A smart prosthetic hand and its control method

Abstract

The application relates to the technical field of intelligent artificial limb control, and discloses an intelligent artificial limb hand and a control method thereof, which comprises the following steps: starting a gripping task through an electromyographic signal, acquiring a forearm motion posture in real time by using an inertial measurement unit and determining a motion context, automatically selecting a gesture prototype to drive the artificial limb hand to preform based on the context, and triggering finger closure after detecting a target object by using a proximity sensor. Through intelligent mapping of the motion context and the gesture prototype, the application realizes physiological-level cooperation between a gripping intention and a limb motion, so that a user can complete natural and smooth gripping operation without deliberately planning actions. Meanwhile, dynamic adjustment of a motion speed feature and an electromyographic trend intensity enables the system to adapt to speed and intensity requirements in different scenes, and significantly reduces operation cognitive load.

Description

Technical Field

[0001] This invention relates to an intelligent prosthetic hand and its control method, belonging to the field of intelligent prosthetic control technology. Background Technology

[0002] In the field of intelligent prosthetic control, existing technologies generally adopt a linear control paradigm of intent-command-execution. This approach requires users to break down continuous grasping actions into independent sequences of operation commands: first, spatial positioning is achieved through electromyographic signals or visual assistance, then preset gestures are actively selected, and finally execution is triggered. Taking everyday grasping scenarios as an example, users need to complete three deliberate steps in sequence: arm movement, gesture selection, and closing operation, resulting in sluggish operation and high cognitive load.

[0003] Therefore, the existing technology has the following fundamental limitations: 1. The existing solution severs the natural coupling process between arm transportation and hand pre-forming, forcing users to make up for the system's defects with non-physiological step-by-step operations; 2. The fixed threshold triggering mechanism cannot adapt to different movement rhythms, such as the speed of industrial grasping and the gentleness of home operation, resulting in delayed response in high-speed scenarios and premature triggering in low-speed scenarios; 3. It relies on users to continuously output fine instructions and fails to explore the intention information contained in the limb movement itself, which increases the burden on neuromuscular control.

[0004] While industry attempts have focused on improving recognition accuracy through multi-sensor fusion, this has exacerbated system complexity and computational overhead, amplifying operational latency issues. Therefore, the technical problem this invention aims to solve is how to establish an instinctive coordination mechanism between grasping intention and limb movement to achieve dynamically adaptive, anthropomorphic operation with zero cognitive burden. Summary of the Invention

[0005] This invention provides an intelligent prosthetic hand and its control method, the main purpose of which is to solve the problems of excessive cognitive load, insufficient dynamic adaptability and redundant intention expression caused by the decomposition of actions in existing prosthetic control.

[0006] To achieve the above objectives, the present invention provides an intelligent prosthetic hand comprising: an inertial measurement unit (IMU) for real-time acquisition of motion posture information of the forearm where the prosthetic hand is located, the motion posture information including the motion velocity characteristics of the forearm during a grasping task, the motion velocity characteristics being determined by processing the dynamic data of the IMU; an electromyography (EMG) sensor for real-time acquisition of EMG signals of specific muscle groups in the forearm and identification of trend intensity changes in the EMG signals, the trend intensity changes reflecting the dynamic intensity of the user's grasping intention; and a control unit electrically connected to the IMU and the EMG sensor, the control unit being configured to: collaboratively determine the grasping timing and grasping speed of the intelligent prosthetic hand based on the motion velocity characteristics and the trend intensity changes in the EMG signals; wherein, when the motion velocity characteristics exhibit high... When the movement speed is high, the control unit determines an earlier grasping timing and a faster grasping speed. When the movement speed is low, the control unit determines a later grasping timing and a slower grasping speed. The trend intensity change of the electromyographic signal dynamically adjusts the grasping timing and grasping speed. When the trend intensity change shows an accelerating increase, the grasping timing is further advanced and the grasping speed is further increased. When the trend intensity change shows a decelerating decrease, the grasping timing is appropriately delayed and the grasping speed is appropriately slowed down. Furthermore, when a target object is detected, the control unit triggers the closing action of the fingers of the intelligent prosthetic hand according to the determined grasping timing and grasping speed. The speed of the finger closing action is matched with the grasping speed to achieve anthropomorphic collaborative grasping of the intelligent prosthetic hand.

[0007] Preferably, the inertial measurement unit is configured to: acquire initial motion data of the forearm at the beginning of the grasping task and establish a reference motion baseline of the forearm based on the initial motion data; during the execution of the grasping task, monitor the changes in the motion data of the inertial measurement unit relative to the reference motion baseline in real time, and use the changes as the basis for judging the motion speed characteristics. The basis for judging the motion speed characteristics includes the rate of change of the motion data within a specific time window, the instantaneous velocity amplitude, or the velocity integral.

[0008] Preferably, the electromyography (EMG) sensor is configured to: filter the EMG signal to eliminate noise, and extract the envelope of the filtered EMG signal to obtain the trend intensity of the EMG signal; perform real-time dynamic analysis on the trend intensity of the EMG signal to identify its rising, falling or stable trend, as well as the rate of change of the trend intensity, thereby obtaining the trend intensity change.

[0009] Preferably, the control unit is configured to: when the motion speed characteristics indicate that the forearm motion speed reaches a preset grasping preparation threshold, cause the intelligent prosthetic hand to enter a grasping pre-response state; in the grasping pre-response state, the initial values ​​of grasping timing and grasping speed are determined by the motion speed characteristics; the trend intensity change of the electromyographic signal finely adjusts the initial values ​​to achieve the flexibility and intention matching of the intelligent prosthetic hand's movements.

[0010] Preferably, it further includes: a proximity sensor disposed on the palm of the intelligent prosthetic hand, used to detect the distance between the target object and the palm of the intelligent prosthetic hand in real time; the control unit is configured such that when the distance is less than the effective trigger distance dynamically adjusted by the motion speed feature, the proximity sensor triggers a finger closing action, wherein a high motion speed feature corresponds to a longer effective trigger distance, and a low motion speed feature corresponds to a shorter effective trigger distance.

[0011] Preferably, the control unit is configured to control the finger closing process of the intelligent prosthetic hand using the following dynamic modes: when the forearm movement speed characteristic indicates rapid grasping, the finger closing action is performed with a high acceleration in the initial stage and closes smoothly with a low acceleration when approaching the target object; when the forearm movement speed characteristic indicates slow grasping, the finger closing action is performed with a smooth and gentle acceleration throughout the entire process.

[0012] Preferably, the control unit is configured to determine the grip timing adjustment amount ΔT according to the following formula:

[0013] ΔT=k v ·V speed +k emg ·S emg_trend ,

[0014] Among them, V speed S represents the quantized value of the motion velocity characteristic. emg_trend k represents the quantized value of the trend intensity change of electromyographic signals. v k is the grasping timing adjustment coefficient for motion speed characteristics. emg This is the adjustment coefficient for the trend intensity change of the electromyographic signal.

[0015] A control method for an intelligent prosthetic hand, characterized by comprising: acquiring motion posture information of the forearm where the prosthetic hand is located in real time through an inertial measurement unit and determining motion speed characteristics; acquiring electromyographic signals of specific muscle groups in the forearm in real time through an electromyography (EMG) sensor and identifying trend intensity changes of the EMG signals; and collaboratively determining the grasping timing and grasping speed of the intelligent prosthetic hand based on the motion speed characteristics and the trend intensity changes of the EMG signals; wherein, when the motion speed characteristics are high, an earlier grasping timing and a faster grasping speed are determined; when the motion speed characteristics are low, a later grasping timing is determined. The grasping speed is relatively slow; the trend intensity change of the electromyographic signal dynamically adjusts the grasping timing and the grasping speed. When the trend intensity change shows an accelerating increase, the grasping timing is further advanced and the grasping speed is further increased; when the trend intensity change shows a decelerating decrease, the grasping timing is appropriately delayed and the grasping speed is appropriately slowed down; when a target object is detected, the closing action of the fingers of the intelligent prosthetic hand is triggered according to the determined grasping timing and the grasping speed. The speed of the finger closing action is matched with the grasping speed to realize the anthropomorphic collaborative grasping of the intelligent prosthetic hand.

[0016] Preferably, the method further includes: continuously monitoring the contact pressure between the intelligent prosthetic hand and the target object after the intelligent prosthetic hand completes the grasp of the target object; dynamically adjusting the gripping force of the fingers of the intelligent prosthetic hand based on the contact pressure to ensure that the target object is held stably without falling off or being excessively squeezed; and maintaining synergistic consistency between the dynamic adjustment of the gripping force and the trend intensity changes of the initial grasping speed characteristics and electromyographic signals.

[0017] Preferably, the method further includes: acquiring contact force information between the intelligent prosthetic hand and the target object in real time through a force sensor installed on the intelligent prosthetic hand; when the contact force information exceeds a preset force threshold, the intelligent prosthetic hand automatically makes a slight adjustment of the gripping force to avoid damaging the target object; the magnitude of the slight adjustment of the gripping force is proportional to the degree to which it exceeds the threshold.

[0018] Preferably, the method can dynamically adapt to task requirements with different motion qualities, including industrial scenarios requiring speed and home scenarios requiring gentleness; the method can automatically adjust the behavior mode of the intelligent prosthetic hand to match the motion quality requirements of the task based on the detected motion speed characteristics, without requiring the user to perform additional mode switching or parameter settings.

[0019] Compared with the prior art, the beneficial effects of the present invention are:

[0020] 1. By capturing the forearm's motion speed characteristics through an inertial measurement unit and combining them with the dynamic trend intensity of electromyography (EMG) signals, the prosthetic hand's grasping timing and speed can match the natural rhythm of human movement in real time. Motion speed characteristics serve as the benchmark for macroscopic movement rhythm, while EMG trend intensity fine-tunes the details of the movement. The two work together to avoid the inherent contradiction between mechanical response lag and intention recognition separation in traditional control. Users do not need to deliberately plan grasping actions; the prosthetic hand can autonomously complete prediction and execution as the arm approaches the target, significantly reducing the cognitive load of operation. The interaction between EMG signal trend intensity and motion speed characteristics forms a multi-level adaptive adjustment capability. During high-speed movement, grasping is triggered in advance to avoid missing the target, while during low-speed movement, delayed response ensures accurate contact. The accelerating upward trend of EMG further enhances response agility, while the decelerating downward trend automatically switches to a buffer mode. This dual feedback mechanism based on physical motion state and neural excitation level enables the system to cope with both fast assembly line operations in industrial scenarios and gentle operation needs in home scenarios, without the need for manual mode switching.

[0021] 2. The technology of dynamically adjusting the proximity sensor trigger distance according to the movement speed solves the problem of coordinating the grasping timing and spatial positioning: high-speed movement corresponds to a longer trigger distance, reserving action time for mechanical closing; low-speed movement corresponds to a shorter trigger distance, realizing precise triggering at the moment of contact. Combined with the finger closing process, the acceleration is adjusted in stages according to the movement speed characteristics (fast first and then slow when high speed, and stable throughout when low speed), forming a complete anthropomorphic action chain from spatial positioning, timing decision to force control, which significantly improves the grasping success rate and the naturalness of operation.

[0022] 3. The inertial measurement unit reuses its dynamic data to generate velocity characteristics based on attitude detection, the electromyography sensor analyzes trend intensity changes in addition to the activation function, and the proximity sensor achieves distance adaptation through trigger logic reconstruction. This in-depth value mining of existing sensor data creates a triple functional gain of motion beat perception, intention intensity analysis and spatial distance coordination without zero hardware increase, avoiding the path dependence of traditional solutions that rely on adding new sensors to improve performance.

[0023] 4. The passive compliant structure of the finger joints automatically compensates for positional errors at the moment of contact, forming a dual guarantee of active positioning and passive adaptation with the grip control based on speed characteristics. When the contact pressure sensor detects abnormal grip force, the system combines the initial speed characteristics and real-time electromyography trends to make dynamic force adjustments. This avoids the computational complexity of traditional force control algorithms and solves the risk of overpressure / slippage in the grip of precision objects, achieving a balanced operating experience. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the overall structure of the intelligent prosthetic hand of the present invention;

[0025] Figure 2 This is a diagram of the transmission structure of the present invention, in which the driving source drives the fixed block to realize the finger joint movement;

[0026] Figure 3 This is a structural diagram showing the coordinated operation of the fixing block, the connecting structure, and the limiting component of the present invention.

[0027] Figure 4 This is a signal processing logic block diagram for the coordinated control of the inertial measurement unit and the electromyography sensor in this invention.

[0028] Figure 5 This is a schematic diagram illustrating the proximity sensor triggering mechanism and dynamic control of finger closure in this invention.

[0029] Figure 6 This is a graph showing the relationship between the grip timing adjustment amount and the forearm movement speed characteristics of the present invention.

[0030] Figure 7 This is a flowchart illustrating the collaborative sensing and dynamic adjustment process of the intelligent prosthetic hand grasping control system of the present invention.

[0031] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0032] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0033] This application provides an intelligent prosthetic hand comprising: an inertial measurement unit (IMU) for real-time acquisition of motion posture information of the forearm where the prosthetic hand is located, the motion posture information including the motion velocity characteristics of the forearm during a grasping task, the motion velocity characteristics being determined by processing the dynamic data of the IMU; an electromyography (EMG) sensor for real-time acquisition of EMG signals of specific muscle groups in the forearm and identification of trend intensity changes in the EMG signals, the trend intensity changes reflecting the dynamic intensity of the user's grasping intention; and a control unit electrically connected to the IMU and the EMG sensor, the control unit being configured to: collaboratively determine the grasping timing and grasping speed of the intelligent prosthetic hand based on the motion velocity characteristics and the trend intensity changes in the EMG signals; wherein, when the motion velocity characteristics exhibit high motion velocity... When the speed of movement is low, the control unit determines an earlier grasping timing and a faster grasping speed. When the movement speed is low, the control unit determines a later grasping timing and a slower grasping speed. The trend intensity change of the electromyographic signal dynamically adjusts the grasping timing and grasping speed. When the trend intensity change shows an accelerating increase, the grasping timing is further advanced and the grasping speed is further increased. When the trend intensity change shows a decelerating decrease, the grasping timing is appropriately delayed and the grasping speed is appropriately slowed down. Furthermore, when a target object is detected, the control unit triggers the closing action of the fingers of the intelligent prosthetic hand according to the determined grasping timing and grasping speed. The speed of the finger closing action is matched with the grasping speed to achieve anthropomorphic collaborative grasping of the intelligent prosthetic hand.

[0034] Preferably, the inertial measurement unit is configured to: acquire initial motion data of the forearm at the beginning of the grasping task and establish a reference motion baseline of the forearm based on the initial motion data; during the execution of the grasping task, monitor the changes in the motion data of the inertial measurement unit relative to the reference motion baseline in real time, and use the changes as the basis for judging the motion speed characteristics. The basis for judging the motion speed characteristics includes the rate of change of the motion data within a specific time window, the instantaneous velocity amplitude, or the velocity integral.

[0035] Preferably, the electromyography (EMG) sensor is configured to: filter the EMG signal to eliminate noise, and extract the envelope of the filtered EMG signal to obtain the trend intensity of the EMG signal; perform real-time dynamic analysis on the trend intensity of the EMG signal to identify its rising, falling or stable trend, as well as the rate of change of the trend intensity, thereby obtaining the trend intensity change.

[0036] Preferably, the control unit is configured to: when the motion speed characteristics indicate that the forearm motion speed reaches a preset grasping preparation threshold, cause the intelligent prosthetic hand to enter a grasping pre-response state; in the grasping pre-response state, the initial values ​​of grasping timing and grasping speed are determined by the motion speed characteristics; the trend intensity change of the electromyographic signal finely adjusts the initial values ​​to achieve the flexibility and intention matching of the intelligent prosthetic hand's movements.

[0037] Preferably, it further includes: a proximity sensor disposed on the palm of the intelligent prosthetic hand, used to detect the distance between the target object and the palm of the intelligent prosthetic hand in real time; the control unit is configured such that when the distance is less than the effective trigger distance dynamically adjusted by the motion speed feature, the proximity sensor triggers a finger closing action, wherein a high motion speed feature corresponds to a longer effective trigger distance, and a low motion speed feature corresponds to a shorter effective trigger distance.

[0038] Preferably, the control unit is configured to control the finger closing process of the intelligent prosthetic hand using the following dynamic modes: when the forearm movement speed characteristic indicates rapid grasping, the finger closing action is performed with a high acceleration in the initial stage and closes smoothly with a low acceleration when approaching the target object; when the forearm movement speed characteristic indicates slow grasping, the finger closing action is performed with a smooth and gentle acceleration throughout the entire process.

[0039] Preferably, the control unit is configured to determine the grip timing adjustment amount ΔT according to the following formula:

[0040] ΔT=k v ·V speed +k emg ·S emg_trend ,

[0041] Among them, V speed S represents the quantized value of the motion velocity characteristic. emg_trend k represents the quantized value of the trend intensity change of electromyographic signals. v k is the grasping timing adjustment coefficient for motion speed characteristics. emg This is the adjustment coefficient for the trend intensity change of the electromyographic signal.

[0042] A control method for an intelligent prosthetic hand, characterized by comprising: acquiring motion posture information of the forearm where the prosthetic hand is located in real time through an inertial measurement unit and determining motion speed characteristics; acquiring electromyographic signals of specific muscle groups in the forearm in real time through an electromyography (EMG) sensor and identifying trend intensity changes of the EMG signals; and collaboratively determining the grasping timing and grasping speed of the intelligent prosthetic hand based on the motion speed characteristics and the trend intensity changes of the EMG signals; wherein, when the motion speed characteristics are high, an earlier grasping timing and a faster grasping speed are determined; when the motion speed characteristics are low, a later grasping timing is determined. The grasping speed is relatively slow; the trend intensity change of the electromyographic signal dynamically adjusts the grasping timing and the grasping speed. When the trend intensity change shows an accelerating increase, the grasping timing is further advanced and the grasping speed is further increased; when the trend intensity change shows a decelerating decrease, the grasping timing is appropriately delayed and the grasping speed is appropriately slowed down; when a target object is detected, the closing action of the fingers of the intelligent prosthetic hand is triggered according to the determined grasping timing and the grasping speed. The speed of the finger closing action is matched with the grasping speed to realize the anthropomorphic collaborative grasping of the intelligent prosthetic hand.

[0043] Preferably, the method further includes: continuously monitoring the contact pressure between the intelligent prosthetic hand and the target object after the intelligent prosthetic hand completes the grasp of the target object; dynamically adjusting the gripping force of the fingers of the intelligent prosthetic hand based on the contact pressure to ensure that the target object is held stably without falling off or being excessively squeezed; and maintaining synergistic consistency between the dynamic adjustment of the gripping force and the trend intensity changes of the initial grasping speed characteristics and electromyographic signals.

[0044] Preferably, the method further includes: acquiring contact force information between the intelligent prosthetic hand and the target object in real time through a force sensor installed on the intelligent prosthetic hand; when the contact force information exceeds a preset force threshold, the intelligent prosthetic hand automatically makes a slight adjustment of the gripping force to avoid damaging the target object; the magnitude of the slight adjustment of the gripping force is proportional to the degree to which it exceeds the threshold.

[0045] Preferably, the method can dynamically adapt to task requirements with different motion qualities, including industrial scenarios requiring speed and home scenarios requiring gentleness; the method can automatically adjust the behavior mode of the intelligent prosthetic hand to match the motion quality requirements of the task based on the detected motion speed characteristics, without requiring the user to perform additional mode switching or parameter settings.

[0046] Example 1: This example provides an intelligent prosthetic hand based on physiological intention perception and multi-level feedback control mechanism. Its structural design and control method are consistent and coordinated as follows, enabling dynamic adaptive anthropomorphic grasping operations without requiring active user mode switching. The intelligent prosthetic hand includes a hand body 1. Four drive sources 2 are disposed on the surface of the hand body 1. The power output end of each drive source 2 is connected to a fixed block 3. The fixed block 3 can move back and forth horizontally on the surface of the hand body 1 to drive subsequent structures to complete finger joint movements. A connecting rod 4 is rotatably connected to the surface of each fixed block 3. The connecting rod 4 is rotatably connected to a connecting seat 5, thus forming a flexible and adjustable transmission structure. A finger joint body 6 is connected to the surface of the connecting seat 5. The finger joint body 6 is rotatably connected to a support seat 7 to enhance overall mechanical stability. The inner side of each fixed block 3 is slidably connected to the surfaces of two limiting rods 8 to control the movement trajectory of the fixed block 3. The structure effectively restricts the movement of the fixed block 3, ensuring stable reciprocating motion in the horizontal direction. During operation, the structure implements the following coordinated control mechanism: When the drive source 2 is running, it moves the fixed block 3 forward in the horizontal direction. The translational action of the fixed block 3 is transmitted to the adapter seat 5 via the adapter rod 4, causing the knuckle body 6 to rotate downwards to achieve a gripping action. When the drive source 2 is running in the reverse direction, the fixed block 3 causes the structure to reset, and the knuckle body 6 straightens accordingly. The limiting rod 8 limits the movement path of the fixed block 3 to the left and right, effectively improving the controllability and stability of the gripping process. The support seat 7 provides support for the knuckle body 6, further improving its structural reliability during the closing and resetting processes.

[0047] In terms of control strategy, the intelligent prosthetic hand integrates an inertial measurement unit (IMU) and an electromyography (EMG) sensor, which are used to collect dynamic motion information and EMG signals of the user's forearm, respectively. The IMU can acquire the forearm posture in real time and extract its motion speed characteristics as a basis for identifying the macroscopic grasping rhythm. The EMG sensor identifies the trend intensity changes of the EMG signal through filtering and envelope extraction, reflecting the dynamic changes of the user's grasping intention. The control unit makes collaborative decisions on the triggering timing and speed parameters of the grasping action based on the above two types of signals. Specifically, when the forearm's motion speed characteristics indicate a high speed and the EMG signal shows an accelerating upward trend, the control unit judges it as a rapid grasping intention and triggers the drive source 2 to work in advance, causing the fixed block 3 to... The system moves forward rapidly to quickly complete the grasping process; conversely, when the movement speed is low and the electromyographic trend is deceleration, the control unit delays triggering and reduces the drive rate to complete a gentle grasping action. At the same time, the prosthetic hand is equipped with a proximity sensor to detect the distance between the target object and the prosthetic hand. When the detected target distance is less than the trigger threshold dynamically adjusted by the control unit based on the movement speed characteristics, the system immediately starts the grasping action. The execution speed of the grasping action is consistent with the detected speed characteristics, realizing adaptive control in different scenarios. During the grasping process, feedback information from the hand contact pressure sensor is also combined to dynamically adjust the gripping force of the fingers to prevent the target object from slipping or being damaged by pressure.

[0048] Example 2: This invention proposes an intelligent prosthetic hand and its control method. This scheme achieves intelligent prediction and dynamic control of grasping movements by deeply fusing the forearm dynamic information provided by the inertial measurement unit (IMU) with the intention signal provided by the electromyography (EMG) sensor. Specifically, the key technical features defined in this invention mainly include the following aspects: First, the IMU is used to acquire the motion posture information of the forearm where the prosthetic hand is located in real time. The most crucial part of this information is the motion velocity feature. This feature is obtained by processing the triaxial acceleration and angular velocity data collected by the IMU through short-time window moving average, differential analysis, and velocity integration. This processing method avoids the instability of directly relying on high-precision positioning equipment, enabling this invention to achieve feasible velocity estimation at a lower cost, thereby supporting the subsequent motion control logic. The presence of motion speed characteristics not only characterizes the overall rhythm of the user's movements but also provides the control unit with a beat reference for triggering grasping. Secondly, electromyography (EMG) sensors are used to collect EMG signals from specific muscle groups in the forearm in real time and further identify their trend intensity changes. In the context of this invention, this term is defined as the rate of change of the EMG signal envelope over time, used to measure the dynamic expression of the user's intention. Specifically, after the raw EMG signal is bandpass filtered to avoid low-frequency drift and high-frequency noise, a trend curve is generated using a sliding window envelope extraction algorithm. Then, through the differentiation operation of this trend curve, the rising, falling, or stationary states and their rates are extracted to obtain the trend intensity change value. This feature can be regarded as a reflection of the degree of neural excitation, complementing the aforementioned macroscopic motion beat, thereby improving the control system's response sensitivity to the user's true intention. Based on the above two key inputs, the control unit undertakes the core scheduling and execution tasks of the grasping action. Its internal algorithm logic is based on the following strategy: when the motion speed characteristics show a high-speed state (e.g., the rate of change of speed is greater than a preset threshold θ), sp Furthermore, the change in electromyographic trend intensity showed a significant increase, such as the first derivative being greater than the threshold. If the system determines that the user's intention is to grasp quickly, the control unit immediately enters the grasping preparation state and triggers the grasping action at a relatively fast speed; otherwise, when the speed characteristics are low and the electromyographic trend slows down, the control unit delays the grasping opportunity and reduces the execution rate accordingly, thereby achieving a flexible response to the grasping action.

[0049] This control system further introduces a quantifiable grasping timing adjustment formula to improve the engineering adjustability and algorithm stability of the control:

[0050] ΔT=k v ·V speed +k emg ·S emg_trend ,

[0051] Among them, V speedS represents the quantized value of the motion velocity characteristics obtained from the inertial measurement unit. emg_trend k is the quantification value of the trend intensity change of electromyography. v With k emg The timing adjustment coefficients are set according to experimental optimization. At the implementation level, this formula can be calculated in real-time by a low-latency algorithm module embedded in the controller, thereby accurately outputting the control delay ΔT, making the system behavior more consistent with the user's current actions and intentions. Simultaneously, to achieve effective response to external target objects, this invention further integrates a proximity sensor to detect the spatial distance between the target object and the prosthetic hand. In the system design, the effective trigger distance set by this proximity sensor is no longer a static fixed value, but a dynamic variable related to motion speed characteristics. Specifically, high speed characteristics correspond to a longer trigger distance to reserve a response window for the prosthetic hand to perform actions in advance; low speed characteristics correspond to close-range triggering to ensure contact accuracy and safety. During finger closure, the control unit introduces a two-stage dynamic control strategy to further enhance the naturalness and anthropomorphism of the operation. The specific control modes are as follows: For fast grasping scenarios, the system initially drives closure with higher acceleration, gradually reducing acceleration as it approaches the target object, forming a closure curve that is fast at first and then slows down; for slow grasping scenarios, the system always executes closure with a stable and gentle acceleration to ensure safe contact with vulnerable targets. In the implementation of this system, to ensure that the prosthetic hand can adapt to different types of movement requirements, such as rapid grasping in industrial scenarios or delicate operations emphasizing gentle contact in home scenarios, the control unit is pre-programmed with a hierarchical judgment logic based on movement speed characteristics. This judgment logic classifies the current task movement into three categories: high speed, normal speed, and low speed, based on the range of speed data provided by the inertial measurement module, and associates them with matching movement strategies and parameter configurations respectively. For example, in high-speed tasks, the system automatically selects an early response strategy, prioritizing the use of faster drive rates and expanding the proximity sensing trigger distance; while in low-speed movements, a delayed response strategy is used and the trigger range is compressed to ensure the accuracy and safety of contact. In addition, the control unit also combines the strength changes of electromyographic signal trends to further fine-tune the timing of movement initiation and the rhythm of finger force application during the response process. These are all extended implementation methods known to those skilled in the art.

[0052] The aforementioned control strategy not only enhances the adaptability of the operation process but also effectively reduces the user's perceptual burden regarding specific action parameters. After the target is grasped, the system continues to monitor the contact pressure information between the fingers and the target (obtained through a force sensor integrated into the fingers) and performs real-time fine-tuning of the gripping force based on the initial grasping speed and electromyographic trends: if the contact force exceeds a set threshold P... thThe system reduces the output force linearly; if the contact force is insufficient, it slightly increases the gripping force, thus ensuring that the target object neither slips nor is excessively squeezed. This invention also possesses automatic task scenario adaptation capabilities: without external parameter configuration or explicit user mode switching, the system can automatically identify different scenarios based on current speed characteristics, such as the high speed requirements of an industrial environment or the preference for gentle movements in a home environment, thereby switching to a matching behavior mode. This feature greatly expands the practicality and adaptability of this invention in different application scenarios.

[0053] Example 3: This example achieves grasping control through a collaborative recognition mechanism of inertial measurement data and electromyographic trend signals. Specifically, in terms of structural composition, the intelligent prosthetic hand includes a hand execution component, a drive module, a control unit, an inertial measurement module, an electromyographic acquisition module, a proximity sensing module, and a grip force detection module. The hand execution component controls the closing and resetting actions of multiple phalanges through several micro linear drive devices. The multiple phalanges adopt a flexible connection structure to form a grasping mechanism with structural compliance capabilities, which can adapt to the surface features of objects of different shapes during execution. In terms of the sensing and processing module, the inertial measurement module is located in the area near the wrist of the user's forearm. It acquires the user's forearm acceleration and angular velocity data during the task using a three-axis accelerometer and gyroscope. The system processes this data based on a set sliding time window algorithm to extract velocity feature values ​​that reflect the rhythm of the forearm movements. These velocity features are used as a beat reference signal for subsequent grasping actions. At the same time, the electromyography (EMG) acquisition module consists of surface EMG sensors deployed in the target muscle group area of ​​the forearm to collect EMG activity signals when the user makes a grasping intention. The acquired raw EMG signals are bandpass filtered to suppress low-frequency drift and high-frequency interference, and then a trend change curve is generated using an envelope extraction method. The rate of change of this trend curve is used to quantify the EMG trend intensity to evaluate the dynamic characteristics of the user's intention change process. The control unit determines the response timing and execution strategy of the grasping task based on the dual inputs of velocity feature values ​​and EMG trend intensity changes.

[0054] The control process is specifically divided into the following four stages: The first stage is the intention perception initialization stage. In the system standby state, the control unit continuously monitors the data from the inertial measurement module and the electromyography (EMG) acquisition module. When the velocity characteristic exceeds the set static recognition threshold and the EMG trend shows a continuous increase, the system enters the grasping preparation state. The second stage is the grasping intention formation and response decision stage. Based on the velocity value extracted by the inertial measurement module and the rate of change of the EMG trend, combined with the set weighting logic, the control unit calculates the trigger delay value of the grasping action. This delay parameter can be reasonably set by those skilled in the art under the conventional parameter setting logic according to the task scenario to ensure that the system can make a matching response when dealing with fast or slow grasping needs. The third stage is the grasping action execution rhythm control stage. The control unit drives the hand knuckles according to the preset acceleration. The curve completes the closing action. When the velocity characteristic value is high, the knuckles perform the closing action with a faster acceleration in the initial stage and automatically decelerate in the final stage to avoid impacting the target object. If the velocity characteristic value is low, the system performs the entire closing process with a relatively constant and gentle acceleration to improve the smoothness of the action. The fourth stage is the closed-loop adjustment stage after grasping. The contact pressure data between the finger and the target object is obtained in real time through contact force sensors deployed in the finger area. Based on this pressure information, initial velocity characteristics and electromyographic trends, the system implements an automatic adjustment strategy for the grasping force: when the detected pressure exceeds the upper limit of the system's preset safety range, the control unit automatically fine-tunes the output of the knuckle driving force to reduce the pressure; if the contact force is insufficient, the grip force is appropriately increased to prevent the object from slipping. The above strategies aim to ensure that the prosthetic hand has sufficient gripping force during operation, while avoiding damage to the target due to excessive force. In addition, the proximity sensing module is placed in the palm of the prosthetic hand. It detects and dynamically adjusts the relative distance between the target object and the hand in combination with the aforementioned speed characteristic information. The control unit sets the dynamic trigger distance according to the current speed characteristics: in the state of rapid movement, the trigger distance is increased accordingly to reserve the finger joint closing response time; when approaching the target at low speed, the trigger distance is automatically reduced to improve the operation accuracy. This distance matching strategy significantly improves the spatiotemporal coupling execution effect in the grasping task.

[0055] Example 4: To verify the technical effects of an intelligent prosthetic hand and its control method in improving the naturalness of grasping, dynamic adaptability, and reducing the cognitive load on users, this experiment designed a verification scheme based on a practical application scenario. The intelligent prosthetic hand control system used in this experiment mainly consists of a forearm motion posture capture module with an inertial measurement unit (IMU), a high-precision surface electromyography (sEMG) signal acquisition module, an intelligent prosthetic hand body with an integrated grasping actuator, and an embedded control unit. The inertial measurement unit is fixed to the middle of the subject's forearm to monitor forearm motion posture information in real time, including the motion velocity characteristics of the forearm during the grasping task. This characteristic is obtained through inertial measurement... Unit dynamics data are processed and determined; the typical range of motion speed characteristics is set to 0.1 m / s to 1.5 m / s to cover common operating scenarios. The electromyography (EMG) signal acquisition module acquires EMG signals of specific muscle groups in the forearm in real time through electrodes placed on the surface of specific muscle groups, and identifies the trend intensity change of the EMG signal. This trend intensity change reflects the dynamic intensity of the user's grasping intention. The quantization range of the trend intensity change of the EMG signal is set to -100% per second to +100% per second. The proximity sensor is set on the palm of the smart prosthetic hand to detect the distance between the target object and the palm of the prosthetic hand in real time. The effective detection distance range is set to 1 cm to 20 cm.

[0056] The control unit is electrically connected to the inertial measurement unit and electromyography (EMG) sensor. It is configured to collaboratively determine the grasping timing and speed of the intelligent prosthetic hand based on motion velocity characteristics and the trend intensity changes of EMG signals. When the motion velocity characteristics are high, the control unit determines an earlier grasping timing and a faster grasping speed; when the motion velocity characteristics are low, the control unit determines a later grasping timing and a slower grasping speed. The trend intensity changes of the EMG signals dynamically adjust the grasping timing and speed. When the trend intensity change shows an accelerating increase, the grasping timing is further advanced and the grasping speed is further increased; when the trend intensity change shows a decelerating decrease, the grasping timing is appropriately delayed and the grasping speed is appropriately slowed. Upon detecting a target object, the control unit triggers the closing action of the fingers of the intelligent prosthetic hand according to the determined grasping timing and speed. The speed of the finger closing action matches the grasping speed, achieving anthropomorphic collaborative grasping of the intelligent prosthetic hand. The grasping timing adjustment amount ΔT is determined according to the formula ΔT = k v ·V speed +k emg ·S emg_trend Proceeding, where V speed S represents the quantized value of the motion velocity characteristic. emg_trend k represents the quantized value of the trend intensity change of electromyographic signals. v k is the grasping timing adjustment coefficient for motion speed characteristics. emgThe adjustment coefficient for the trend intensity change of electromyographic (EMG) signal is k, which is used in engineering practice. v It can be set to 0.2 seconds per meter per second, k emg It can be set to 0.005 seconds per percent per second. The selection of these parameters is intended to balance the influence of various factors on the timing of the grip.

[0057] This experiment invited several amputees to participate, simulating everyday household scenarios (such as picking up cups and books) and light industrial sorting scenarios (such as quickly grasping small parts). In the everyday household scenario, subjects typically operate with a low forearm movement speed, and the change in electromyographic signal intensity is relatively gradual. Under these conditions, the system, through dynamic adjustment of the control unit, usually triggers grasping when the target object is about 5 cm away from the palm and completes the grasp at a relatively slow speed. This makes the grasping action more gentle and precise, meeting the needs of daily fine motor operations. In the light industrial sorting scenario, subjects typically move their forearms faster to improve efficiency, and the change in electromyographic signal intensity shows a rapid upward trend. Under these conditions, the system, based on the synergistic effect of movement speed characteristics and electromyographic trend intensity, significantly advances the grasping timing, generally triggering grasping when the target object is about 15 cm away from the palm, and correspondingly accelerating the finger closing speed. This dynamic adjustment strategy effectively avoids errors that may be caused by response lag in high-speed grasping. Experimental results show that, under different combinations of quantified values ​​of motion velocity characteristics and quantified values ​​of electromyographic signal trend intensity changes, the grasping timing adjustment exhibits a dynamic change consistent with expectations. When the quantified value of motion velocity characteristics is high and the quantified value of electromyographic signal trend intensity changes is positive (increasing trend), the grasping timing adjustment tends to be negative, meaning the grasping timing is advanced. Conversely, when the quantified value of motion velocity characteristics is low and the quantified value of electromyographic signal trend intensity changes is negative (decreasing trend) or close to zero, the grasping timing adjustment tends to be positive, meaning the grasping timing is delayed. This verifies that the control unit dynamically adjusts the grasping timing. The effectiveness of the grip is enhanced. Furthermore, after grasping, the intelligent prosthetic hand continuously monitors the contact pressure between itself and the target object, and dynamically adjusts the gripping force of its fingers based on the contact pressure to ensure that the target object is held stably without slipping or being excessively squeezed. This dynamic adjustment of gripping force is coordinated with the initial grasping speed characteristics and the trend intensity changes of electromyographic signals. For example, when grasping fragile items, the system can automatically adjust the gripping force to a level that is just enough to hold them stably; when grasping heavy objects, it can provide sufficient gripping force to prevent slippage. This dynamic gripping force adjustment mechanism further improves the reliability and safety of the grasp.

[0058] Example 5: This example combines Figures 6 to 7 This paper describes the implementation of an intelligent prosthetic hand and its control method. Figure 6As shown, the horizontal axis represents the forearm movement speed characteristic (m / s), and the vertical axis represents the grasping timing adjustment amount ΔT (seconds). This reflects the trend that as the forearm movement speed characteristic increases, the grasping timing adjustment amount ΔT gradually decreases, indicating that under high movement speed conditions, the grasping action needs to be triggered in advance. The figure uses different shapes to mark three types of application scenarios, including home scenarios (marked with circles), industrial scenarios (marked with squares), and transition intervals (marked with triangles), which respectively represent typical action rhythm types. Each data point is labeled with a distance value (unit: centimeters), indicating that the system triggers the grasping action at that distance under the corresponding speed characteristic. For example, in the home scenario, the triggering distance corresponding to the lower speed is 15 centimeters, and in the transition interval, which is in the middle speed range (0.5 m / s), the triggering distance is 10 centimeters.

[0059] like Figure 7 As shown, in this system, the inertial measurement unit is used to acquire the forearm movement speed characteristics, providing a motion rhythm reference for the control unit; the electromyography sensor identifies the trend intensity of the grasping intention, providing key data for fine-tuning the details of the movement; the proximity sensor is used to detect the distance to the target object to trigger finger closure. Based on the above three input signals, the control unit issues adaptive grasping commands (timing, speed, acceleration) to drive the intelligent prosthetic hand to perform anthropomorphic collaborative grasping actions. During the execution, the system receives gripping force feedback from contact pressure monitoring (dynamic adjustment of gripping force) in real time, further closing the loop to regulate the execution behavior of the intelligent prosthetic hand and ensure that the grasping action is safe, natural, and stable.

[0060] Example 6: In this example, focusing on typical application scenarios of intelligent prosthetic hands in daily life assistance, a multimodal perception input system integrating electromyographic signals, multi-point tactile feedback, and wrist dynamic posture data was constructed. Based on this, precise control and dynamic adjustment of the prosthetic hand's grasping movements were achieved, enhancing its controllability, adaptability, and operational stability in actual use. During operation, firstly, multiple electromyographic signal acquisition electrodes placed on the surface of the residual limb simultaneously collect electromyographic data. The electrode arrangement can be personalized according to the individual residual limb muscle distribution to ensure signal acquisition quality and user comfort. Simultaneously, the prosthetic hand... The flexible tactile sensing unit at the palm position continuously collects hand contact pressure information and simultaneously integrates real-time spatial posture data measured by the posture sensor located at the wrist, including but not limited to dynamic parameters such as tilt angle, rotational angular velocity, and linear acceleration. The above-mentioned multi-source sensing data is processed by the data fusion processing module of the present invention for feature extraction and dynamic correlation modeling. This module preferably adopts a sliding window and weighted filtering strategy to enhance signal robustness, and combines a real-time threshold judgment mechanism to construct an electromyographic trigger signal model that reflects the user's current intention. After determining that the current electromyographic signal is in a high-confidence intention trigger state, the control command will be generated and transmitted to the prosthetic finger driving unit.

[0061] In the actual control strategy, the drive command not only considers the intensity of electromyographic intent but also combines the current tactile perception results to quickly estimate the hardness of the target object. Based on this, the drive rate and upper limit of the force applied by each finger execution unit are dynamically adjusted. For example, when the target object is identified as a lightweight and fragile material, the system will reduce the initial clamping force and further slow down the force application rate after detecting contact feedback to avoid damage to the object due to control lag or force overshoot. Conversely, if the target has obvious hardness characteristics, the system can prioritize the use of a higher initial drive force and enter the holding force stabilization control stage after the feedback confirms contact to improve operational efficiency. It is worth noting that, in order to improve control accuracy and response speed, this embodiment further introduces a local prediction adjustment mechanism. That is, before the electromyographic signal shows a significant trend change and before reaching the intent trigger threshold, the system can predict the trend in advance based on the historical signal sequence and preload the control channel to shorten the trigger response delay. At the same time, when the prediction error exceeds the set threshold, the system will automatically cancel the preload channel and restore the standard trigger mechanism to ensure safety and control stability. In several typical scenarios, such as picking up a glass, holding a key, or turning a doorknob, the system demonstrated good responsiveness to action intent, stable grip control, and automatic adjustment strategies under different target characteristics, verifying the effectiveness and engineering feasibility of the control logic under actual operating conditions.

[0062] Example 7: In this example, to further clarify the definition and mechanism of key parameters in the control strategy, this example provides a detailed technical explanation of the grasping timing adjustment process of the intelligent prosthetic hand, based on existing control structures. Firstly, regarding the motion speed characteristic parameter used in the intelligent prosthetic hand control system, this example clarifies that its physical source is the forearm dynamic acceleration and angular velocity data acquired by the inertial measurement unit. Specifically, the system extracts the speed change trend reflecting the user's current movement rhythm by performing moving average and differential analysis on the sampled values ​​within a short time window. In actual engineering implementation, this parameter represents the activity level of the user's arm movement during task initiation, guiding the overall timing selection of the grasping behavior. Simultaneously, to respond to changes in the intensity of the user's intent during movement, the system introduces the intensity of the electromyographic trend change as another important input parameter. In this example, this parameter originates from the slope of the envelope curve of the electromyographic signal on the forearm surface, i.e., the rate of increase or decrease of the electromyographic signal intensity per unit time, with a strong trend. A positive increase in the degree usually indicates that the user's intention is gradually strengthening, while a downward trend indicates that the user's grasping intention is becoming more relaxed. To improve recognition accuracy, the system uses a dynamic sliding window algorithm to suppress short-term jitter and combines it with a threshold strategy to avoid accidental triggering of control logic due to occasional fluctuations. In grasping control, these two types of parameters work together to adjust the timing and speed of the grasping action. The macro rhythm is determined based on speed characteristics, and the trend intensity is used to refine the immediacy and stability of the grasping response. When the system detects that the user's arm movement speed is in a high range and the electromyography trend is significantly increasing, the system will prepare to drive the execution in advance and generate the grasping action command before the target distance enters the traditional trigger threshold. This approach compensates for the timing error caused by sensor response delay under high-speed movements, effectively ensuring the synchronization and stability of the grasping action. When the speed characteristics are low and the electromyography trend changes slowly or decreases, the system automatically delays the action response, making the grasping process gentler, thus adapting to low-speed fine movement scenarios such as daily household operations. In this embodiment, all parameters involved in the adjustment process can be output in real time by the algorithm module in the control unit according to the set calculation logic. Without the introduction of external devices or special algorithms, the necessary logic judgment and execution control can be completed by the existing microcontroller chip, ensuring that the present invention can run stably on a standard computing resource platform. In addition, the set parameter adjustment range is not a fixed value, but can be reasonably adjusted according to the individual differences and usage habits of different users, achieving flexible adaptation without affecting the control logic structure.

[0063] Regarding parameter interpretation, this embodiment further provides a technical explanation of the adjustment coefficients involved in the grasping timing adjustment process. The coefficient used to adjust the degree of influence of speed characteristics reflects the system's responsiveness to the rhythm of user movements; the coefficient used to adjust the degree of influence of electromyographic trends reflects the system's immediate response to changes in user intent. In the control unit, both types of coefficients can be preset or optimized through self-learning based on feedback from long-term use. The goal of parameter setting is not to obtain a specific ideal result, but to achieve engineering adaptation to different usage scenarios without introducing computational burden. From the perspective of the physical implementation mechanism of the grasping action itself, the response strategy adopted in this embodiment aims to maintain the coordination between finger closure and the rhythm of external movements. For example, under high-speed tasks, the prosthetic hand will… A rapid start and slow finish closing strategy is adopted to ensure the integrity of the grasping action. In low-speed tasks, the closing action remains stable throughout, reducing contact impact caused by rapid closing. This closing mechanism does not rely on complex calculations, but is the result of coordinated control through acceleration output settings and sensing trigger distance. Finally, to enhance the adaptive capability of the control strategy, the grasping trigger mechanism used in this embodiment does not set a fixed threshold, but dynamically adjusts the target approach distance based on the current speed characteristics, realizing coupled control of target distance and reaction timing. The significance of this setting is that the system can autonomously adjust response conditions at different rhythms without user input or manual intervention, thereby truly achieving high sensitivity operation under low cognitive load. All of these are extended implementation methods known to those skilled in the art.

[0064] 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 present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent prosthetic hand, characterized in that, The application relates to a prosthetic hand, comprising: an inertial measurement unit for acquiring real-time motion posture information of a forearm where the prosthetic hand is located, the motion posture information comprising motion speed characteristics of the forearm in a grasping task, the motion speed characteristics being determined by processing dynamics data of the inertial measurement unit; an electromyography sensor for acquiring real-time electromyography signals of a forearm muscle group and identifying trend intensity changes of the electromyography signals, the trend intensity changes reflecting dynamic intensity of a user's grasping intention; a control unit electrically connected with the inertial measurement unit and the electromyography sensor, the control unit being configured to cooperatively determine grasping timing and grasping speed of the intelligent prosthetic hand based on the motion speed characteristics and the trend intensity changes of the electromyography signals; wherein when the motion speed characteristics represent high motion speed, the control unit determines early grasping timing and fast grasping speed; when the motion speed characteristics represent low motion speed, the control unit determines late grasping timing and slow grasping speed; the trend intensity changes of the electromyography signals dynamically adjust the grasping timing and the grasping speed; when the trend intensity changes represent accelerated rising, the grasping timing is further advanced and the grasping speed is further accelerated; when the trend intensity changes represent decelerated falling, the grasping timing is appropriately delayed and the grasping speed is appropriately slowed down; and when a target object is detected, the control unit triggers a closing action of fingers of the intelligent prosthetic hand according to the determined grasping timing and grasping speed, the speed of the finger closing action matches the grasping speed, and the humanization cooperative grasping of the intelligent prosthetic hand is realized.

2. The intelligent prosthetic hand of claim 1, wherein The inertial measurement unit is configured to acquire initial motion data of the forearm at a starting stage of the grasping task, and establish a reference motion baseline of the forearm based on the initial motion data; in the process of executing the grasping task, the inertial measurement unit is configured to monitor changes of motion data relative to the reference motion baseline in real time, and the changes are taken as a judgment basis of the motion speed characteristics, the judgment basis of the motion speed characteristics comprising a change rate, an instantaneous speed amplitude or a speed integral of the motion data within a set time window.

3. The intelligent prosthetic hand of claim 1, wherein The electromyography sensor is configured to filter the electromyography signals to eliminate noise, and obtain trend intensity of the electromyography signals by envelope extraction of the filtered electromyography signals. The trend intensity of the electromyography signals is dynamically analyzed in real time to identify rising, falling or stable change trends and rates of the change trends, so as to obtain trend intensity changes.

4. The intelligent prosthetic hand of claim 1, wherein, The control unit is configured to make the intelligent prosthetic hand enter a grasping pre-response state when the motion speed characteristics indicate that the motion speed of the forearm reaches a preset grasping preparation threshold; in the grasping pre-response state, initial values of the grasping timing and the grasping speed are determined by the motion speed characteristics; and the trend intensity changes of the electromyography signals fine-tune the initial values.

5. The intelligent prosthetic hand of claim 1, wherein, Further comprising: a proximity sensor arranged on a palm of the intelligent prosthetic hand and used for detecting a distance between the target object and the palm of the intelligent prosthetic hand in real time; the control unit is configured to trigger the finger closing action when the distance is less than an effective triggering distance dynamically adjusted by the motion speed characteristics, wherein high motion speed characteristics correspond to a farther effective triggering distance, and low motion speed characteristics correspond to a closer effective triggering distance.

6. A smart prosthetic hand according to claim 5, characterized in that The control unit is configured to control the finger closing process of the intelligent artificial hand to adopt the following dynamic mode: when the current arm movement speed feature indicates a fast grasp, the finger closing action is performed at a high acceleration in the initial stage and smoothly closes at a low acceleration when approaching the target object; when the current arm movement speed feature indicates a slow grasp, the finger closing action is performed at a smooth and gentle acceleration throughout. The method further comprises: after the intelligent artificial hand completes the grasping of the target object, continuously monitoring the contact pressure between the intelligent artificial hand and the target object; based on the contact pressure, dynamically adjusting the finger holding force of the intelligent artificial hand to ensure that the target object is stably held without falling off or being excessively squeezed; the dynamic adjustment of the holding force is consistent with the initial grasping speed feature and the trend intensity change of the electromyographic signal.

7. The intelligent prosthetic hand of claim 1, wherein The control unit is configured to determine the gripping timing adjustment amount in accordance with the following equation : , wherein is a quantification of the movement velocity characteristic, is a quantification of the change in the trend intensity of the myoelectric signal, is a grip timing adjustment coefficient for the movement velocity characteristic, is a myoelectric signal trend intensity change adjustment coefficient for the change in the trend intensity of the myoelectric signal.

8. The control method of the intelligent prosthetic hand according to claim 1, characterized by, The method further comprises: acquiring the motion posture information of the forearm where the artificial hand is located in real time through an inertial measurement unit, and determining the motion speed feature; The method further comprises: acquiring the electromyographic signal of the forearm muscle group in real time through an electromyographic sensor, and identifying the trend intensity change of the electromyographic signal; based on the motion speed feature and the trend intensity change of the electromyographic signal, the grasping timing and grasping speed of the intelligent artificial hand are determined; when the motion speed feature shows high motion speed, the grasping timing and grasping speed are determined; when the motion speed feature shows low motion speed, the grasping timing and grasping speed are determined; the trend intensity change of the electromyographic signal dynamically adjusts the grasping timing and grasping speed, when the trend intensity change shows acceleration rising, the grasping timing is further advanced and the grasping speed is further accelerated; when the trend intensity change shows deceleration descending, the grasping timing is appropriately delayed and the grasping speed is appropriately slowed down; When the target object is detected, the closing action of the fingers of the intelligent artificial hand is triggered according to the determined grasping timing and grasping speed, the speed of the finger closing action matches the grasping speed, and the anthropomorphic cooperative grasping of the intelligent artificial hand is realized. The method further comprises: after the intelligent artificial hand completes the grasping of the target object, continuously monitoring the contact pressure between the intelligent artificial hand and the target object; based on the contact pressure, dynamically adjusting the finger holding force of the intelligent artificial hand to ensure that the target object is stably held without falling off or being excessively squeezed; the dynamic adjustment of the holding force is consistent with the initial grasping speed feature and the trend intensity change of the electromyographic signal.

9. The control method of the intelligent prosthetic hand according to claim 8, characterized in that, The method further comprises: acquiring the contact force information between the intelligent artificial hand and the target object in real time through the force sensor arranged on the intelligent artificial hand; when the contact force information exceeds the preset force threshold, the intelligent artificial hand automatically adjusts the holding force by a small amplitude to avoid damaging the target object; the amplitude of the small amplitude holding force adjustment is proportional to the degree of exceeding the threshold.

10. The control method of the intelligent prosthetic hand according to claim 9, characterized in that, ​

Images