Bionic multi-finger underwater manipulator device, control system, grabbing strategy and target identification method

Abstract

The invention relates to a bionic multi-finger underwater manipulator device, a control system, a grabbing strategy and a target identification method.The manipulator is mainly composed of a flange plate and an under-actuated mechanical finger of a seven-connecting-rod mechanism, and the mechanical finger comprises a base, a triangular piece, a fingertip piece and a sliding block; two corners of the triangular piece are connected with the fingertip piece and the sliding block through connecting rod mechanisms respectively, the third corner of the triangular piece is connected with the base through a trapezoidal connecting piece, the sliding block is driven by a threaded linear long-axis motor, a magnetic touch sensor is arranged in a fingertip, and the magnetic touch sensor receives fingertip touch information through a fingertip receptor unit made of a magnetic powder silica gel mixture. The device is controlled by a control system comprising a microprocessor, naturally adapts to objects with different shapes and sizes by formulating corresponding adaptive force grabbing strategies, and classifies and identifies underwater objects through a deep learning algorithm. Compared with the prior art, the method has the advantages of high sensitivity, interference resistance and the like, and has the capabilities of tactile perception recognition and self-adaptive grabbing.

Description

Technical Field

[0001] The invention relates to a multi-finger manipulator, in particular to a bionic multi-finger underwater manipulator device, a control system, a grasping strategy and a target recognition method. Background Art

[0002] With the increasing demand for underwater operations, underwater robots have been widely used in deep-sea exploration, scientific research, resource exploration, and environmental monitoring. In these applications, the design and control of manipulators have become one of the key technologies. In underwater environments, the working conditions of manipulators are more complex and are affected by multiple factors such as water pressure, flow, and object characteristics. Therefore, manipulators with rigid designs can provide stable grasping force and reliable manipulation performance. This design usually includes a joint structure with multiple degrees of freedom, which can simulate the movement of human fingers, thereby achieving effective grasping and manipulation of objects. At the same time, the robustness of rigid manipulators gives them better durability and reliability when facing high water pressure and strong water flow.

[0003] In recent years, with the rapid development of artificial intelligence and sensor technology, intelligent manipulators combined with tactile perception have gradually become a research hotspot. By integrating tactile sensors, the manipulator can monitor the force applied during the grasping process in real time, and adjust the grasping strategy based on the feedback information to achieve adaptive grasping. This method significantly improves the manipulator's operational performance in complex underwater environments, enabling it to better complete a variety of tasks, such as grasping and identifying objects.

[0004] However, the current underwater manipulator still has some shortcomings, such as the manipulator's driving structure is relatively complex and the maintenance cost is high; the underwater tactile sensor is not sensitive, has weak anti-interference ability, and is not durable enough; the integration, control, and optimization of the recognition algorithm of the tactile sensor are not fully combined; the manipulator cannot accurately adjust the grasping angle and force according to the characteristics of the object, is not sensitive enough when facing slippage, and is not accurate enough when identifying objects. Therefore, facing the flexible configuration requirements of multiple working modes, developing a lightweight under-actuated underwater bionic manipulator combined with efficient tactile perception and intelligent control has become an urgent need to improve the efficiency and intelligence of underwater operations. Summary of the invention

[0005] The purpose of the present invention is to overcome the problems that existing manipulators do not have highly sensitive and anti-interference tactile perception and recognition capabilities and adaptive grasping capabilities, and to provide a new lightweight bionic multi-finger underwater manipulator device, control system, grasping strategy and target recognition method with tactile perception function, flexible adjustment of flange size and integration at the end of various underwater robots or underwater equipment.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] According to one aspect of the present invention, there is provided a bionic multi-finger underwater manipulator device, comprising a fixed flange and a plurality of manipulator fingers connected to the side of the fixed flange;

[0008] The mechanical finger includes a base, a triangle, a fingertip and a slider. The three corners of the triangle are connected to the fingertip, the slider and the base through a connecting rod mechanism. The point where the triangle is connected to the base through the connecting rod mechanism is a fixed rotation point. The slider is used to push one corner of the triangle to rotate around the fixed rotation point through linear motion, thereby driving the fingertip to swing.

[0009] The slider is driven by a threaded linear long-axis motor. A magnetic tactile sensor is arranged in the fingertip part. The magnetic tactile sensor receives fingertip tactile information through a fingertip receptor unit. The fingertip receptor unit is a magnetic powder and silicone mixture.

[0010] Furthermore, the mechanical finger is a seven-bar linkage, the base is fixedly connected to the fixed flange, and the seven-bar linkage includes a driving long link, a triangular piece, a first finger joint link, a first finger key, a second finger web, a third finger joint link and a fingertip link; one end of the triangular piece is connected to the slider through the driving long link, and the other end is connected to the first finger key, and the middle is connected to the base through a trapezoidal connecting piece;

[0011] One end of the first finger key is connected to the triangle piece, and the other end is connected to the second finger web; one end of the second finger web is connected to the first finger key, and the other end is connected to the fingertip connecting rod, and the middle is connected to the first knuckle connecting rod; one end of the first knuckle connecting rod is connected to the middle of the triangle piece, one end is connected to the third knuckle connecting rod, and the middle end is connected to the second finger web; one end of the third knuckle connecting rod is connected to the fingertip connecting rod, and the other end is connected to the end of the first knuckle connecting rod; the fingertip connecting rod connects the second finger web and the third knuckle connecting rod;

[0012] The slider is located on the base and has the freedom of reciprocating motion along the axial direction of the mechanical finger, and the finger tip piece is embedded in the surface of the finger tip connecting rod.

[0013] Furthermore, the mechanical finger also includes a finger pulp, which is fixedly connected to the first finger joint connecting rod.

[0014] Furthermore, the fingertip piece includes a fingertip base, a magnetic tactile sensor and an isolation bracket which are connected in sequence. A notch is provided on the fingertip base, and the fingertip piece is embedded in the fingertip connecting rod through the notch on the fingertip base.

[0015] Furthermore, a groove is provided on the isolation bracket, and the fingertip receptor unit is located in the groove, and the isolation bracket separates the magnetic tactile sensor and the fingertip receptor unit.

[0016] Furthermore, a motor compartment is provided on the side of the fixed flange, a threaded linear long-axis motor is fixed in the motor compartment, and the motor shaft extends from the motor compartment and is connected to the slider, and an AB two-phase Hall encoder is provided at the rear of the threaded linear long-axis motor.

[0017] Furthermore, the motor compartment includes a motor seat and a shell, the motor seat is fixedly connected to the fixed flange, the shell is connected to the motor seat through a mortise and tenon structure, the AB two-phase Hall encoder is connected to the threaded linear long-axis motor, and the motor seat and the shell seal the AB two-phase Hall encoder and the threaded linear long-axis motor.

[0018] According to another aspect of the present invention, a control system for a bionic multi-finger underwater manipulator device is provided, characterized in that it includes a microprocessor, a voltage stabilizing module, a crystal oscillator circuit, a reset circuit, a motor drive module, a Flash memory, a tactile sensor and an indicator light; the microprocessor uses STM32F103C8T6 as a main control chip, is equipped with a voltage stabilizing module, a crystal oscillator circuit and a reset circuit, and uses a SWD interface as a programmer, and is connected to the motor drive module through an IO port; the motor drive module generates PWM to drive the threaded linear long-axis motor 2 of each mechanical finger; the microprocessor is also connected to the flash memory through the SPI bus interface, and is connected or reads the magnetic tactile sensor 152 corresponding to each mechanical finger through the IIC bus interface and address selection; the control system displays status information through the indicator light.

[0019] According to another aspect of the present invention, a grasping strategy of a bionic multi-finger underwater manipulator device is provided, characterized in that it includes the following steps:

[0020] The change of friction between the fingertips and the object during grasping is monitored in real time through magnetic tactile sensors;

[0021] When the system detects a change in friction force, it determines whether the change in friction force exceeds a slip detection threshold;

[0022] When the friction force change exceeds the slip detection threshold, the gripping force adjustment amount of the manipulator device is calculated by combining the target gripping force and using the gripping force adjustment formula;

[0023] The threaded linear long-axis motor is controlled to adjust the friction force until the friction force change does not exceed the slip detection threshold.

[0024] According to another aspect of the present invention, a method for target recognition of a bionic multi-finger underwater manipulator device is provided, characterized in that it comprises the following steps:

[0025] Extracting feature vectors from contact force data collected by a magnetic tactile sensor;

[0026] Obtain the convolutional layer output from the feature vector through a convolutional neural network deep learning model;

[0027] Downsample the output of the convolutional layer to get the output of the pooling operation;

[0028] The extracted features are mapped to the output layer through a series of fully connected layers. The number of neurons in the output layer corresponds to the number of categories of the target objects.

[0029] Activate the output and generate the classification result C of the target object.

[0030] Compared with the prior art, the present invention has the following advantages:

[0031] 1. The mechanical finger of the present invention adopts a seven-link under-actuated mechanism design, simplifies the transmission scheme, uses a single driver to control multiple joints, can accurately adjust the grasping force and angle according to the characteristics of the object, and the multi-finger coordinated grasping and single-finger grasping methods enable it to have better grasping reliability. The robustness of the rigid-link manipulator enables it to have better durability and reliability when facing high water pressure and strong water flow; in addition, the magnetic tactile sensor provided at the fingertips of the present invention has the advantages of high sensitivity, strong anti-interference ability, and high durability, and can provide accurate multi-dimensional force perception in complex environments, thereby improving its perception ability.

[0032] 2. The invention proposes a strategy for adjusting the grasping force in real time according to the slip detection data. The slip can be detected in real time by a magnetic tactile sensor, and the grasping force can be adaptively adjusted according to the slip data to improve its grasping stability.

[0033] 3. The underwater bionic three-finger manipulator with tactile perception function designed by the present invention adopts a deep learning algorithm to extract the features required for grasping an object. By evaluating the grasping force applied to different objects, the system can infer the shape, hardness, etc. of the object, thereby achieving accurate recognition of the target and improving its intelligent operation level. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a structural schematic diagram of the present invention;

[0035] Figure 2 It is a schematic diagram of the structure of the mechanical finger of the present invention;

[0036] Figure 3 An exploded view of the mechanical finger structure of the present invention;

[0037] Figure 4 An exploded view of the fixed flange structure of the present invention;

[0038] Figure 5 An exploded view of the fingertip member of the present invention;

[0039] Figure 6 It is a schematic diagram of the control system framework of the present invention;

[0040] Figure 7 is a control block diagram of the adaptive gripping force adjustment strategy of the present invention;

[0041] Figure 8 It is a schematic diagram of the deep learning algorithm flow of the present invention;

[0042] In the attached drawings: 1-fixed flange; 2-threaded linear long axis motor; 3-base; 4-slider cover; 5-slider; 6-driving long connecting rod; 7-trapezoidal connecting piece; 8-triangular piece; 9-finger pulp; 10-first finger joint connecting rod; 11-first finger key; 12-second finger pulp; 13-third finger joint connecting rod; 14-fingertip connecting rod; 15-fingertip piece; 151-fingertip base; 152-magnetic tactile sensor; 153-isolation bracket; 16-motor compartment; 17-AB two-phase Hall encoder; 18-hexagonal nut. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0044] The present invention provides an underwater bionic three-finger manipulator with tactile sensing function, the main body of the manipulator includes a fixed flange 1 and a manipulator arranged on the side of the fixed flange 1. The manipulator adopts an underactuated anthropomorphic hand design, which is a seven-link mechanism, including a triangular member 8, and realizes functions such as three-finger collaborative grasping, single-finger grasping, and soil excavation and sand excavation through the motion transmission of the connecting rod. Among them, the first finger joint connecting rod 10 is embedded in a finger pulp 9, and the fingertip connecting rod 14 is embedded in a fingertip member 15. In particular, the fingertip member 15 is embedded with a magnetic tactile sensor 152 and a fingertip receptor unit, and the fingertip receptor unit is made by adding magnetic powder to silica gel and then casting, and is isolated from the magnetic tactile sensor 152 by a bracket. The microprocessor is installed in a waterproof and sealed cabin of various underwater robots or underwater equipment carrying this manipulator, and is used to control the movement of the manipulator. The manipulator is driven by three threaded linear long-axis motors 2 respectively, and is provided with an AB two-phase Hall encoder 17 for accurate reading of posture. The robot arm realizes multiple functions such as three-finger collaborative grasping, single-finger grasping, and soil and sand digging through ingenious connecting rod motion transmission.

[0045] The working principle of the present invention is: an underwater bionic three-finger manipulator with tactile sensing function is fixed to the end of an underwater robot or underwater equipment. The manipulator transmits through a clever connecting rod motion, accurately adjusts the grasping force and angle according to the characteristics of the object, and realizes three-finger collaborative grasping. It is also possible to grasp slender, spherical objects independently with a single finger. For objects hidden in soft sand or soil, they can also be preliminarily excavated and grasped by the end of the finger. In addition, the manipulator can detect slippage in real time through the magnetic tactile sensor 152, and adaptively adjust the grasping force according to the slippage data. When the sensor detects that the object has a slight slip at the fingertips, the system gradually increases the grasping force through feedback control to prevent the object from falling off. Finally, the manipulator collects the contact force data during the grasping process in real time through the magnetic tactile sensor 152, and uses this as a basis for identifying and classifying the target object.

[0046] like Figure 1 As shown, the threaded linear long-axis motor 2 is fixed to the motor compartment 16 on the side of the fixed flange 1. The motor compartment 16 is composed of a motor seat and a housing. The motor seat and the housing are connected by the mortise and tenon principle to seal the threaded linear long-axis motor 2. The threaded linear long-axis motor 2 is connected to the slider 5 through a hexagonal nut 18. The hexagonal nut 18 is installed inside the slider 5. A slider cover 4 is provided above the slider 5 to fit the hexagonal nut 18 to prevent it from falling out. Figure 2 The rotational motion of the threaded linear long-axis motor 2 is converted into the vertical up-and-down motion of the slider 5 .

[0047] like Figure 3 As shown, the slider 5 reciprocates up and down on the base 3, and the trapezoidal connector 7 is fixed below the base 3 by a positioning pin. The slider 5 is connected to the driving long connecting rod 6 by a pin, and the other side of the driving long connecting rod 6 is connected to one end of the triangular member 8, and the top of the triangular member 8 is connected to the trapezoidal connector 7 and the first finger joint connecting rod 10 by a screw. The first finger joint connecting rod 10 is fixed to the finger pulp 9 by a positioning pin. The other end of the triangular member 8 is fixed to one side of the first finger key 11 by a screw. The other side of the first finger key 11 is fixed to one side of the second finger pulp 12 by a positioning pin, and the middle hole of the second finger pulp 12 is fixed to the middle hole of the first finger joint connecting rod 10 by a positioning pin. The other side of the first finger joint connecting rod 10 is connected to the third finger joint connecting rod 13 by a positioning pin. The other side of the third finger joint connecting rod 13 is connected to a positioning hole of the finger tip connecting rod 14 by a positioning pin. The other side of the second finger pulp 12 is connected to another positioning hole of the finger tip connecting rod 14 by a positioning pin. The surface of the fingertip link 14 is embedded with a fingertip base 151, which is the first layer of the fingertip member 15. Figure 5As shown, the second layer of the fingertip member 15 is provided with a magnetic tactile sensor 152, the third layer is an isolation bracket 153, and the fourth layer is a fingertip receptor unit, which is a mixed magnetic powder silica gel body. The fingertip receptor unit is made by AB silica gel reverse molding.

[0048] As a preferred embodiment, a finger pulp 9 is fixedly connected to the first knuckle link 10, and the finger pulp 9 is used for single-finger grasping. When the finger is bent to a certain degree, the finger pulp 9 will contact the trapezoidal connecting member 7, thereby fixing the first knuckle link 10 of the finger, and the triangular member 8 continues to drive the first finger key 11, the second finger pulp 12, the third knuckle link 13, the fingertip link 14 and the fingertip member 15 to further rotate, thereby realizing single-finger bending.

[0049] like Figure 2 As shown, when the threaded linear long-axis motor 2 drives the slider 5 to translate backward on the base 3, the slider 5 drives the driving long connecting rod 6 to move backward, and the triangular member 8 rotates counterclockwise around the connection between the triangular member 8 and the trapezoidal connecting member 7 under the pull of the driving long connecting rod 6. At this time, the connection between the triangular member 8 and the driving long connecting rod 6 moves backward, and the connection between the triangular member 8 and the first finger key 11 moves forward. The first finger key 11 moves forward under the push of the triangular member 8, and the finger pulp 9, the first knuckle connecting rod 10, the first finger key 11, the second finger pulp 12, the third knuckle connecting rod 13, the fingertip connecting rod 14, and the fingertip member 15 bend inward as a whole. When the bending movement reaches a certain extent, the finger pulp 9 will contact the trapezoidal connecting member 7, thereby limiting the further movement of the first knuckle connecting rod 10 of the finger, and the triangular member 8 continues to push the first finger key 11, and the first finger key 11 drives the second finger pulp 12, the third knuckle connecting rod 13, the fingertip connecting rod 14 and the fingertip member 15 to further rotate around the connection between the second finger pulp 12 and the first knuckle connecting rod 10, thereby realizing single finger bending.

[0050] like Figure 4 As shown, a microprocessor is provided at the bottom of the fixed flange 1, which is used to control the threaded linear long-axis motor 2 to drive multiple joints and adjust the movement of each joint. An AB two-phase Hall encoder 17 is mounted above the threaded linear long-axis motor 2 for accurate reading of the manipulator posture. The microprocessor is connected to the voltage regulator module (PW2126, TPS54302), indicator light, reset circuit, crystal oscillator circuit, programming interface, motor drive module (TB6612FNG) through the IO port, connected to the three-way magnetic tactile sensor 152 through the IIC bus, and connected to the Flash memory (AT45DB64) through the SPI. The microprocessor generates three PWM waves through the timer, and the PWM waves control the threaded linear long-axis motor 2 to achieve adjustable speed rotation after passing through the motor drive module.

[0051] As mentioned above, the fingertip 15 is embedded with a magnetic tactile sensor 152. The magnetic tactile sensor 152 is composed of four ultra-small digital three-dimensional Hall sensors MLX90393 and an FPC interface. It senses force or contact through the magnetic field changes of the fingertip sensor unit, and is usually composed of a magnetic field sensing element (such as a Hall effect sensor) and an embedded magnetic source. When force or contact is applied from the outside, the position of the magnetic source or the magnetic field strength changes, and the magnetic field sensor converts it into an electrical signal, which provides information on the magnitude and direction of the force after processing.

[0052] In particular, the present invention also includes a grasping perception and recognition algorithm equipped by the manipulator, which, through the coordinated work of the microprocessor and the magnetic tactile sensor 152, can adjust the grasping force in real time according to the slip detection data, thereby adapting to objects of different shapes and sizes and effectively grasping and identifying underwater objects.

[0053] The grasping strategy proposed in the present invention can detect slip in real time through the magnetic tactile sensor 152, and adaptively adjust the grasping force according to the slip data. When the sensor detects that the object has a slight slip at the fingertips, the system gradually increases the grasping force through feedback control to prevent the object from falling off. The slip is sensed by monitoring the magnetic field changes and pressure fluctuations, and the increment and rate of the grasping force are adjusted according to the degree of slip to ensure that the grasping force is moderate and does not damage the object. In addition, the friction coefficient estimation is used to further optimize the force adjustment to make the grasping process more stable. Through this closed-loop feedback mechanism, the robot can flexibly respond to the grasping needs of different objects in an underwater environment. The following steps are included:

[0054] S1. The magnetic tactile sensor 152 is used to monitor the change of friction between the fingertips and the object in real time during the grasping process. The friction calculation formula is:

[0055] F f =μ·F n (1)

[0056] Where μ is the friction coefficient (depending on the contact state between the surface material of the object and the tactile sensor), F n is the normal force (the pressure exerted by the fingertip on the object).

[0057] S2. When the system detects a change in friction, it determines whether the change in friction exceeds the slip detection threshold. The slip detection threshold can be set using the following formula:

[0058] S th =α·μ·F n (2)

[0059] Among them, S th is the slip threshold, and α is a preset constant, usually 1.1 to 1.5, which is used to set the safety margin.

[0060] S3. When the friction force changes beyond the slip detection threshold, the gripping force adjustment amount of the manipulator device is calculated by combining the target gripping force and using the gripping force adjustment formula. The gripping force adjustment can be performed based on the slip detection feedback:

[0061] F g (t) = F g (t-1)+k·(F target -F g (t-1)) (3)

[0062] Among them, F g (t) is the grasping force F at the current moment g (t-1) is the grasping force at the previous moment, F target is the target grasping force (based on slip detection and friction estimation), and k is the adjustment rate constant (between 0 and 1, controlling the increment).

[0063] S4, controlling the threaded linear long-axis motor 2 to adjust the friction force until the friction force change does not exceed the slip detection threshold.

[0064] The present invention also relates to a target recognition method, which controls the robot to keep its grasping angle unchanged, uses a deep learning algorithm to perform target recognition detection, and uses a magnetic tactile sensor 15-2 to collect contact force data during the grasping process in real time, and uses this as a basis for identifying and classifying the target object. The system uses a deep learning model to analyze sensor data and extract the features required for the object when grasping, such as the required grasping force and the physical properties of the object. By evaluating the grasping force applied to different objects, the system can infer the shape, hardness, etc. of the object, thereby achieving accurate recognition of the target. The following steps are included:

[0065] S1. Extracting a feature vector from the contact force data collected by the magnetic tactile sensor 152. The feature vector X extracted by the data processing module is composed of the following contents:

[0066] X=[F g ,ΔF g ,T,V,…] (4)

[0067] Where, ΔF g is the rate of change of grasping force, T is the contact time, and V is the grasping speed;

[0068] S2. The convolutional layer output is obtained from the feature vector through the convolutional neural network (CNN) deep learning model. The feature extraction process can be expressed as:

[0069] Z=σ(W*X+b) (5)

[0070] Where Z is the output of the convolution layer, W is the convolution kernel, b is the bias, and σ is the activation function;

[0071] S3. Downsample the output of the convolutional layer to reduce the data dimension and the amount of calculation, and obtain the output of the pooling operation. The output is expressed as:

[0072] P=pool(Z) (6)

[0073] Among them, pool is a pooling operation (such as maximum pooling or average pooling);

[0074] S4. Map the extracted features to the output layer through a series of fully connected layers. The number of neurons in the output layer corresponds to the number of categories of the target objects. The output calculation can be expressed as:

[0075] O=σ(W f ·P+b f ) (7)

[0076] Where W f is the weight matrix of the fully connected layer, b f is the bias, O is the final output;

[0077] S5. Activate the output to generate the classification result C of the target object, usually using the Softmax function:

[0078]

[0079] Among them, C i is the probability of the target category, k is the total number of categories, O i is the value of the i-th neuron in the output layer.

[0080] The specific implementation mode of the present invention is described below by way of examples:

[0081] Example 1

[0082] like Figure 1-Figure 5 As shown, this embodiment provides an embodied sensing, multi-mode flexible configuration bionic multi-finger underwater manipulator, including a fixed flange 1 and a three-finger manipulator arranged on the side of the fixed flange 1. The manipulator adopts an under-actuated design method and is a seven-link mechanism structure, including a triangular component 8, and the first finger joint connecting rod 10 is fixedly connected to the finger pulp 9. The surface of the fingertip connecting rod 14 is embedded in the fingertip base 151 of the fingertip member 15, and the fingertip base 151 is the first layer of the fingertip member 15. The second layer of the fingertip member 15 is placed with a magnetic tactile sensor 152, the third layer is an isolation bracket 153, and the fourth layer is a fingertip receptor unit. The fingertip receptor unit is a mixed magnetic powder silica gel body, which is made by AB silica gel reverse molding. The threaded linear long-axis motor 2 is fixed to the motor compartment 16 on the side of the fixed flange 1, and the motor compartment 16 is connected using the mortise and tenon principle to seal the threaded linear long-axis motor 2.

[0083] The control system of the manipulator is based on a microprocessor, which is installed in a waterproof and sealed cabin of various underwater robots or underwater equipment carrying the manipulator to control the movement of the manipulator. Figure 6 As shown, the microprocessor uses STM32F103C8T6 as the main control chip, equipped with a power supply voltage regulator module (TPS54302), a crystal oscillator circuit (8MHZ) and a reset circuit (power-on reset), uses the SWD interface as a programmer, and connects to the motor driver module (TB6612FNG) through the IO port. The motor driver module generates PWM to drive the threaded linear long-axis motor 2 and the threaded linear long-axis motors corresponding to the second and third mechanical fingers. The main control chip is connected to the flash memory (AT45DB641E) through the SPI bus interface and the magnetic tactile sensor 15-2 (MLX90393) through the IIC bus interface. In addition, through address selection, the magnetic tactile sensor 152 corresponding to the second and third mechanical fingers can be read respectively. In addition, the control system displays the device status information through the LED indicator.

[0084] Example 2

[0085] On the basis of Example 1, this example further tests the three-finger collaborative grasping ability of the underwater bionic manipulator. The underwater bionic three-finger manipulator designed in this study adopts a precise seven-bar linkage mechanism, which can grasp a variety of objects by controlling the movement angles of three mechanical fingers separately and simultaneously. The manipulator can intelligently adjust the grasping force and grasping angle according to the characteristics of the target object (such as shape and size). For example, when facing milk bottles, small metal balls (such as ping-pong balls), rubber jellyfish, and rubber shrimps, the manipulator can adjust the grasping angles of the three mechanical fingers according to the characteristics of the objects to ensure that the grasping force is evenly distributed and avoid objects from slipping due to concentrated force.

[0086] Example 3

[0087] like Figure 1 and Figure 2 As shown, based on Example 1 or Example 2, the single-finger design of the manipulator provides flexible soil-moving and grasping functions for hidden objects in a soft sand environment. The manipulator embeds the magnetic tactile sensor 152 shown in the fingertips to sense the physical properties of the sand in real time. During operation, the manipulator first uses the fingertips to move the soil in order to clear the layer of sand covering the target object. For example, when looking for a small ceramic vase buried in fine sand, the manipulator can easily move the sand and obtain the hardness information of the sand with the help of the built-in tactile sensor 152 to optimize the subsequent grasping strategy. In addition, the precise angle control of the multi-link ensures that the manipulator can grasp slender or small spherical objects, such as small agate blocks or ceramic blocks, with a single finger.

[0088] Example 4

[0089] like Figure 7 As shown, on the basis of Example 1 or Example 2 or Example 3, in order to ensure the stability of the object during the grasping process, the magnetic tactile sensor 152 equipped by the manipulator can monitor the slippage that may occur during the grasping process in real time. When grasping heavy objects (such as metal blocks), after the system detects the change in the friction force applied to the object, the control algorithm will automatically calculate and gradually increase the grasping force to prevent the object from slipping. In a specific implementation, the manipulator dynamically adjusts the grasping force according to the change in friction, contact time and dynamic characteristics of the object. For example, when slippage of an object is detected, the manipulator will optimize the grasping force in real time through a closed-loop feedback mechanism to ensure flexible response to different grasping needs in an underwater environment.

[0090] Example 5

[0091] like Figure 8 As shown, based on Example 1 or Example 2 or Example 3, the manipulator classifies objects through contact force data collected in real time. During the grasping process, the manipulator analyzes and extracts the characteristics of different objects, such as small rubber underwater creatures, small metal balls, milk bottles, small agate blocks, pebbles, keys, gel pens, etc. Through the deep learning model, the system can classify and identify the target objects according to feature vectors such as changes in grasping force, contact time, and grasping speed. In the implementation process, a convolutional neural network (CNN) is used to process the collected multidimensional data to accurately infer the shape, hardness and other physical properties of the object. This algorithm not only improves the manipulator's ability to recognize a variety of underwater objects, but also provides data support for subsequent autonomous decision-making, which helps to efficiently complete tasks in complex underwater environments.

[0092] In the specific contents of the above-mentioned specific implementation methods, the various technical features can be combined in any non-contradictory manner. In order to make the description concise, not all possible combinations of the above-mentioned technical features are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0093] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A bionic multi-finger underwater manipulator device, characterized in that: It comprises a fixed flange (1) and a plurality of mechanical fingers connected to the side of the fixed flange (1); The mechanical finger comprises a base (3), a trapezoidal connecting member (7), a triangular member (8), a fingertip member (15) and a slider (5); two corners of the triangular member (8) are respectively connected to the fingertip member (15) and the slider (5) through a connecting rod mechanism; the third corner of the triangular member (8) is connected to the base (3) through the trapezoidal connecting member (7) and the connection point is a fixed rotation point; the slider (5) is used to push one corner of the triangular member (8) to rotate around the fixed rotation point through linear motion, thereby driving the fingertip member (15) to swing; The slider (5) is driven by a threaded linear long-axis motor (2); a magnetic tactile sensor (152) is provided in the fingertip member (15); the magnetic tactile sensor (152) receives fingertip tactile information via a fingertip receptor unit; the fingertip receptor unit is a magnetic powder-silicone mixture.

2. A bionic multi-finger underwater manipulator device according to claim 1, characterized in that: The mechanical finger is a seven-bar linkage, the base (3) is fixedly connected to the fixed flange (1), and the seven-bar linkage comprises a driving long link (6), a triangular piece (8), a first finger joint link (10), a first finger key (11), a second finger web (12), a third finger joint link (13) and a fingertip link (14); one end of the triangular piece (8) is connected to the slider (5) through the driving long link (6), and the other end is connected to the first finger key (11), and the middle is connected to the base (3) through a trapezoidal connecting piece (7); One end of the first finger key (11) is connected to the triangular piece (8), and the other end is connected to the second finger pulp (12); one end of the second finger pulp (12) is connected to the first finger key (11), and the other end is connected to the fingertip link (14), and the middle is connected to the first finger joint link (10); one end of the first finger joint link (10) is connected to the middle of the triangular piece (8), and one end is connected to the third finger joint link (13), and the middle end is connected to the second finger pulp (12); one end of the third finger joint link (13) is connected to the fingertip link (14), and the other end is connected to the end of the first finger joint link (10); the fingertip link (14) connects the second finger pulp (12) and the third finger joint link (13); the slider (5) is located on the base (3) and has the freedom to reciprocate along the axial direction of the mechanical finger, and the fingertip piece (15) is embedded in the surface of the fingertip link (14).

3. A bionic multi-finger underwater manipulator device according to claim 2, characterized in that: The mechanical finger also includes a finger pulp (9), and the finger pulp (9) is fixedly connected to the first finger joint connecting rod (10).

4. The bionic multi-finger underwater manipulator device according to claim 2, characterized in that: The fingertip piece (15) comprises a fingertip base (151), a magnetic tactile sensor (152) and an isolation bracket (153) which are connected in sequence; a notch is provided on the fingertip base (151); and the fingertip piece (15) is embedded in the fingertip connecting rod (14) through the notch on the fingertip base (151).

5. The bionic multi-finger underwater manipulator device according to claim 4, characterized in that: The isolation bracket (153) is provided with a groove, the fingertip sensor unit is located in the groove, and the isolation bracket (153) separates the magnetic tactile sensor (152) and the fingertip sensor unit.

6. The bionic multi-finger underwater manipulator device according to claim 1, characterized in that: A motor compartment (16) is provided on the side of the fixed flange (1), the threaded linear long-axis motor (2) is fixed in the motor compartment (16), and the motor shaft extends from the motor compartment (16) and is connected to the slider (5), and an AB two-phase Hall encoder (17) is provided at the rear of the threaded linear long-axis motor (2).

7. The bionic multi-finger underwater manipulator device according to claim 6, characterized in that: The motor compartment (16) comprises a motor seat and a housing, the motor seat is fixedly connected to the fixed flange (1), the housing is connected to the motor seat via a mortise and tenon structure, the AB two-phase Hall encoder (17) is connected to the threaded linear long-axis motor (2), and the motor seat and the housing seal the AB two-phase Hall encoder (17) and the threaded linear long-axis motor (2).

8. A control system for a bionic multi-finger underwater manipulator device according to any one of claims 1 to 7, characterized in that: It includes a microprocessor, a voltage stabilizing module, a crystal oscillator circuit, a reset circuit, a motor drive module, a Flash memory, a tactile sensor and an indicator light; the microprocessor uses STM32F103C8T6 as the main control chip, is equipped with a voltage stabilizing module, a crystal oscillator circuit and a reset circuit, and uses the SWD interface as a programmer, and is connected to the motor drive module through the IO port; the motor drive module generates PWM to drive the threaded linear long-axis motor 2 of each mechanical finger; the microprocessor is also connected to the flash memory through the SPI bus interface, and is connected or reads the magnetic tactile sensor 152 corresponding to each mechanical finger through the IIC bus interface and address selection; the control system displays status information through the indicator light.

9. A grasping strategy of a bionic multi-finger underwater manipulator device according to any one of claims 1 to 7, characterized in that: The following steps are involved: Real-time monitoring of the change in friction between the fingertip and the object during grasping by means of a magnetic tactile sensor (152); When the system detects a change in friction force, it determines whether the change in friction force exceeds a slip detection threshold; When the friction force change exceeds the slip detection threshold, the gripping force adjustment amount of the manipulator device is calculated by combining the target gripping force and using the gripping force adjustment formula; The threaded linear long-axis motor (2) is controlled to adjust the friction force until the friction force change does not exceed the slip detection threshold.

10. A target recognition method for a bionic multi-finger underwater manipulator device according to any one of claims 1 to 7, characterized in that: The following steps are involved: extracting a feature vector from contact force data collected by a magnetic tactile sensor (152); Obtain the convolutional layer output from the feature vector through a convolutional neural network deep learning model; Downsample the output of the convolutional layer to get the output of the pooling operation; The extracted features are mapped to the output layer through a series of fully connected layers. The number of neurons in the output layer corresponds to the number of categories of the target objects. Activate the output and generate the classification result C of the target object.

Images