A multi-stage self-adaptive interactive cavity adjusting system for AI bionic robots

Abstract

The application provides a kind of multi-stage self-adapting interactive cavity adjusting system for AI bionic robot, relating to the technical field of AI bionic robot, which includes bionic cavity module, multi-modal driving module, sensing and detecting module, central control module and user preference storage module, the user preference storage module is electrically connected with the central control module, used for storing the interactive parameter preference of different users, supporting multi-user switching: the bionic cavity module is made of medical grade silicone material, internally provided with deformable elastic support structure, used for simulating the softness and elasticity of human tissue. Through the interactive stage recognition algorithm, the five different stages of the interactive process can be accurately identified and the suction force, vibration frequency and temperature parameters are automatically adjusted, simulating the physiological response process of real human body. This dynamic self-adaptive adjustment significantly improves the realism and immersion of physical interaction.

Description

Technical Field

[0001] This invention relates to the field of AI bionic robot technology, specifically to a multi-stage adaptive interactive cavity adjustment system for AI bionic robots. Background Technology

[0002] With the development of bionic robot technology, the physical interaction experience of bionic robot puppets is constantly improving. In existing technologies, the physical interaction systems of bionic robot puppets mostly adopt a fixed-parameter driving method, meaning that parameters such as suction force, vibration frequency, and temperature are all preset values, and cannot be dynamically adjusted according to the actual situation during the interaction process. This fixed-parameter driving method has the following obvious drawbacks: 1. The interactive experience is monotonous and unnatural: It is unable to simulate the physiological changes of the real human body at different stages of interaction, resulting in a rigid and unrealistic physical interaction experience; 2. Unable to adapt to the needs of different users: Different users have significant differences in their preferences for intensity, frequency, and temperature, and fixed parameters cannot meet personalized needs; 3. Lack of interactive feedback mechanism: The inability to adjust interaction parameters based on real-time user feedback can easily lead to a poor or even uncomfortable user experience; 4. Inconsistent interaction process: The transitions between different interaction stages are abrupt and lack smooth transitions, affecting the overall immersive experience. Summary of the Invention

[0003] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a multi-stage adaptive interactive cavity adjustment system for AI bionic robot puppets, which solves the problem that the physical interaction system parameters of existing bionic robot puppets are fixed and cannot be adaptively adjusted according to the interaction stage.

[0004] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a multi-stage adaptive interactive cavity adjustment system for AI bionic robots, comprising a bionic cavity module, a multimodal driving module, a sensing and detection module, a central control module, and a user preference storage module. The user preference storage module is electrically connected to the central control module and is used to store the interaction parameter preferences of different users, supporting multi-user switching. The biomimetic cavity module is made of medical-grade silicone material and has a deformable elastic support structure inside to simulate the softness and elasticity of human tissue. The multimodal driving module includes a suction driving unit, a vibration driving unit, and a temperature regulation unit, which are used to provide three interactive modes: suction force, vibration feedback, and temperature regulation, respectively. The sensing and detection module includes a pressure sensor array, a displacement sensor, and a contact sensor, which are used to collect pressure distribution, displacement deformation, and contact state data in real time during the interaction process. The central control module has a built-in interaction stage recognition algorithm and a multi-parameter adaptive control model. It is electrically connected to the multimodal driving module and the sensing and detection module, respectively. It is used to identify the current interaction stage based on the data collected by the sensing and detection module and output the corresponding control parameters to the multimodal driving module. During the interaction, the sensing module collects pressure distribution, displacement deformation, and contact state data in real time and transmits them to the central control module. The central control module identifies the current interaction stage based on the changing characteristics of the sensing data. The multi-parameter adaptive control model outputs corresponding control parameters for suction force, vibration frequency, and temperature based on the identified interaction stage and user preferences. The multi-modal drive module operates according to the control parameters, achieving dynamic adjustment synchronized with the interaction stage. The system monitors user feedback signals in real time. If abnormal pressure is detected or a stop command is issued by the user, the drive parameters are immediately reduced or the system stops operating.

[0005] Preferably, during system initialization, the central control module reads parameters from the user preference storage module and sets the initial interaction parameters; After the interaction ends: The system automatically records the parameter settings and user feedback for this interaction and updates the user preference model.

[0006] Preferably, the suction clamp driving unit consists of multiple micro servo motors and a flexible transmission mechanism, which can realize independent contraction and relaxation of different parts of the cavity, and the suction clamp force adjustment range is 0.5N-15N.

[0007] Preferably, the vibration drive unit consists of multiple miniature vibration motors distributed at different positions in the cavity, with a vibration frequency adjustment range of 10Hz-200Hz, enabling multiple vibration modes.

[0008] Preferably, the temperature regulation unit consists of a flexible heating film and a temperature sensor, with a temperature regulation range of 32℃-42℃ and a temperature control accuracy of ±0.5℃.

[0009] Preferably, the pressure sensor array consists of 16-64 thin-film pressure sensors arranged in a matrix on the inner wall surface of the cavity.

[0010] Preferably, the interaction stage identification algorithm adopts a temporal classification method based on a hidden Markov model or a long short-term memory network. The input features include the mean, variance, and rate of change of pressure, the amount of displacement deformation and its rate of change, the duration of the contact state, the spatial features of the pressure distribution and their temporal combination, and adopts a sliding window mechanism with a window length of 1-5 seconds and a step size of 0.1-0.5 seconds to achieve real-time stage identification with an identification delay of ≤1 second.

[0011] Preferably, the multi-parameter adaptive control model adopts fuzzy PID control or model predictive control algorithm, and includes four control strategies: stage reference parameter mapping, user preference correction, real-time feedback fine-tuning and gradual transition control. The gradual transition control adopts an S-curve or exponential curve gradual transition method, and the transition time is 2-10 seconds.

[0012] (III) Beneficial Effects This invention provides a multi-stage adaptive interactive cavity adjustment system for AI bionic robots. It offers the following advantages: 1. Multi-stage adaptive adjustment: It can accurately identify different stages of the interaction process and automatically adjust the suction force, vibration frequency and temperature, simulating the physiological reaction process of the real human body, significantly improving the realism and immersion of physical interaction.

[0013] 2. Multimodal collaborative control: It realizes the collaborative control of three interaction modes: suction clamp, vibration and temperature, and provides a richer and more three-dimensional physical interaction experience.

[0014] 3. Strong personalization and adaptation capabilities: It supports the storage of multiple user preferences and custom parameter settings, which can meet the personalized needs of different users.

[0015] 4. Safe and reliable: Equipped with a complete sensor detection and anomaly protection mechanism, it can monitor the interaction status in real time, avoid user discomfort due to excessive parameters, and ensure the safety of the interaction process.

[0016] 5. Smooth transition: The parameter changes between different interaction stages are gradual, avoiding abrupt transitions and making the interaction process more natural and smooth. Detailed Implementation

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] A multi-stage adaptive interactive cavity adjustment system for AI bionic robots includes a bionic cavity module, a multimodal driving module, a sensing and detection module, a central control module, and a user preference storage module. The user preference storage module is electrically connected to the central control module and is used to store the interaction parameter preferences of different users, supporting multi-user switching. The user preference storage module uses non-volatile memory. The biomimetic cavity module is made of medical-grade silicone material and has a deformable elastic support structure inside to simulate the softness and elasticity of human tissue. The specific medical-grade silicone material meets the ISO 10993 biocompatibility standard, with a hardness range of Shore A10-30, tensile strength ≥5MPa, and elongation at break ≥400%, ensuring the safety, softness, and durability of the material. The elastic support structure has a honeycomb microporous structure inside, with a micropore diameter of 0.5-2mm and a porosity of 30%-60%, which enhances the cavity's deformability and pressure uniformity, while also providing a certain degree of air permeability. The multimodal drive module includes a suction clamping drive unit, a vibration drive unit, and a temperature regulation unit, which are used to provide three interactive modes: suction force, vibration feedback, and temperature regulation, respectively. The suction clamping drive unit consists of multiple micro servo motors and a flexible transmission mechanism, enabling independent contraction and relaxation of different parts of the cavity. The suction force adjustment range is 0.5N-15N. Specifically, there are 4-12 micro servo motors evenly distributed along the circumference of the cavity. Each motor is connected to the inner wall of the cavity through a flexible transmission mechanism (such as a silicone bellows, flexible connecting rod, or shape memory alloy wire). Each micro servo motor can be independently controlled to achieve differentiated suction clamping in different parts of the cavity, simulating the non-uniform contraction characteristics of real human muscles. The vibration drive unit consists of multiple micro vibration motors distributed at different positions in the cavity. The vibration frequency adjustment range is 10Hz-200Hz, enabling multiple vibration modes. Specifically, there are 6-16 micro vibration motors, divided into deep vibration groups and surface vibration groups. Three to eight deep vibration groups are installed in the middle layer support structure of the cavity, providing low-frequency deep vibration (10Hz-80Hz); three to eight surface vibration groups are installed near the inner wall of the cavity, providing high-frequency surface vibration (50Hz-200Hz). Each vibration motor can independently control the frequency, amplitude, and phase to achieve multiple vibration modes, including but not limited to: same-frequency and same-phase vibration mode, same-frequency and opposite-phase vibration mode, frequency-sweeping vibration mode, multi-point alternating vibration mode, and random noise vibration mode. The vibration drive unit supports coordinated control of vibration mode and suction clamp action, such as synchronously increasing vibration intensity when the suction clamp tightens to simulate the stress response of a real human body. The temperature regulation unit consists of a flexible heating film and a temperature sensor, with a temperature regulation range of 32℃-42℃ and a temperature control accuracy of ±0.5℃. The flexible heating film is a resistance heating film with a polyimide (PI) substrate, a thickness of 0.1-0.3mm, and a power density of 0.5-2W / cm². 2 The temperature sensors are NTC thermistors or PT100 platinum resistance thermometers, numbered 4-8, and distributed in different positions in the cavity to monitor the temperature distribution in different areas of the cavity in real time. During the temperature control process, natural heat dissipation or micro fan-assisted heat dissipation is used for cooling. The sensing and detection module includes a pressure sensor array, displacement sensors, and contact sensors. It is used to collect pressure distribution, displacement deformation, and contact state data in real time during the interaction process. The pressure sensor array consists of 16-64 thin-film pressure sensors arranged in a matrix on the inner wall surface of the cavity. These thin-film pressure sensors employ capacitive or piezoresistive principles, exhibiting good flexibility and conformability. The pressure distribution is used to analyze pressure concentration areas and pressure change trends during the interaction process. The displacement sensors detect the deformation of the cavity, including changes in cavity diameter and axial compression. These sensors employ flexible strain gauges or inductive displacement sensors, numbering 4-8, arranged along the circumference and axial direction of the cavity. The contact sensors detect the contact state between the user and the cavity, including contact position, contact area, and contact duration. These contact sensors are distributed at the cavity entrance and key locations on the inner wall, and can identify the user's contact patterns (such as point contact, surface contact, sliding contact, etc.), providing auxiliary judgment for the interaction phase. The central control module has a built-in interaction stage recognition algorithm and a multi-parameter adaptive control model. It is electrically connected to the multimodal drive module and the sensing and detection module, respectively. It is used to identify the current interaction stage based on the data collected by the sensing and detection module and output the corresponding control parameters to the multimodal drive module.

[0019] During system initialization, the central control module reads parameters from the user preference storage module and sets initial interaction parameters. During interaction, the sensing module collects pressure distribution, displacement deformation, and contact state data in real time and transmits them to the central control module. The central control module identifies the current interaction stage based on the changing characteristics of the sensing data. The multi-parameter adaptive control model outputs corresponding control parameters for suction force, vibration frequency, and temperature based on the identified interaction stage and user preferences. The multi-modal drive module operates according to the control parameters, achieving dynamic adjustment synchronized with the interaction stage. The system monitors user feedback signals in real time. If abnormal pressure is detected or a stop command is issued by the user, the drive parameters are immediately reduced or operation is stopped. After the interaction ends, the system automatically records the parameter settings and user feedback for this interaction and updates the user preference model.

[0020] The specific interaction process is as follows: Initial contact phase: The contact sensor detects a contact signal, and the pressure sensor detects a small pressure (<1 N / cm). 2 The displacement sensor shows minute deformation (<5%). This stage is characterized by contact establishment, and interaction has just begun.

[0021] Gradual interaction stage: Pressure gradually increases (1-5 N / cm) 2 Displacement and deformation increase (5%-20%), and the contact area expands. This stage is characterized by a gradual increase in interaction intensity.

[0022] Stable interaction phase: Pressure fluctuates within a relatively stable range (3-8 N / cm). 2 The displacement and deformation tend to stabilize (15%-25%), and the contact state continues. This stage is characterized by the interaction entering a stable period.

[0023] Peak interaction phase: Pressure reaches a relatively high level (6-15 N / cm) 2 During this stage, displacement deformation reaches a relatively large value (20%-30%), and the frequency of pressure fluctuations increases. The characteristic of this stage is that the interaction intensity reaches its peak.

[0024] Recovery phase: Pressure gradually decreases (<5 N / cm) 2 The displacement deformation decreases (<15%), and the contact area shrinks. This stage is characterized by weakened interaction intensity, marking the beginning of the final stage.

[0025] In the above, the interaction stage identification algorithm adopts a temporal classification method based on Hidden Markov Model or Long Short-Term Memory Network. The input features include pressure mean, variance, rate of change, displacement deformation and its rate of change, contact state duration, spatial features of pressure distribution and their temporal combination. A sliding window mechanism is adopted, with a window length of 1-5 seconds and a step size of 0.1-0.5 seconds, to achieve real-time stage identification with an identification delay of ≤1 second. The multi-parameter adaptive control model outputs corresponding parameters for suction force, vibration frequency, and temperature control based on the identified interaction stage and user preferences. The multi-parameter adaptive control model employs fuzzy PID control or model predictive control algorithms, and includes the following control strategies: Stage reference parameter mapping: Each interaction stage corresponds to a set of reference parameters (clamping force, vibration frequency, temperature), forming a stage-parameter mapping table.

[0026] User preference correction: Based on the data in the user preference storage module, the baseline parameters are personalized and the correction coefficient ranges from 0.5 to 1.5.

[0027] Real-time feedback fine-tuning: The output parameters are dynamically fine-tuned based on the real-time data from the sensing and detection module, with a fine-tuning range of ±20%.

[0028] Gradual transition control: When switching between interactive phases, parameter changes are performed using an S-curve or exponential curve gradual transition with a transition time of 2-10 seconds to avoid sudden parameter changes.

[0029] Implementation Plan 1: Standard Multi-Stage Adjustment Mode.

[0030] Initial contact stage: temperature 36℃, suction force 2N, vibration frequency 20Hz, gentle vibration; Gradual interaction phase: The temperature rises to 37℃, the clamping force gradually increases to 5N, and the vibration frequency increases to 50Hz; Stable interaction phase: The temperature is maintained at 37.5℃, the clamping force changes periodically between 5N and 8N, and the vibration frequency is 80Hz-120Hz. Peak interaction phase: The temperature rises to 38℃, the clamping force reaches its maximum value of 12N, the vibration frequency is 150Hz, and the vibration is high-intensity. Recovery phase: The temperature gradually drops to 36℃, and the suction force and vibration frequency gradually decrease until they stop.

[0031] Implementation Plan 2: Gentle Mode. For users who are more sensitive to force and vibration, the suction force is reduced by 30% and the vibration frequency is reduced by 20% in all stages, and the temperature adjustment range is narrowed to 35℃-37℃.

[0032] Implementation Plan 3: Personalized Customization Mode. Users can customize the suction force, vibration frequency, temperature, and duration of each interaction stage through the accompanying app, and the settings can be recalled with a single click after being saved by the system.

[0033] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-stage adaptive interactive cavity adjustment system for AI bionic robots, characterized in that: It includes a bionic cavity module, a multimodal driving module, a sensing and detection module, a central control module, and a user preference storage module. The user preference storage module is electrically connected to the central control module and is used to store the interaction parameter preferences of different users, supporting multi-user switching. The biomimetic cavity module is made of medical-grade silicone material and has a deformable elastic support structure inside to simulate the softness and elasticity of human tissue. The multimodal driving module includes a suction driving unit, a vibration driving unit, and a temperature regulation unit, which are used to provide three interactive modes: suction force, vibration feedback, and temperature regulation, respectively. The sensing and detection module includes a pressure sensor array, a displacement sensor, and a contact sensor, which are used to collect pressure distribution, displacement deformation, and contact state data in real time during the interaction process. The central control module has a built-in interaction stage recognition algorithm and a multi-parameter adaptive control model. It is electrically connected to the multimodal driving module and the sensing and detection module, respectively. It is used to identify the current interaction stage based on the data collected by the sensing and detection module and output the corresponding control parameters to the multimodal driving module. During the interaction, the sensing module collects pressure distribution, displacement deformation, and contact state data in real time and transmits them to the central control module. The central control module identifies the current interaction stage based on the changing characteristics of the sensing data. The multi-parameter adaptive control model outputs corresponding control parameters for suction force, vibration frequency, and temperature based on the identified interaction stage and user preferences. The multi-modal drive module operates according to the control parameters, achieving dynamic adjustment synchronized with the interaction stage. The system monitors user feedback signals in real time. If abnormal pressure is detected or a stop command is issued by the user, the drive parameters are immediately reduced or the system stops operating.

2. The AI ​​bionic robot occasional multi-stage adaptive interactive cavity adjustment system according to claim 1, characterized in that: During system initialization, the central control module reads parameters from the user preference storage module and sets the initial interaction parameters; After the interaction ends: The system automatically records the parameter settings and user feedback for this interaction and updates the user preference model.

3. The AI ​​bionic robot occasional multi-stage adaptive interactive cavity adjustment system according to claim 2, characterized in that: The suction clamp drive unit consists of multiple micro servo motors and a flexible transmission mechanism, which can realize independent contraction and relaxation of different parts of the cavity, and the suction clamp force adjustment range is 0.5N-15N.

4. The AI ​​bionic robot occasional multi-stage adaptive interactive cavity adjustment system according to claim 3, characterized in that: The vibration drive unit consists of multiple miniature vibration motors distributed at different positions in the cavity, with a vibration frequency adjustment range of 10Hz-200Hz, enabling various vibration modes.

5. The AI ​​bionic robot occasional multi-stage adaptive interactive cavity adjustment system according to claim 4, characterized in that: The temperature regulation unit consists of a flexible heating film and a temperature sensor, with a temperature regulation range of 32℃-42℃ and a temperature control accuracy of ±0.5℃.

6. The AI ​​bionic robot occasional multi-stage adaptive interactive cavity adjustment system according to claim 5, characterized in that: The pressure sensor array consists of 16-64 thin-film pressure sensors arranged in a matrix on the inner wall surface of the cavity.

7. The AI ​​bionic robot occasional multi-stage adaptive interactive cavity adjustment system according to claim 6, characterized in that: The interaction phase identification algorithm adopts a temporal classification method based on Hidden Markov Model or Long Short-Term Memory Network. The input features include pressure mean, variance, rate of change, displacement deformation and its rate of change, contact state duration, spatial features of pressure distribution and their temporal combination. A sliding window mechanism is adopted with a window length of 1-5 seconds and a step size of 0.1-0.5 seconds to achieve real-time phase identification with an identification delay of ≤1 second.

8. The AI ​​bionic robot occasional multi-stage adaptive interactive cavity adjustment system according to claim 7, characterized in that: The multi-parameter adaptive control model adopts fuzzy PID control or model predictive control algorithm, and includes four control strategies: stage reference parameter mapping, user preference correction, real-time feedback fine-tuning and gradual transition control. The gradual transition control adopts S-curve or exponential curve gradual transition method, and the transition time is 2-10 seconds.