Soft micro-robot multi-mode magnetic driving and motion optimization method oriented to natural orifice environment

Abstract

The invention discloses a soft micro-robot multi-mode magnetic driving and motion optimization method oriented to a natural orifice environment, and relates to the technical field of micro-robot control. The method comprises the steps that multi-source information of a natural cavity environment where the soft micro-robot is located is obtained, wherein the multi-source information at least comprises parameters reflecting the narrowing degree, the bending angle or the wrinkle distribution of the cavity; extracting a topographic feature vector representing the environment; based on a preset environment feature weight, a form-motion weight matrix and the topographic feature vector, an optimal robot form primitive and a corresponding motion mode are determined through a dynamic matching decision-making mechanism, and the motion mode at least comprises one of curling rolling, bending inrush or stretching crawling; according to the determined motion mode, a preset driving magnetic field corresponding to the motion mode is selected and applied, the driving magnetic field at least comprises one of a rotating magnetic field, an alternating pulse magnetic field or an oscillating magnetic field, and the driving magnetic field can be generated through a gradient-torque decoupling modulation strategy. According to the method, the optimal motion mode can be adaptively selected according to the complex and changeable cavity environment, and the motion efficiency, stability and operation performance of the soft micro-robot in the natural cavity environment are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of micro-robots, and particularly to soft micro-robots for navigation, detection or task execution in complex biological channels (such as digestive tracts, blood vessels, etc.). Specifically, it relates to a multi-modal magnetic drive and motion optimization method for soft micro-robots facing the natural channel environment. Background Art

[0002] With the continuous progress of the medical and engineering fields, micro-robots, especially magnetically-driven soft micro-robots, are increasingly widely used in biomedicine. Especially in complex biological channels, soft micro-robots can adapt to irregular environments such as narrow, curved, and wrinkled ones due to their flexibility and deformability, and can complete tasks that traditional mechanical robots cannot achieve. However, existing magnetically-driven soft micro-robots still face many challenges in multi-modal motion capabilities and environmental adaptability, especially in the application of complex biological channel environments. The existing technologies mainly focus on a single magnetic field driving mode, and the motion modes of micro-robots are relatively single, making it difficult to cope with the dynamically changing channel environment. In addition, existing micro-robots have hysteresis in environmental response, and their motion capabilities, accuracy, and stability are affected by various factors such as mechanical effects and environmental geometric characteristics, severely restricting their application effects in complex environments. Therefore, how to break through the bottleneck of existing technologies and design soft micro-robots with multi-modal motion and adaptive capabilities has become the key to solving these problems. Summary of the Invention

[0003] The present invention provides a multi-modal magnetic drive and motion optimization method for soft micro-robots facing the natural channel environment, and the method includes the following steps:

[0004] (1) Obtain multi-source information of the natural channel environment where the soft micro-robot is located, and the multi-source information reflects the geometric features and / or mechanical properties of the channel;

[0005] (2) Based on the multi-source information, extract the terrain feature vector characterizing the natural channel environment;

[0006] (3) Based on the preset environmental feature weights, morphology-motion weight matrix, and the terrain feature vector, through a dynamic matching decision mechanism, determine the optimal target form primitive and the corresponding optimal target motion mode of the soft micro-robot;

[0007] (4) According to the determined optimal target motion mode, select and apply the corresponding driving magnetic field that can drive this mode. Preferably, the driving magnetic field is generated by a gradient-torque decoupling modulation strategy.

[0008] In a preferred embodiment, the multi-source information can come from micro-sensors integrated on the micro-robot body (such as pressure sensors, flexible strain sensors) or external observation devices (such as a micro-camera combined with an image processing algorithm), and is used to obtain parameters reflecting the degree of channel stenosis, bending angle, fold distribution, etc.

[0009] In a preferred embodiment, the terrain feature vector S x is obtained by normalizing and encoding the original multi-source information. For example, it can be defined as a vector containing the following components:

[0010]

[0011] where P is the measured value of the pressure sensor, and P max is the preset stenosis threshold; θ is the bending angle extracted by image processing, and θ th is the critical value of bending density; N fold is the number of folds extracted by image processing, and N max is the maximum fold density. This vector intuitively characterizes the environmental features through the numerical range (0-1), providing a quantitative input for subsequent decision-making.

[0012] In a preferred embodiment, the dynamic matching decision-making mechanism is based on the "environment-morphology-motion" triple relationship. Define the environmental feature weight vector W = [w1, w2, w3] to characterize the contribution degrees of stenosis, bending, and folds to the decision-making, and define the morphology-motion weight matrix Q (with dimensions of m×n, where m is the number of morphological primitives and n is the number of motion modes), and its element q ij represents the matching degree between morphology i and motion mode j. The matching decision output Y (representing the index of the optimal combination) is calculated by the following formula:

[0013] Y = argmax(S x ·W·Q),

[0014] where "·" represents an appropriate matrix or vector operation (for example, first calculate the weighted feature S' x = S x ⊙W, where ⊙ represents element-wise multiplication or a defined weighting method, and then calculate the matrix multiplication of S' x ·Q), and the argmax function returns the index corresponding to the largest element in the result vector.

[0015] In a preferred embodiment, to drive different motion modes, a specific driving magnetic field is designed:

[0016] Coiling and rolling mode: Apply a rotating magnetic field B r (t), and its form can be:

[0017] B rB(t) = B m [cos(2πft r u + sin(2πft r v)],

[0018] where B m is the magnetic field strength, f r is the rotational magnetic field frequency, and u and v are two orthogonal basis vectors in the rotation plane.

[0019] Bending surge mode: Apply an alternating pulsed magnetic field B s (t), such as a piecewise-defined square wave or sine pulse:

[0020]

[0021] where T is the waveform period and n0 is the unit vector in the magnetic field direction.

[0022] Stretching crawl mode: Apply an oscillating magnetic field B s (t) in a specific direction, which can be in the form of:

[0023] B w (t) = B m [k w m w + sin(2πft w t)n w ,

[0024] where f w is the oscillating magnetic field frequency, k w is the coefficient to determine the oscillation amplitude angle, n w is the unit vector in the central symmetry axis direction of the oscillating magnetic field sweeping plane, and m w is the unit vector in the plane orthogonal to n w .

[0025] In a preferred embodiment, the gradient-torque decoupling modulation strategy aims to independently generate the required magnetic field gradient (for applying force) and magnetic field direction / strength (for applying torque) by precisely controlling the current magnitude, frequency, and phase of multiple electromagnetic coils, thereby achieving decoupled control of the force and torque of the micro-machine to drive complex motion modes. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a schematic diagram of the multi-modal motion design of a soft micro-robot according to an embodiment of the present invention, showing three motion modes of curling and rolling, bending surge, and stretching crawl and the corresponding driving magnetic field types.

[0027] Figure 2It is a schematic diagram of the "environment - morphology - motion" dynamic matching mechanism process according to an embodiment of the present invention, showing the process from environmental feature extraction to motion mode selection. Detailed implementation manners

[0028] The technical solutions of the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and do not limit the scope of the present invention.

[0029] Embodiment 1: Adaptive motion mode selection based on dynamic matching

[0030] In this embodiment, the soft micro - robot moves in a simulated natural cavity environment.

[0031] Step S1: Acquisition and feature extraction of environmental information

[0032] The robot obtains environmental information through sensors and / or external observations. Assume that at a certain moment t, the information is obtained and the terrain feature vector is calculated as:

[0033] S x =[0.8, 0.67, 0.5].

[0034] It indicates that the environment is relatively narrow (0.8), moderately curved (0.67), and moderately wrinkled (0.5).

[0035] Step S2: Dynamic matching decision

[0036] Assume that the preset environmental feature weight vector W = [0.4, 0.4, 0.2] (indicating that narrowness and curvature have a greater impact on the decision, and wrinkling has a smaller impact).

[0037] Assume that the preset morphology - motion weight matrix Q (3x3, rows for morphology: closed, bent, stretched; columns for motion: rolling, surging, crawling) is:

[0038]

[0039] This matrix indicates that the closed morphology is most suitable for rolling, the bent morphology is most suitable for surging, and the stretched morphology is most suitable for crawling.

[0040] Calculate the matching score vector:

[0041] Score = S x ·W·Q≈[0.378, 0.328, 0.229].

[0042] According to the formula Y = argmax(Score). Under the calculation results of this example, the first element is the largest, and the corresponding index is (morphology 1, motion 1), that is, the combination of "closed morphology" and "curling and rolling" mode.

[0043] Step S3: Application of driving magnetic field

[0044] According to the decision result, a rotating magnetic field that drives the "curling and rolling" mode is selected. The system transmits the corresponding driving parameters back to the three-dimensional programmable magnetic field generation system. The magnetic field system generates a rotating magnetic field according to the parameters, driving the robot to roll forward in a closed form.

Claims

1. A multi-modal magnetic drive and motion optimization method for a soft micro-robot in a natural cavity environment, characterized in that It includes the following steps: Obtain multi-source information of the natural cavity environment where the soft micro-robot is located; Based on the multi-source information, extract a terrain feature vector characterizing the natural cavity environment; Based on the preset weight information and the terrain feature vector, through a dynamic matching decision mechanism, determine the target form primitive and the corresponding target motion mode of the soft micro-robot; According to the determined target motion mode, select and apply a corresponding driving magnetic field to drive the soft micro-robot to move according to the target form primitive and target motion mode.

2. The method according to claim 1, wherein The multi-source information includes at least one or more of the following information: the pressure value obtained by a pressure sensor, the bending angle obtained by an image acquisition device, the number or distribution information of folds obtained and processed by an image acquisition device.

3. The method according to claim 1 or 2, characterized in that The step of extracting the terrain feature vector characterizing the natural cavity environment includes: normalizing the obtained multi-source information and encoding it as the terrain feature vector.

4. The method according to claim 1, characterized in that, The preset weight information includes an environmental feature weight vector and a form-motion weight matrix; the dynamic matching decision mechanism includes: performing matrix operations on the terrain feature vector, the environmental feature weight vector, and the form-motion weight matrix, and determining the index of the optimal target form primitive and the corresponding target motion mode combination through the argmax function.

5. The method according to claim 1 or 4, characterized in that The target form primitive is selected from at least one of: closed, bent, and unfolded.

6. The method according to claim 1 or 4, characterized in that, The target motion mode is selected from at least one of: curling and rolling, bending and surging, and unfolding and crawling.

7. The method according to claim 1, characterized in that, The driving magnetic field is selected from at least one of: a rotating magnetic field, an alternating pulse magnetic field, and an oscillating magnetic field; wherein, the rotating magnetic field is used to drive the curling and rolling mode; the alternating pulse magnetic field is used to drive the bending and surging mode; the oscillating magnetic field is used to drive the unfolding and crawling mode.

8. The method according to claim 1 or 7, characterized in that The driving magnetic field is generated and controlled by a gradient-torque decoupling modulation strategy.

Images