm interval T in the training structure parameters, return to steps S3 to S6 until all training balls set in the training structure parameters are completed.

[0006] Furthermore, in step S2, the specific process of recognizing and semantically understanding natural language instructions and outputting structured training parameters P includes: The speech recognition module converts voice commands into text sequences. Based on a pre-defined knowledge base of golf training parameters, a combination of rule parsing and machine learning is used to extract and map at least one of the following from the text sequence: training mode, target distance, landing area, number of balls, tee height, ball position offset, rhythm interval T, and safety constraint distance D_safe, forming structured training parameters P.

[0007] Furthermore, in step S3, the multi-source security signal includes signals from at least two of the following types of sensors: Ultrasonic sensors, millimeter-wave radar, or vision sensors are used to detect the approach of a golf club. Infrared sensors, ToF sensors, or vision sensors used to detect the approach of a human body; A force sensor or touch sensor is installed on the ball-positioning actuator; Visual sensors, photoelectric sensors, or pressure sensors used for position detection.

[0008] Furthermore, in step S3, the preset interlock determination rule includes: If the distance between the ball stick or human body and the actuator is detected to be less than the safety constraint distance D_safe, then the feeding mechanism and the tee adjustment mechanism are prohibited from moving. If the force sensor or touch sensor at the ball end detects abnormal pressure or touch signal, the shutdown protection will be triggered. If any abnormality is detected in the feeding mechanism, ball placement mechanism, or tee adjustment mechanism, such as ball jamming, failure to reach the correct position, or failure to return to the correct position, a shutdown protection will be triggered. If the voice wake-up or command is not confirmed by an authorized sound source, only a prompt will be given without any action will be performed; If the ball's position verification fails, the system will be prevented from entering the "start of round" state.

[0009] Further, in step S4, the logic for generating the action sequence A is as follows: Based on the ball position parameters in the current training parameters P, a sequence of actions including feeding, tee height adjustment, and ball placement is planned. Configure the corresponding actuator drive command and expected position verification logic for each timing action; The timing actions, driving instructions, and position verification logic are combined into an executable continuous action sequence A.

[0010] Furthermore, in step S6, the data bound to the round ID includes at least: Training configuration data P_i: includes mode, target, ball position parameters, and training structure parameters; The data A_i for ball placement and positioning includes the status of the actuator, adjustment amount, positioning verification result, and safety interlock trigger record. Shot result data O_i: includes at least one of the following: ball trajectory, landing coordinates, deviation, and swing action analysis indicators collected after the shot.

[0011] Furthermore, the method also includes: The training effect is evaluated based on one or more consecutive rounds of bound data. The evaluation indicators include at least one of the following: the deviation trend between the shot result and the target parameters, the convergence degree of the swing action analysis indicators, and the systematic deviation of the ball path direction. When the evaluation results meet the preset adjustment conditions, the training parameters P_{i+1} for the next round or the next set of rounds are automatically generated to drive the system to perform subsequent training and form a training closed loop.

[0012] Furthermore, the method for automatically generating training parameters P_{i+1} includes: Based on the deviation trend between the shot result and the target parameters, the system automatically fine-tunes the target distance or landing area; based on the convergence degree of the swing action analysis indicators, it automatically adjusts the tee height, ball position offset, or training rhythm; and based on the systematic deviation of the ball trajectory direction, it automatically generates a specialized training mode that includes path correction.

[0013] A voice-interactive intelligent golf ball-laying and training control system, applied to any one of the methods described in the voice-interactive intelligent golf ball-laying and training control method, the system comprising: The voice acquisition module is used to receive the user's wake word and natural language training instructions; A semantic understanding and parameter generation module, connected to the speech acquisition module, is used to parse the natural language training instructions into structured training parameters P; An action planning module, connected to the semantic understanding and parameter generation module, is used to generate an executable action sequence A based on the training parameters P; The ball feeding and ball placement actuator is connected to the motion planning module and is used to execute the motion sequence A to complete the feeding, ball placement and ball position adjustment. The sensor and safety interlock module, connected to the ball loading and placement actuator, is used to collect multi-source safety signals and perform safety verification according to preset interlock rules; and the round binding and data recording module, connected to the semantic understanding and parameter generation module, the action planning module and the sensor and safety interlock module, is used to generate a unique round ID for each successful ball placement, and bind the training parameter P, the execution log of the action sequence A, the safety interlock related data and the ball hitting result data to the round ID for storage.

[0014] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any one of the methods described in the "Voice Interactive Smart Golf Ball-Taking and Training Control Method".

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. Through an automatic ball loading and placement mechanism, precise and consistent ball positioning (such as T-shirt height and position) is achieved, avoiding errors caused by manual placement, ensuring the comparability of training data, and reducing the time trainees spend bending over to retrieve and place balls, significantly improving training rhythm and efficiency. Users can directly control the system using natural language commands containing rich training intentions (such as target distance, landing area, number of balls, rhythm, etc.), simplifying complex parameter settings into a single voice command, greatly reducing the operational threshold and learning cost, and enabling training intentions to be quickly and accurately translated into equipment actions.

[0016] 2. By integrating signals from multiple sensors, including those for club approach, human approach, force / touch, voiceprint authorization, and device status, and performing real-time safety checks based on preset rules, this system effectively prevents malfunctions or safety incidents caused by voice misrecognition, unauthorized personnel entry, or device abnormalities, ensuring safe use in complex practice environments. The system introduces the concept of a "round ID," binding and storing the voice training parameters, device execution logs, safety status snapshots, and shot results (trajectory, landing point, motion analysis, etc.) corresponding to each successful ball placement. This provides a refined data foundation for reviewing training effectiveness, analyzing movement techniques, and providing personalized evaluations.

[0017] 3. Building upon the basic closed loop, the system can automatically evaluate training effectiveness (such as deviation trends and movement stability) based on the binding data from consecutive rounds, and intelligently generate the next set of training parameters, enabling dynamic and personalized adjustments to the training content. This forms an autonomous progressive training closed loop of "setting-execution-evaluation-optimization," improving the scientific nature and effectiveness of training. Attached Figure Description

[0018] Figure 1 This is a system structure block diagram of the present invention; Figure 2 This is a flowchart of the method of the present invention; Figure 3 This is a schematic diagram of the safety interlock of the present invention. Detailed Implementation

[0019] 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.

[0020] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0021] Example 1: Basic Method Flow This embodiment describes the basic implementation process of the method of the present invention, such as... Figure 2 As shown, the specific steps include: S101: Voice wake-up and command input.

[0022] Users activate the system using a preset wake word (e.g., "Golf Assistant"). Once activated, the user issues training commands in natural language. These commands can contain a variety of training intentions, such as: "Chip practice, target 60 yards, hit 10 balls, lower the tee slightly, 6-second interval between each ball." The system collects speech signals through a near-field microphone array and performs noise reduction and enhancement processing to improve recognition rates in noisy practice environments.

[0023] S102: Semantic parsing generates training parameters P.

[0024] The speech recognition engine converts the collected speech into a text sequence. Subsequently, the semantic understanding module parses the text. This module incorporates a golf-specific dictionary and grammar rules, and integrates a user habit learning model. After parsing, it outputs a set of structured, machine-executable training parameters P.

[0025] Taking the above instruction as an example, the generated parameter P can include: Mode parameters: Cut-off mode.

[0026] Target parameters: Target distance d = 60 yards.

[0027] Ball position parameters: tee height h = -10% (adjusted 10% lower than the standard height).

[0028] Training structure parameters: number of balls per group K=10, rhythm interval T=6 seconds.

[0029] Safety constraint parameters: safety distance D_safe = 1 meter, voiceprint verification enabled.

[0030] For any part of the instruction that is not specified (such as the number of groups N), the system will call the user's historical default value (such as N=1) to complete it.

[0031] S103: Safety interlock judgment.

[0032] Before generating an action instruction, the system must pass the safety check. As Figure 3 shown, the safety interlock module collects multi-source sensor signals in real time, including: Distance detection: through the millimeter-wave radar installed on the side of the ball placing mechanism, detect the distances D_club and D_human between the tip of the club or the human torso and the mechanism.

[0033] Status detection: detect abnormal contact F_touch through the micro pressure sensor at the ball placing end; detect whether a golf ball is in place OK_ball through the photoelectric sensor on the ball placing platform; verify whether the current instruction comes from an authorized user AUTH through voiceprint comparison.

[0034] The system makes a judgment according to preset rules: if D_club<D_safe or D_human<D_safe or F_touch is abnormal, it is judged as "prohibited from execution", and all moving mechanisms are locked. If AUTH fails, it is judged as "only prompt", the system broadcasts "unauthorized instruction" but does not execute the action. Only when all safety signals pass the check, it is judged as "execution allowed", and the process enters the next step.

[0035] S104 / S105: Action sequence generation and execution.

[0036] The motion planning module generates a specific, time-sequenced action sequence A according to the parameter P. For the instruction "chip shot 60 yards, lower the tee", the sequence A may be: Feeding action: control the solenoid valve of the material bin to release a golf ball to the feeding guide rail.

[0037] Ball position adjustment action: control the stepping motor that drives the telescopic tee to lower the tee height to the target value (h=-10%).

[0038] Ball placing action: control the ball placing mechanical arm or pushing mechanism to accurately place the ball at the end of the guide rail on the top of the lowered tee.

[0039] In-place check action: trigger the photoelectric sensor again to confirm that the ball blocking signal is stable, and read the data of the pressure sensor to confirm that the ball is stably placed on the tee.

[0040] If all checks pass, the system performs an "unlock" operation, that is, allows the current round to enter the hittable state.

[0041] S106 / S107 / S108: Abnormal handling and round ready.

[0042] If any sensor feedback timeout or anomaly occurs during the execution of the action (such as ball jamming during feeding, tee not reaching the designated position, or ball not being in place), the process jumps to S107, triggering the anomaly handling procedure (such as emergency stop of the mechanism, reverse action to remove the jammed ball), and prompts the user to intervene via voice or light. If all actions are successful and the verification passes (S106 successful), the system generates a globally unique round ID (such as R_20250321_001) in S108 and announces "Position 1 is ready" to prompt the user to start hitting the ball.

[0043] S109 / S110: Shots and Data Recording.

[0044] The user swings the club to hit the ball. Optional shot acquisition modules (such as high-speed cameras or radar velocimeters) begin collecting data such as initial velocity, launch angle, sidespin, flight trajectory, and landing coordinates, denoted as O_i. Subsequently, the round binding module packages and binds all data from this round: training setting parameters P_i, action execution log A_i, sensor snapshot data S_i during the safety interlock process, and the shot result O_i. All this data is indexed by round IDR_i and stored in the database.

[0045] S111: Cyclic control.

[0046] The system checks whether all K balls in this group have been played. If not, it waits for a set interval T (e.g., 6 seconds) and then automatically returns to S105 to generate and execute the next ball's placement sequence based on the same parameter P. This cycle repeats until all K balls are played, at which point the process ends.

[0047] Example 2: Implementation of a complete system integrating intelligent training closed loop This embodiment describes a complete system implementation that integrates an intelligent training closed loop. Its hardware configuration can be found in [reference needed]. Figure 1 The modular architecture shown.

[0048] 1. System hardware composition: Voice acquisition module (1): includes a directional microphone array and a wake word processing chip, used to accurately pick up user commands in ambient noise.

[0049] Semantic understanding and parameter generation module (2): Implemented by an embedded processor or cloud server, it runs a natural language processing model and completes the mapping from text to structured parameters P.

[0050] Motion planning module (3): Receives parameter P, combines it with the kinematic model of the actuator, and plans a collision-free and highly efficient motion sequence A.

[0051] Ball loading and placement actuator (4): includes a ball storage bin, a feeding mechanism driven by a motor, a ball placement robotic arm or push rod that can be finely adjusted in three dimensions, and a telescopic tee unit or lifting platform whose height can be precisely controlled by a servo motor.

[0052] Sensor and safety interlock module (5): It integrates the aforementioned millimeter-wave radar, photoelectric sensor, pressure sensor, voiceprint recognition unit, etc. All signals are connected to a dedicated safety PLC (programmable logic controller) to realize millisecond-level dangerous condition judgment and hardware-level interlock.

[0053] Round binding and data recording module (6): It consists of a database server and data analysis software, and is responsible for data storage, management and analysis.

[0054] 2. Intelligent training closed-loop workflow: This embodiment enhances the functions of steps S6 and S7 based on embodiment one, forming a self-evolving training system.

[0055] The user's first input command: "Practice a 100-yard full swing, hit 5 balls, and see my trajectory." The system completed the ball placement, hitting, and data binding for the first group of 5 balls (R_1 to R_5).

[0056] Evaluation Phase: The intelligent training adjustment module analyzed the bound data from R_1 to R_5. The evaluation revealed: ① The average landing distance was 97 yards, systematically 3 yards short; ② The launch angle dispersion was large; ③ The swing rhythm of the last two balls (R_4, R_5) was significantly faster.

[0057] Parameter tuning phase: Based on the rule model, the system automatically generates the next set of training parameters P': The target distance was adjusted from 100 yards to 103 yards to compensate for systematic bias.

[0058] Add a voice prompt to the instructions to “focus on swing rhythm stability” (as an implicit constraint).

[0059] Increase the tempo interval T by 2 seconds from the default value to smooth out the user's gradually increasing tempo.

[0060] Closed-loop execution phase: Without requiring the user to speak again, the system automatically uses the new parameter P' to drive the actuator, setting a new target (103 yards) for the 6th ball (R_6) and placing the ball. After the user hits R_6, the data is recorded, and the system re-enters the evaluation-adjustment loop.

[0061] In this way, training shifts from passive "voice control" to active "AI coach-led" training, allowing the system to dynamically respond to the user's real-time performance and provide personalized adjustments to the training content.

[0062] Example 3: Multi-scenario Adaptation Application Example The method and system of this invention have good scene generalization ability, and can be adapted to different training areas by adjusting the parameter P and the action sequence A. Driving range hitting position: Use a telescopic tee, mainly responding to commands such as "full swing" and "target distance", with a relatively large safety distance D_safe setting (e.g., 1.5 meters).

[0063] Chip / Pitch area: Replaced with a flat ball placement platform or short grass simulation mat. The semantic understanding module focuses on parsing instructions such as "high lob", "low lob", and "landing area". The shot result O_i focuses on recording "roll / stop performance".

[0064] Putting Green: Using a tee-less ball-placement platform, the system interprets instructions such as "putt 10 feet, downhill left line". The ball position parameter P incorporates "slope level s" and "grass texture effect g". The shot result mainly records the initial direction and the final stopping point.

[0065] The core logic of the safety interlock rules remains unchanged in different scenarios, but the sensor configuration can be optimized accordingly (e.g., adding visual anti-trampling detection in the ridge area).

[0066] 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.

[0067] 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 voice-interactive intelligent golf ball-laying and training control method, characterized in that: Includes the following steps: S1. Voice wake-up and command input: The system is activated by a wake word and receives natural language commands from the user that contain training intentions; S2. Semantic parsing and parameter generation: Recognize and semantically understand the received natural language training instructions, and output structured current training parameters P, wherein the current training parameters P include at least pattern parameters, target parameters and ball position parameters; S3. Safety Interlock Judgment: Collects multi-source safety signals through sensors and performs safety verification based on preset interlock judgment rules. If the verification passes, proceed to the next step; if the verification fails, enters a waiting or prompting state. S4. Action Sequence Generation and Execution: Based on the current training parameters P, generate and execute action sequence A to drive the ball feeding and placement mechanism to complete feeding, ball placement, and ball position adjustment, and verify whether the ball is in place using a sensor; if the verification fails, enter the abnormal handling or shutdown state; if the verification succeeds, proceed to the next step; S5. Round ID Generation and Notification: The round binding and data recording module generates a unique round ID, and the system notifies the user that "ready". S6. Shot Result Acquisition and Round Data Binding: Collect the result data after the user hits the ball. The round binding and data recording module binds the structured training parameter P, the execution log of the action sequence A, the safety interlock judgment related data and the shot result data to the round ID for storage. S7. Loop execution: After waiting according to the rhythm interval T in the training structure parameters, return to steps S3 to S6 until all training balls set in the training structure parameters are completed.

2. The voice-interactive intelligent golf ball-laying and training control method according to claim 1, characterized in that: In step S2, the specific process of recognizing and semantically understanding natural language instructions and outputting structured training parameters P includes: The speech recognition module converts voice commands into text sequences. Based on a pre-defined knowledge base of golf training parameters, a combination of rule parsing and machine learning is used to extract and map at least one of the following from the text sequence: training mode, target distance, landing area, number of balls, tee height, ball position offset, rhythm interval T, and safety constraint distance D_safe, forming structured training parameters P.

3. The voice-interactive intelligent golf ball-laying and training control method according to claim 1, characterized in that: In step S3, the multi-source security signal includes signals from at least two of the following types of sensors: Ultrasonic sensors, millimeter-wave radar, or vision sensors are used to detect the approach of a golf club. Infrared sensors, ToF sensors, or vision sensors used to detect the approach of a human body; A force sensor or touch sensor is installed on the ball-positioning actuator; Visual sensors, photoelectric sensors, or pressure sensors used for position detection.

4. The voice-interactive intelligent golf ball-laying and training control method according to claim 3, characterized in that: In step S3, the preset interlock determination rules include: If the distance between the ball stick or human body and the actuator is detected to be less than the safety constraint distance D_safe, then the feeding mechanism and the tee adjustment mechanism are prohibited from moving. If the force sensor or touch sensor at the ball end detects abnormal pressure or touch signal, the shutdown protection will be triggered. If any abnormality is detected in the feeding mechanism, ball placement mechanism, or tee adjustment mechanism, such as ball jamming, failure to reach the correct position, or failure to return to the correct position, a shutdown protection will be triggered. If the voice wake-up or command is not confirmed by an authorized sound source, only a prompt will be given without any action will be performed; If the ball's position verification fails, the system will be prevented from entering the "Start of Round" state.

5. The voice-interactive intelligent golf ball-laying and training control method according to claim 1, characterized in that: In step S4, the logic for generating the action sequence A is as follows: Based on the ball position parameters in the current training parameters P, a sequence of actions including feeding, tee height adjustment, and ball placement is planned. Configure the corresponding actuator drive command and expected position verification logic for each timing action; The timing actions, driving instructions, and position verification logic are combined into an executable continuous action sequence A.

6. The voice-interactive intelligent golf ball-laying and training control method according to claim 1, characterized in that: In step S6, the data bound to the round ID includes at least: Training configuration data P_i: includes mode, target, ball position parameters, and training structure parameters; The data A_i for ball placement and positioning includes the status of the actuator, adjustment amount, positioning verification result, and safety interlock trigger record. Shot result data O_i: includes at least one of the following: ball trajectory, landing coordinates, deviation, and swing action analysis indicators collected after the shot.

7. The voice-interactive intelligent golf ball-laying and training control method according to claim 6, characterized in that: The method further includes: The training effect is evaluated based on one or more consecutive rounds of bound data. The evaluation indicators include at least one of the following: the deviation trend between the shot result and the target parameters, the convergence degree of the swing action analysis indicators, and the systematic deviation of the ball path direction. When the evaluation results meet the preset adjustment conditions, the training parameters P_{i+1} for the next round or the next set of rounds are automatically generated to drive the system to perform subsequent training and form a training closed loop.

8. The voice-interactive intelligent golf ball-laying and training control method according to claim 7, characterized in that: The methods for automatically generating training parameters P_{i+1} include: Based on the deviation trend between the shot result and the target parameters, automatically fine-tune the target distance or landing area; Based on the convergence of swing action analysis metrics, the tee height, ball offset, or training pace are automatically adjusted. Based on the systematic deviation of the ball's trajectory, a specialized training mode including path correction is automatically generated.

9. A voice-interactive intelligent golf ball-laying and training control system, characterized in that: The system, used to implement the method as described in any one of claims 1 to 8, comprises: The voice acquisition module is used to receive the user's wake word and natural language training instructions; A semantic understanding and parameter generation module, connected to the speech acquisition module, is used to parse the natural language training instructions into structured training parameters P; An action planning module, connected to the semantic understanding and parameter generation module, is used to generate an executable action sequence A based on the training parameters P; The ball feeding and ball placement actuator is connected to the motion planning module and is used to execute the motion sequence A to complete the feeding, ball placement and ball position adjustment. The sensor and safety interlock module, connected to the ball loading and placement actuator, is used to collect multi-source safety signals and perform safety verification according to preset interlock rules; and the round binding and data recording module, connected to the semantic understanding and parameter generation module, the action planning module and the sensor and safety interlock module, is used to generate a unique round ID for each successful ball placement, and bind the training parameter P, the execution log of the action sequence A, the safety interlock related data and the ball hitting result data to the round ID for storage.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.

Images