Biomimetic combat robot and method of use thereof

Abstract

The application discloses a kind of bionic combat robot and its use method, including bionic mechanical body, multimodal perception module, data fusion pretreatment module, martial arts game layered autonomous decision module, motion execution control module, power management module, it is characterized by: the bionic mechanical body is multi-degree-of-freedom humanoid combat structure, can complete parry, split, dodge, rush, retreat, side attack and defense martial arts combat action, multimodal synchronous fusion perception innovation: break through the limitation of traditional single visual perception, integrate five-dimensional perception unit of vision, distance, force perception, posture, speed, realize the full-dimension accurate perception of combat scene enemy state, environment state, equipment self state, eliminate single sensor error by data fusion algorithm, greatly improve the perception stability and accuracy under complex confrontation scene.

Description

Technical Field

[0001] This invention relates to the field of robot-related products, specifically to a biomimetic fighting robot and its usage method. Background Technology

[0002] Most bionic fighting robots rely on pre-set motion libraries, fixed logic control, or a single visual perception mode, exhibiting significant shortcomings in intelligence and autonomous game-playing capabilities. Firstly, their perception is limited, depending solely on camera vision to identify target position and movements. They cannot perceive multi-dimensional information such as opponent's force, speed, distance, posture, joint status, and environmental interference, making them highly susceptible to recognition errors due to changes in ambient light and opponent feints, resulting in extremely poor anti-interference capabilities. Secondly, their control logic is rigid, often involving pre-programmed fighting actions and the cyclical execution of fixed attack and defense commands. They lack autonomous game-playing thinking and cannot dynamically adjust tactics based on the opponent's real-time moves and rhythm, making them highly vulnerable to defeat in flexible combat scenarios. Thirdly, their decision-making mechanisms are flat, lacking the layered thinking of martial arts attack and defense logic. They cannot implement anthropomorphic fighting strategies such as prediction, probing, attack-defense switching, move deconstruction, and endgame strategy, resulting in an intelligence level far below that of real-life martial arts game-playing.

[0003] In existing technologies, some robots attempt to incorporate multi-sensor fusion perception, but these only achieve simple overlay of environmental data without optimizing perception for combat scenarios. Furthermore, they lack hierarchical decision-making algorithms adapted to the martial arts offensive and defensive system, failing to distinguish between different levels of combat such as probing, defense, counterattack, offensive, and evasion. This results in problems such as decision lag, chaotic moves, and a disconnect between offense and defense, making it difficult to meet the demands of high-intensity, high-dynamic, and intelligent autonomous combat. Therefore, there is an urgent need to develop a biomimetic combat robot with multimodal precise perception and hierarchical autonomous decision-making capabilities in martial arts combat, enabling human-like, intelligent, and autonomous combat operations. Summary of the Invention

[0004] The purpose of this invention is to provide a biomimetic fighting robot and its usage method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a bionic fighting robot, comprising a bionic mechanical body, a multimodal perception module, a data fusion preprocessing module, a martial arts game hierarchical autonomous decision-making module, a motion execution control module, and a power management module, characterized in that: the bionic mechanical body is a multi-degree-of-freedom bionic fighting structure, capable of performing martial arts fighting actions such as blocking, striking, dodging, rushing, retreating, and side-stepping attack and defense;

[0006] The multimodal perception module includes a visual perception unit, a distance perception unit, a force perception unit, a posture perception unit, and a speed perception unit, which are used to synchronously collect multi-dimensional real-time data on the opponent's status, the distance between the opponent and ourselves, the intensity of attack and defense, our own posture, and the speed of movement in the fighting scene.

[0007] The data fusion preprocessing module is used to receive multimodal raw data, complete timestamp synchronization, noise filtering, data normalization and feature weighted fusion, and generate a standardized fighting scene feature matrix.

[0008] The martial arts game-playing layered autonomous decision-making module is the core intelligent decision-making unit. It has a built-in five-layered decision-making architecture, which completes autonomous game-playing decisions through scene perception, situation analysis, tactical decision-making, move execution, and dynamic correction in sequence, and outputs attack and defense execution commands.

[0009] The motion execution control module receives attack and defense commands, drives the bionic mechanical body to complete the corresponding fighting actions, and provides real-time feedback on the action execution status.

[0010] The power management module provides stable power to each module and has overload protection.

[0011] As a preferred embodiment of the present invention, the martial arts game hierarchical autonomous decision-making module includes a five-layer architecture from top to bottom, namely, scene perception layer, situation analysis layer, tactical decision-making layer, move execution layer, and dynamic correction layer.

[0012] The scene perception layer is used to identify the types of fighting scenes such as confrontation, probing, attack, entanglement, and evasion, and to complete the state labeling;

[0013] The situation assessment layer quantifies the opponent's threat level, the enemy's and our own security coefficient, and our own stability coefficient through a multi-dimensional scoring model to determine the current game's tone.

[0014] The tactical decision-making layer corresponds to different game themes, matches preset martial arts offensive and defensive tactics, and outputs tactical instructions for probing, defense, counterattack, strong attack, and risk avoidance.

[0015] The move execution layer breaks down tactical instructions into specific martial arts fighting action combinations, generating joint motion trajectories, servo parameters, and action speed parameters.

[0016] The dynamic correction layer dynamically corrects tactical and move parameters based on real-time perception data and motion feedback data, achieving closed-loop adaptive adjustment.

[0017] As a preferred embodiment of the present invention, the multimodal perception module adopts a 100Hz millisecond-level synchronous sampling mechanism, and the data of each sensor are calibrated with a unified timestamp. The data fusion preprocessing module adopts an improved Kalman filter algorithm and a feature weighted fusion algorithm to enhance the weight of core data in high-threat scenarios.

[0018] As a preferred embodiment of the present invention, the martial arts game-playing hierarchical autonomous decision-making module has a built-in martial arts move library, which includes basic offensive and defensive moves and combination moves, supports iterative updates, and adapts to game-playing needs with different levels of confrontation.

[0019] A method for using a biomimetic fighting robot includes the following steps:

[0020] S1. Equipment initialization: Start the robot and complete sensor self-test, algorithm initialization, model loading and servo motor calibration.

[0021] S2, Multimodal Real-time Perception: The multimodal perception module synchronously collects multi-dimensional raw data of the combat scene;

[0022] S3. Data fusion preprocessing: The raw data is denoised, synchronized, and normalized, and then fused to generate a fighting scene feature matrix;

[0023] S4. Layered Autonomous Game Decision-Making: Through a five-layer martial arts game decision-making framework, the system sequentially completes scene recognition, situation assessment, tactical matching, move decomposition, and parameter correction, and outputs attack and defense execution commands.

[0024] S5. Combat Action Execution: Drives the bionic mechanical body to complete corresponding martial arts offensive and defensive actions, achieving autonomous combat;

[0025] S6. Dynamic closed-loop correction: Iteratively perceives, integrates, decides, and executes processes, adapting to changes in the adversarial situation in real time;

[0026] S7. End Standby: After detecting the termination signal, the robot resets its posture and enters a low-power standby state.

[0027] As a preferred embodiment of the present invention, in step S4, the situation analysis layer quantifies and generates six game themes: safe standoff, passive defense, active probing, precise counterattack, all-out offensive, and emergency avoidance, and matches them with differentiated offensive and defensive tactics.

[0028] As a preferred embodiment of the present invention, in step S6, the decision module completes parameter updates and action corrections every 10ms to achieve rapid adaptive response to opponent changes and environmental interference.

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

[0030] 1. Multimodal synchronous fusion perception innovation: Breaking through the limitations of traditional single visual perception, it integrates five-dimensional perception units of vision, distance, force, posture and speed to achieve full-dimensional accurate perception of enemy and friendly status, environmental status and equipment status in combat scenarios. Through data fusion algorithms, it eliminates the error of single sensors and greatly improves the perception stability and accuracy in complex combat scenarios.

[0031] 2. Innovation of Layered Autonomous Decision-Making Algorithm in Martial Arts Game: Abandoning the traditional flat and fixed control logic, it pioneers a five-layered game decision-making architecture, which completely replicates the game thinking of real martial arts: "viewing the scene - judging the situation - determining tactics - selecting moves - adjusting actions". It achieves a technological breakthrough from passively executing instructions to actively thinking about the game and can autonomously adapt to various confrontation scenarios and opponent changes. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the structure of a biomimetic fighting robot according to the present invention;

[0033] Figure 2 This is a schematic diagram of the bionic mechanical body of a bionic fighting robot according to the present invention;

[0034] Figure 3 This invention relates to a biomimetic fighting robot and its usage method. Detailed Implementation

[0035] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0036] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," "front end," "rear end," "both ends," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and for 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. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0037] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0038] Please see Figure 1-3This invention provides an embodiment of a biomimetic fighting robot and its usage method. The biomimetic mechanical body adopts a 17-DOF humanoid structure, with 1 DDOF for the head, 6 DDOF for each arm, 2 DDOF for the torso, and 2 DDOF for the legs, adaptable to various basic martial arts attack and defense movements. The multimodal perception module includes a 1080P resolution binocular camera equipped with a YOLO real-time target recognition model, capable of accurately identifying opponent limb movements and attack / defense trajectories; an infrared ranging sensor array of 8 groups, covering all-around monitoring of the robot's front, back, left, and right sides; a force sensor with a range of 0-50N, accurately collecting the force of attack and defense contact; and a six-axis IMU sensor monitoring the robot's posture in real time with a sampling accuracy of 0.01°.

[0039] The data fusion preprocessing module uses an improved Kalman filter algorithm to filter out interference noise such as light jitter and mechanical vibration. Through a feature-weighted fusion model, it strengthens the weight of distance and force data for close-range assault scenarios and strengthens the weight of visual and speed data for long-range probing scenarios, thereby improving scenario adaptability.

[0040] Workflow of the hierarchical autonomous decision-making module in martial arts game theory:

[0041] 1. Scene Awareness Layer: Real-time identification of scenarios such as standoff, probing, attack, entanglement, and risk avoidance, with a recognition response time of ≤20ms;

[0042] 2. Situational Assessment Layer: Threat level is quantified in real time through a five-dimensional scoring model. When the opponent rushes forward quickly and the speed of limb swings exceeds the standard, it is judged as a high-threat offensive situation; when the opponent sways slightly and does not make obvious offensive movements, it is judged as a probing situation; when the distance between the enemy and ourselves is less than the safety threshold and our own posture is unbalanced, it is judged as a dangerous risk avoidance situation.

[0043] 3. Tactical Decision-Making Level: In response to a high-threat offensive situation, initiate passive defensive tactics, prioritizing blocking and dodging actions; in response to a probing situation, initiate active probing tactics, using small feints to exploit the opponent's weaknesses; in response to one's own advantageous situation, initiate strong offensive combo tactics to continuously suppress the opponent.

[0044] 4. Move Execution Layer: Matches corresponding martial arts moves according to tactical instructions, accurately outputs servo motor angle, movement speed, and force parameters to ensure standard moves and smooth transitions;

[0045] 5. Dynamic Correction Layer: Data updates and parameter corrections are completed every 10ms. It can quickly switch offensive and defensive tactics in response to the opponent's temporary changes in moves, avoiding lag in actions and disconnect in moves.

[0046] In actual use, after the robot initiates its self-check, it autonomously completes the entire process of perception, analysis, decision-making, and offense / defense without human remote control intervention. In high-intensity dynamic combat scenarios, it can effectively deal with complex situations such as opponent feints, rapid changes in moves, and continuous attacks, autonomously completing probing maneuvers, defensive counterattacks, and tactical suppression. Its anthropomorphic combat maneuvering effect is significantly better than that of traditional combat robots.

[0047] 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.

Claims

1. A biomimetic fighting robot, comprising a biomimetic mechanical body, a multimodal perception module, a data fusion preprocessing module, a martial arts game-themed hierarchical autonomous decision-making module, a motion execution control module, and a power management module, characterized in that: The bionic mechanical body is a multi-degree-of-freedom bionic fighting structure that can perform martial arts fighting actions such as blocking, striking, dodging, rushing forward, retreating, and side-stepping attack and defense. The multimodal perception module includes a visual perception unit, a distance perception unit, a force perception unit, a posture perception unit, and a speed perception unit, which are used to synchronously collect multi-dimensional real-time data on the opponent's status, the distance between the opponent and ourselves, the intensity of attack and defense, our own posture, and the speed of movement in the fighting scene. The data fusion preprocessing module is used to receive multimodal raw data, complete timestamp synchronization, noise filtering, data normalization and feature weighted fusion, and generate a standardized fighting scene feature matrix. The martial arts game-playing layered autonomous decision-making module is the core intelligent decision-making unit. It has a built-in five-layered decision-making architecture, which completes autonomous game-playing decisions through scene perception, situation analysis, tactical decision-making, move execution, and dynamic correction in sequence, and outputs attack and defense execution commands. The motion execution control module receives attack and defense commands, drives the bionic mechanical body to complete the corresponding fighting actions, and provides real-time feedback on the action execution status. The power management module provides stable power to each module and has overload protection.

2. The bionic fighting robot according to claim 1, characterized in that: The martial arts game-playing layered autonomous decision-making module includes a five-layer architecture from top to bottom, namely the scene perception layer, situation analysis layer, tactical decision-making layer, move execution layer, and dynamic correction layer. The scene perception layer is used to identify the types of fighting scenes such as confrontation, probing, attack, entanglement, and evasion, and to complete the state labeling; The situation assessment layer quantifies the opponent's threat level, the enemy's and our own security coefficient, and our own stability coefficient through a multi-dimensional scoring model to determine the current game's tone. The tactical decision-making layer corresponds to different game themes, matches preset martial arts offensive and defensive tactics, and outputs tactical instructions for probing, defense, counterattack, strong attack, and risk avoidance. The move execution layer breaks down tactical instructions into specific martial arts fighting action combinations, generating joint motion trajectories, servo parameters, and action speed parameters. The dynamic correction layer dynamically corrects tactical and move parameters based on real-time perception data and motion feedback data, achieving closed-loop adaptive adjustment.

3. The bionic fighting robot according to claim 1, characterized in that: The multimodal perception module adopts a 100Hz millisecond-level synchronous sampling mechanism, and the data from each sensor are calibrated with a unified timestamp. The data fusion preprocessing module adopts an improved Kalman filter algorithm and a feature weighted fusion algorithm to enhance the weight of core data in high-threat scenarios.

4. The biomimetic fighting robot according to claim 1, characterized in that: The martial arts game-playing layered autonomous decision-making module has a built-in martial arts move library, which includes basic offensive and defensive moves and combos. It supports iterative updates and can adapt to game-playing needs with different levels of confrontation.

5. The method of using a bionic fighting robot according to claim 1, characterized in that: Includes the following steps: S1. Equipment initialization: Start the robot and complete sensor self-test, algorithm initialization, model loading and servo motor calibration. S2, Multimodal Real-time Perception: The multimodal perception module synchronously collects multi-dimensional raw data of the combat scene; S3. Data fusion preprocessing: The raw data is denoised, synchronized, and normalized, and then fused to generate a fighting scene feature matrix; S4. Layered Autonomous Game Decision-Making: Through a five-layer martial arts game decision-making framework, the system sequentially completes scene recognition, situation assessment, tactical matching, move decomposition, and parameter correction, and outputs attack and defense execution commands. S5. Combat Action Execution: Drives the bionic mechanical body to complete corresponding martial arts offensive and defensive actions, achieving autonomous combat; S6. Dynamic closed-loop correction: Iteratively perceives, integrates, decides, and executes processes, adapting to changes in the adversarial situation in real time; S7. End Standby: After detecting the termination signal, the robot resets its posture and enters a low-power standby state.

6. The method of using a bionic fighting robot according to claim 5, characterized in that: In step S4, the situation analysis layer quantifies and generates six game themes: safe standoff, passive defense, active probing, precise counterattack, all-out offensive, and emergency avoidance, and matches them with differentiated offensive and defensive tactics.

7. The method of using a biomimetic fighting robot according to claim 5, characterized in that: In step S6, the decision module completes parameter updates and action corrections every 10ms to achieve rapid adaptive response to opponent changes and environmental interference.

Images