A robot reflection suppression and tunnel control system based on biological quasi-state

Abstract

This invention provides a biomimetic-based robot reflection suppression and tunneling control system, comprising an interaction module, a reflection module, and a survival module. The interaction module receives and processes user intentions, executes behavioral decisions, and does not participate in maintaining motor torque. The reflection module fuses inertial sensor data to generate an anti-interference balance torque signal to maintain robot posture stability. The survival module monitors the energy core status and has the highest execution authority. The technical effects of this invention are: preventing the algorithm from misjudging robot instability and thus outputting a corrective torque, leading to motor overheating or even burnout; preventing complex action commands (such as "do a set of push-ups") from being interrupted by concurrent heartbeat packets or sensor noise, resulting in distorted movements; and preventing main program deadlock, which would cause battery protection to fail and lead to physical safety hazards.

Description

Technical Field

[0001] This invention relates to a robot control system, specifically to a robot reflection suppression and tunneling control system based on biomimicry. Background Technology

[0002] As robots become increasingly complex, the conflict between their "cognitive system" (responsible for interaction, performance, and decision-making) and "cerebellum system" (responsible for basic balance and fall prevention) will become a core bottleneck. Existing problems include: 1. Control split: When a robot attempts to perform artistic, non-standard movements (such as the violent shaking during "dancing"), the bottom module's balance algorithm may misinterpret this as instability, resulting in a reverse output of corrective torque. This internal conflict can lead to motor overheating or even burnout. 2. Fragmented instructions: Under network latency or multi-terminal control, a series of continuous complex action instructions (such as "do a set of push-ups") are easily interrupted by concurrent heartbeat packets or sensor noise, leading to distorted movements. 3. Lack of survival mechanisms: Traditional control logic typically treats "low-voltage protection" as a normal task. Once the main program deadlocks, the battery protection function immediately fails, posing a physical safety hazard. Summary of the Invention

[0003] To address the aforementioned problems, this invention provides a robot reflection suppression and tunneling control system based on biomimicry, comprising an interaction module, a reflection module, and a survival module;

[0004] The interaction module receives and processes user intent information, executes behavioral decisions, and does not participate in maintaining motor torque.

[0005] The reflection module integrates inertial sensor data to generate an anti-interference balance torque signal, maintaining the stability of the body's attitude;

[0006] The survival module monitors the status of the energy core and has the highest execution authority.

[0007] Furthermore, the system operation includes a steady-state survival mode and an atomic behavior release mode;

[0008] The steady-state survival mode is the default state, where the reflection module completely occupies the data path to the limbs, and the feedback of sensor data directly acts on the actuator to form a closed-loop control. The motion intention generated by the interaction module is only injected into the reflection module as a reference target value, and the reflection module converts it into a movement that conforms to physical balance.

[0009] When an interaction module initiates an indivisible complex action, the system enters atomic action release mode:

[0010] Furthermore, the atomic behavior release mode performs the following steps:

[0011] S1 Tunnel Establishment: The interaction module projects a suppression signal to the system, establishing a direct control tunnel between the command source and the driver;

[0012] S2 Reflection Suppression: When the reflection module receives the suppression signal, it immediately suspends its own balance maintenance function and cuts off the closed loop between the sensor feedback and the motor output. At this time, it completely obeys the preset motion trajectory.

[0013] S3 Open-Loop Execution: Complex action sequences are transmitted to the drive unit without loss through a pre-established tunnel;

[0014] S4 Functional Recovery: As the action sequence ends, the inhibition signal fades, the tunnel automatically disintegrates, the reflection module takes over the data path, and the body's autonomous balance is restored.

[0015] Furthermore, the survival module continuously monitors the health status of the energy system. Once it detects an energy depletion signal, the survival module will ignore any current behavior pattern and directly trigger a physical-level circuit blockage, forcibly releasing all joint torques and putting the body into a protective state.

[0016] Furthermore, when the survival module detects a critical state, it forcibly shuts down all motion outputs.

[0017] The technical advantages of this invention are: preventing the algorithm from misjudging machine instability and thus outputting a corrective torque, which could lead to motor overheating or even burnout; preventing complex action commands (such as "do a set of push-ups") from being interrupted by concurrent heartbeat packets or sensor noise, resulting in distorted movements; and preventing main program deadlock, which could cause battery protection to fail and lead to physical safety hazards. This architecture has extremely low implementation cost and high feasibility on a microcontroller: the formulas do not involve complex matrix inversion operations, and arbitration can be completed in less than 1μs on a 72MHz MCU. Only additional maintenance is required. and Two state variables, with negligible RAM usage. The survival module directly embeds the highest priority timer interrupt, with response latency controllable to the microsecond level, far faster than the battery voltage drop rate (millisecond level), ensuring the protection mechanism is absolutely effective. Attached Figure Description

[0018] Figure 1 This is a flowchart of the atomic behavior release mode of the present invention. Detailed Implementation

[0019] The inventive point of this invention is the construction of a three-tiered biomimetic architecture consisting of an "interaction module - reflex module - survival module," with the core being the introduction of a "Neural Inhibition Tunnel" mechanism: The Reflex Layer simulates the biological spinal cord, defaulting to high-frequency closed-loop balance control. The Inhibition Tunnel actively sends inhibition signals when the cognitive module initiates "atomic behavior," temporarily severing the closed-loop circuit of the Reflex Layer and establishing an open-loop tunnel directly to the driver. The Survival Layer simulates the biological brainstem, operating independently of consciousness and possessing absolute control over the hardware.

[0020] The system consists of three parallel but distinct physiological units: The Cognitive Interaction Module: Acting as the system's "cerebral cortex module," it receives and processes complex, high-latency, non-real-time user intentions (such as voice interaction and long-term action planning). This module focuses on behavioral decision-making and does not directly participate in millisecond-level motor torque maintenance. The Spinal Reflex Module: Acting as the system's "cerebellum and spinal cord," it operates continuously at a high-frequency rhythm. In its natural state, this module continuously integrates inertial sensor data to generate anti-interference balancing torques, maintaining bodily posture stability. The Brainstem Survival Module: Acting as the system's "life center," it independently monitors the energy core state and heart rate rhythm. This module is at the lowest level of hardware control and has inalienable execution authority.

[0021] This invention manages control through a dynamic "channel switching" and "function suppression" mechanism:

[0022] In steady-state survival mode (default state), during normal operation, the reflex module completely occupies the neural pathways to the limbs. Feedback from sensor data directly affects the actuators, forming a closed-loop control. The motor intention generated by the cognitive module is only injected into the reflex module as a reference target value, which is then converted into a movement that conforms to physical equilibrium.

[0023] Atomic Behavior Release Mode (Tunnel Activation): When the cognitive module decides to initiate an indivisible complex behavior (such as a stunt), the system enters tunnel control state:

[0024] S1 Tunnel Establishment: The cognitive module projects a suppression signal to the system's bottom module, constructing a direct control tunnel between the instruction source and the driver.

[0025] S2 Reflection Suppression: Upon sensing the presence of the suppression signal, the reflection module suspends its own balance maintenance function, cutting off the closed-loop circuit between sensor feedback and motor output. At this point, the robot temporarily loses its "pain sensation" and "balance sense," completely obeying the preset motion trajectory.

[0026] S3 Open-Loop Execution: Complex motion sequences are transmitted to the drive unit without loss through a pre-established tunnel, accurately reproducing the preset artistic trajectory, unaffected by body tilt or vibration.

[0027] S4 functional recovery: As the action sequence ends, the inhibitory signal fades, and the tunnel automatically disintegrates. The reflex module takes over the neural pathway again, instantly restoring the body's autonomous balance.

[0028] The system of this invention has an abnormal circuit breaker mechanism. The survival module operates independently of the above logic and continuously monitors the health status of the energy system. Once it detects signs of energy depletion or loss of vital signs (heartbeat), the survival module will ignore any current behavior pattern (whether balancing or dancing) and directly trigger a physical-level neural blockade, forcibly releasing all joint torques and putting the body into a flaccid protective state, thereby preventing battery over-discharge or motor burnout.

[0029] The mathematical model for the control switching of this invention is presented below, along with a feasibility analysis.

[0030] The following explains the arbitration formula for system control in this invention.

[0031] System output torque This can be described by the following arbitration formula:

[0032]

[0033] in:

[0034] : Represents the closed-loop equilibrium torque (unit: N·m) generated by the spinal reflex module. Its calculation is based on attitude feedback.

[0035]

[0036] ( For PD control parameters, (for joint angles)

[0037] : Expected joint angle, i.e., the target attitude angle issued by the cognitive module (unit: rad).

[0038] : Current joint angle, i.e., the actual attitude angle fed back by the sensor in real time (unit: rad).

[0039] : Desired joint angular velocity, i.e., the derivative of the target attitude angle (unit: rad / s).

[0040] : Current joint angular velocity, i.e., the derivative of the actual attitude angle (unit: rad / s).

[0041] : Represents the open-loop stunt motion torque (preset trajectory) generated by the cognitive interaction module.

[0042] : Indicates a neural inhibitory signal (i.e., the "inhibition tunnel" activation state).

[0043] when (Steady-state mode): System output To maintain balance.

[0044] when (Atomic Behavior Pattern): System Output , indicating the performance of a special skill.

[0045] : This indicates the hardware enable signal for the survival module.

[0046] Under normal conditions .

[0047] When the battery voltage hour, The torque is forced to 0, at which point the output torque is reduced. (Motor unloading), regardless of the output of other module levels.

[0048] Real-time battery voltage (unit: V).

[0049] : Preset low-voltage protection threshold voltage for the battery (unit: V).

[0050] The stability of the system state switching of the present invention will be described below.

[0051] The core challenge facing this system is the instantaneous state transition. The sudden change in torque that may occur when changing from 0 to 1 or from 1 to 0.

[0052] Feasibility Derivation:

[0053] Assuming at the switching time This suppresses signal jumps.

[0054] Starting from steady state to enter the tunnel ( ):

[0055] At this point, the robot is usually in a state of controlled equilibrium. (Maintaining torque).

[0056] The action generator of the cognitive module (such as S-curve planning) is designed to start from the current pose, i.e. .

[0057] Therefore, torque continuity is guaranteed, and mechanical impact will not occur.

[0058] Switching back to steady state from the tunnel ( ):

[0059] When the atomic behavior ends, the robot may be in a non-equilibrium posture.

[0060] At this point, the reflection module takes over again. Instantly turned .

[0061] because It is a closed-loop feedback based on PD control, if the error A larger value will result in a larger restoring torque.

[0062] Optimization strategy: In In the first few periods after resetting to zero, we can introduce a smoothing factor. ,make ,in The transition from 0 to 1 achieves a soft landing. This is also a key guarantee for the feasibility of this invention.

[0063] The following is a feasibility analysis of the hardware resources for this invention.

[0064] This architecture has extremely low implementation cost and high feasibility on a microcontroller:

[0065] Computational overhead: The formula only involves simple weighted summation and logical judgment, without complex matrix inversion operations, and can complete an arbitration in less than 1μs on a 72MHz MCU.

[0066] Storage overhead: Only requires additional maintenance and Two state variables, RAM usage is negligible.

[0067] Real-time performance: The survival module directly embeds the highest priority timer interrupt, and the response delay can be controlled at the microsecond level, which is much faster than the battery voltage drop speed (millisecond level), ensuring that the protection mechanism is absolutely effective.

Claims

1. A robot reflection suppression and tunneling control system based on biomimicry, comprising an interaction module, a reflection module, and a survival module; The interaction module receives and processes user intent information, executes behavioral decisions, and does not participate in maintaining motor torque. The reflection module integrates inertial sensor data to generate an anti-interference balance torque signal, maintaining the stability of the body's attitude; The survival module monitors the status of the energy core and has the highest execution authority.

2. The biomimetic-based robot reflection suppression and tunneling control system as described in claim 1, The system operation includes a steady-state survival mode and an atomic behavior release mode; The steady-state survival mode is the default state, where the reflection module completely occupies the data path to the limbs, and the feedback of sensor data directly acts on the actuator to form a closed-loop control. The motion intention generated by the interaction module is only injected into the reflection module as a reference target value, and the reflection module converts it into a movement that conforms to physical balance. When an interactive module initiates an indivisible complex action, the system enters atomic action release mode.

3. A robot reflection suppression and tunneling control system based on biomimicry as described in claim 2, The atomic behavior release mode performs the following steps: S1 Tunnel Establishment: The interaction module projects a suppression signal to the system, establishing a direct control tunnel between the command source and the driver; S2 Reflection Suppression: When the reflection module receives the suppression signal, it immediately suspends its own balance maintenance function and cuts off the closed loop between the sensor feedback and the motor output. At this time, it completely obeys the preset motion trajectory. S3 Open-Loop Execution: Complex action sequences are transmitted to the drive unit without loss through a pre-established tunnel; S4 Functional Recovery: As the action sequence ends, the inhibition signal fades, the tunnel automatically disintegrates, the reflection module takes over the data path, and the body's autonomous balance is restored.

4. A robot reflection suppression and tunneling control system based on biomimicry as described in claim 1. The survival module continuously monitors the health status of the energy system. Once it detects an energy depletion signal, the survival module will ignore any current behavior pattern and directly trigger a physical-level circuit blockage, forcibly releasing all joint torques and putting the body into a protective state.

5. The biomimetic robot reflection suppression and tunneling control system as described in claim 1, wherein the survival module forcibly shuts off all motion outputs when it detects a critical state.

Images