Microstructure automatic assembly system and method

Abstract

The present application relates to a kind of microstructure automatic assembly system, comprising: control terminal module, imaging identification module, acoustic driving module, wherein: the control terminal module is used to build the three-dimensional assembly layout model of target microstructure, and according to the initial position information of microstructure obtained by imaging identification module, the carrying target position and carrying path of microstructure are calculated, while generating acoustic field control parameter;The imaging identification module is used to collect the real-time image information of microstructure, identify its position, shape and state in three-dimensional space, the image recognition result is as the input of control logic, and is used for feedback control;The acoustic driving module is used to receive acoustic field parameter, and drive ultrasonic transducer array, generate three-dimensional acoustic field, realize the three-dimensional non-contact capture and carrying of microstructure.The present application also relates to a kind of microstructure automatic assembly method.The present application can improve the flexibility, precision and stability of organization construction.

Description

Technical Field

[0001] This invention relates to an automated assembly system and method for microstructures. Background Technology

[0002] In fields such as biomanufacturing, disease modeling, and drug screening, various microstructures (such as cells, organoids, and microspheres) have played a crucial role. To construct functional models that more closely resemble the structures of real tissues or systems, researchers are increasingly focusing on the orderly and precise assembly of multiple microstructures in space. Achieving a stable and controllable three-dimensional assembly process is of great significance for reconstructing complex structures and enhancing system function. Three-dimensional acoustic manipulation technology utilizes ultrasound to form standing waves or acoustic radiation force fields, thereby enabling non-contact manipulation of microparticles or biological tissues. This method is characterized by strong controllability, wide adaptability, and gentleness towards biological samples, and has shown significant application potential in cell manipulation, particle sequencing, and tissue engineering.

[0003] Currently, the three-dimensional construction of microstructures largely relies on bio-3D printing technology, which forms structures by depositing bio-ink and stacking layers one by one. While this can achieve a certain degree of spatial arrangement, it suffers from problems such as high shear force, poor adaptability to microstructure particles, and limited assembly precision. Furthermore, it is difficult to flexibly identify and dynamically reconstruct the generated microstructures, limiting its applicability to complex tissue configurations. In other words: Existing microstructure construction technologies primarily rely on microscale 3D printing, which, while enabling the creation of structured tissues, suffer from several key limitations. The construction process depends on specific bio-ink materials, exhibiting poor adaptability to pre-formed microstructure particles, and shear stress during printing can affect the integrity and activity of the microstructures. Furthermore, this technology struggles to achieve precise positioning and three-dimensional spatial reconstruction of multiple microstructures and lacks the ability to identify and adaptively adjust the actual positions of microstructures in real time. Summary of the Invention

[0004] In view of this, it is necessary to provide an automatic assembly system and method for microstructures, which can realize non-contact spatial identification, path planning and three-dimensional assembly of multiple microstructures in three-dimensional space, thereby improving the flexibility, accuracy and stability of tissue construction.

[0005] This invention provides an automated assembly system for microstructures. The system includes a control terminal module, an imaging recognition module, and an acoustic drive module. The control terminal module constructs a three-dimensional assembly layout model of the target microstructure and calculates the transport target position and transport path of the microstructure based on the initial position information of the microstructure obtained by the imaging recognition module, while simultaneously generating acoustic field control parameters. The imaging recognition module acquires real-time image information of the microstructure, identifies its position, shape, and state in three-dimensional space, uses the image recognition results as input to the control logic, and employs feedback control. The acoustic drive module receives the acoustic field parameters and drives an ultrasonic transducer array to generate a three-dimensional sound field, enabling three-dimensional non-contact capture and transport of the microstructure.

[0006] Specifically, the control terminal module is used for: Perform microstructure assembly structural modeling; Generate a transport path and dynamically adjust it based on feedback from the imaging recognition module; Based on the spatial layout of the target and the required sound field shape, control parameters that match the acoustic drive module are generated.

[0007] The aforementioned microstructure assembly structure modeling includes: Before performing the microstructure assembly operation, a physical 3D model of the expected assembly form is completed in the control terminal module, and the spatial information of the assembly is obtained from the expected microstructure model. The spatial information of the assembly mainly includes: the number of microstructures and the three-dimensional spatial coordinates (x, y, z) of each microstructure. After modeling is completed, the expected coordinate information of the microstructure is used as the target point for acoustic manipulation, and the sound field of the target point and the corresponding emission delay parameters are calculated.

[0008] The generated transport path includes: The spatial coordinates of the microstructure identified in the image are (x1, y1, z1), and the target point of one of the microstructures in the assembly is located at (x2, y2, z2). First, using the above two positions as the starting and ending coordinates, n movement steps are calculated according to the acoustic control movement path. The number of steps n is calculated based on the distance between the two points / step length. After obtaining the focal position corresponding to each step, the ultrasonic phase information during ultrasonic emission is calculated using the time inversion IB algorithm. Then, the phase relationship of the ultrasound is converted into the delay parameters corresponding to each channel, thereby calculating n sets of delay parameter data for m channels; m is the total number of emission channels.

[0009] Specifically, the imaging recognition module is used for: Acquire real-time images; Image recognition algorithms are used to identify and detect microstructures; It provides real-time feedback on positional deviations and participates in path correction.

[0010] The method of using image recognition algorithms to identify and detect microstructures includes: Image recognition algorithms are used to extract microstructure target information in the imaging field of view, including its current center position coordinates and edge morphology; the recognition model is pre-trained using a labeled dataset for image recognition training; after training, the data is imported and run for real-time image recognition and inference.

[0011] The real-time feedback position deviation, which participates in path correction, includes: The system continuously acquires images of the current field of view, updates the actual position of detected microstructures in real time, and compares them with the planned path. If an offset occurs, an error feedback signal is generated to indicate to the control terminal module that the current sample position has shifted. The control terminal module then adjusts the sound field parameters to achieve closed-loop control of the movement path.

[0012] Specifically, the acoustic driving module is used for: Based on acoustic control parameters, an excitation signal is generated and the ultrasonic transducer is controlled. Provides a two-way communication interface with the control terminal module; The generated excitation signal is amplified and output matched.

[0013] The process of generating an excitation signal and controlling the ultrasonic transducer based on acoustic control parameters includes: Based on the acoustic control parameters sent by the control terminal module, a corresponding driving voltage pulse is generated; based on the acoustic control parameters, the acoustic drive module controls the ultrasonic transducer to realize the position movement of the sound field focal point and the switching of the sound field shape.

[0014] This invention provides an automated assembly method for microstructures, the method comprising: Step S1: Construct a three-dimensional assembly layout model of the target microstructure, and calculate the transport target position and transport path of the microstructure based on the obtained initial position information of the microstructure, while generating sound field control parameters. Step S2: Acquire real-time image information of the microstructure, identify its position, shape and state in three-dimensional space; use the image recognition results as input to the control logic and for feedback control; Step S3: Receive sound field parameters and drive the ultrasonic transducer array to generate a three-dimensional sound field, thereby achieving three-dimensional non-contact capture and transport of the microstructure and completing the three-dimensional structure assembly.

[0015] This application enables non-contact manipulation and precise 3D assembly of multiple microstructures, avoiding the tissue stress and deformation caused by mechanical extrusion during 3D printing. Acoustic tweezers technology facilitates dynamic handling and positional calibration of particle-level microstructures, improving the spatial flexibility and structural accuracy of assembly. Combined with image recognition and automatic path planning, it can adapt to microstructures with different initial positions and morphologies, enhancing versatility and automation. Compared to 3D printing's reliance on bio-inks and fixed templates, this application is more suitable for the assembly and reconstruction of heterogeneous microstructures. Attached Figure Description

[0016] Figure 1 This is a hardware architecture diagram of the automatic assembly system for microstructures of the present invention; Figure 2 This is a schematic diagram of the automatic assembly method for microstructures according to an embodiment of the present invention; Figure 3 This is a schematic diagram of non-contact manipulation of microstructures according to an embodiment of the present invention; Figure 4 This is a schematic diagram of inter-channel delay switching according to an embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the movement of the sound field focal point in an embodiment of the present invention. Figure 6 This is a schematic diagram of sound field shape switching according to an embodiment of the present invention; Figure 7 This is a flowchart of the automatic assembly method for microstructures according to the present invention. Detailed Implementation

[0017] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0018] See Figure 1 The diagram shown is a hardware architecture diagram of the automatic microstructure assembly system of the present invention. The system includes: a control terminal module, an imaging recognition module, and an acoustic drive module.

[0019] The control terminal module is used to construct a three-dimensional assembly layout model of the target microstructure, and calculate the transport target position and transport path of the microstructure based on the initial position information of the microstructure obtained by the imaging recognition module, while generating sound field control parameters. Specifically, the control terminal module is the core control terminal of the system, providing a UI control interface for setting and transmitting / receiving system parameters. (1) The control terminal module performs microstructure assembly structure modeling, supporting preset templates or custom layouts: Before assembling microstructures using the system, a physical 3D model of the expected assembly form is completed in the control terminal module, and the spatial information of the assembly is obtained from the expected microstructure model. The spatial information of the assembly mainly includes: the number of microstructures and the three-dimensional spatial coordinates (x, y, z) corresponding to each microstructure, with the unit of the spatial coordinates being the same as the scale unit of the current field of view.

[0020] After modeling is completed, the expected coordinates of the acquired microstructure are used as the target point for acoustic manipulation, and the sound field and corresponding emission delay parameters of the target point are calculated. The assembly modeling of the microstructure can use a preset assembly template or be customized, allowing for free setting of the assembly position and shape of the microstructure.

[0021] (2) The control terminal module generates the transport path and dynamically adjusts it based on feedback from the imaging recognition module: After obtaining the spatial coordinates of the moving target, the positions of the microstructures identified by the imaging recognition module are acquired in real time. For example... Figure 3 The spatial coordinates of the microstructure identified by the image are (x1, y1, z1), and the target point of one of the microstructures in the assembly is located at (x2, y2, z2). First, using these two positions as the starting and ending coordinates, n movement steps are calculated according to the acoustic manipulation movement path. The number of steps n is calculated based on the distance between the two points / step length. After obtaining the focal position corresponding to each step, the ultrasonic phase information during ultrasonic emission is calculated using the time-inversion IB algorithm. Then, the phase relationship of the ultrasound is converted into the delay parameters corresponding to each channel. Thus, n sets of delay parameter data for m channels (m is the total number of emission channels) are calculated. During pulse emission, the pulse delay of each step corresponding to all channels is shown in Table 1. The temporal relationship of the emission delay before and after pulse emission is as follows: Figure 4 As shown.

[0022] Table 1

[0023] Based on the real-time position updates of the manipulated microstructures by the imaging recognition module, the control path is corrected. If the currently manipulated microstructure cell deviates from the set control path, a corrected control path is calculated based on the current deviation position and the nearest point in the path. The control delay parameters of the corrected path are then sent down to control the microstructure back to the control path, and the movement continues along the path.

[0024] (3) The control terminal module generates control parameters that match the acoustic drive module based on the spatial layout of the target and the required sound field shape: The control terminal module has a sound field parameter generation function, responsible for generating control parameters that match the acoustic drive module based on the spatial layout of the target and the required sound field shape. This control terminal module, combined with the hardware characteristics of the acoustic drive module (such as the number of transducers, center frequency range, phase resolution, etc.), sets basic drive parameters such as center frequency (Fc), pulse repetition frequency (PRF), pulse duty cycle, and voltage excitation amplitude (V). Simultaneously, based on the preset target position of the microstructure to be controlled in three-dimensional space, it generates the relative delay time and phase adjustment parameters for each drive channel to construct the desired focused, standing wave, or rotating sound field shape. Furthermore, this control terminal module supports the selection of multiple sound field modes, such as single-point focusing, trajectory tracking, and multi-point synchronous control. Through parameter combinations, it flexibly generates sound field structures that meet different control requirements, achieving precise transport and three-dimensional positioning of the microstructure. All generated parameters are sent to the acoustic drive module in the form of control commands to achieve real-time sound field control.

[0025] The imaging recognition module is used to acquire real-time image information of microstructures and identify their position, shape, and state in three-dimensional space. The image recognition results are used as input to the control logic and for feedback control. Specifically, this includes: (1) The imaging recognition module acquires real-time images, including: confocal microscopy and inverted microscopy imaging. Using microscopic imaging devices, such as inverted microscopes and confocal imaging modules, images of the microstructure manipulation area are acquired in real time to obtain image information of the particles in the current experiment. The acquired real-time images are displayed in the UI interface of the control terminal module to facilitate observation of the movement of the microstructure.

[0026] (2) The imaging recognition module uses an image recognition algorithm to identify and detect microstructures: Image recognition algorithms, such as lightweight recognition models based on convolutional neural networks (CNN) or YOLO, are used to extract microstructure target information from the imaging field of view, including its current center coordinates and edge morphology. The recognition model is pre-trained using a labeled dataset to improve the accuracy of microstructure recognition. After training, the data is imported and run for real-time image recognition and inference. The spatial coordinate data of the recognized microstructures is standardized to provide the control terminal module with information for generating transport paths for the microstructures. In addition to spatial coordinate information, the detected microstructure cell diameter and edge morphology information are also output.

[0027] (3) The imaging recognition module feeds back the positional deviation to the layout planning module in real time to participate in path correction: During the acoustic manipulation and handling of the microstructure, the imaging recognition module continuously acquires images of the current field of view, updates the actual position of the detected microstructure in real time, and compares it with the planned path. If an offset occurs, an error feedback signal is generated to indicate to the control terminal module that the current sample position has been offset. The control terminal module then adjusts the parameters of the sound field to achieve closed-loop control of the movement path.

[0028] The acoustic drive module receives sound field parameters and drives the ultrasonic transducer array to generate a three-dimensional sound field, enabling three-dimensional non-contact capture and transport of microstructures. Specifically: (1) The acoustic drive module generates excitation signals and controls the ultrasonic transducer according to the acoustic control parameters: Based on acoustic manipulation parameters such as frequency, amplitude, phase, and pulse repetition frequency sent by the control terminal module, a corresponding driving voltage pulse is generated. The acoustic driving module supports independent or synchronous excitation of multiple array elements, ensuring the accuracy of the sound field in three-dimensional space. According to the acoustic manipulation parameters, the acoustic driving module controls the ultrasonic transducer to achieve the movement of the sound field focal point and the switching of the sound field morphology. Figure 5 , Figure 6 ), to be suitable for a variety of acoustic control scenarios.

[0029] (2) The acoustic drive module provides a bidirectional communication interface with the control terminal module: The acoustic drive module communicates via wired or wireless data transmission. The control terminal module can send adjustment parameters in real time and can also read back data to ensure the correctness of the acoustic control parameters.

[0030] (3) The acoustic drive module amplifies the generated excitation signal and performs output matching: The acoustic drive module amplifies the generated excitation signal to provide sufficient energy to excite the front-end transducer, while reducing reflections during transmission through impedance matching to minimize energy loss on the line.

[0031] See Figure 7 The diagram shown is a flowchart of a preferred embodiment of the automatic assembly method for microstructures of the present invention. Please refer to it as well. Figure 2 .

[0032] Step S1 involves constructing a 3D assembly layout model of the target microstructure and, based on the acquired initial position information of the microstructure, calculating the target location and transport path for the microstructure, while simultaneously generating acoustic field control parameters. Specifically, this includes: (1) Perform microstructure assembly structure modeling, supporting preset templates or custom layouts: Before assembling the microstructures, a physical 3D model of the expected assembly form is completed, and the spatial information of the assembly is obtained from the expected microstructure model. The spatial information of the assembly mainly includes: the number of microstructures and the three-dimensional spatial coordinates (x, y, z) corresponding to each microstructure. The unit of the spatial coordinates is the same as the scale unit of the current field of view.

[0033] After modeling is completed, the expected coordinates of the acquired microstructure are used as the target point for acoustic manipulation, and the sound field and corresponding emission delay parameters of the target point are calculated. The assembly modeling of the microstructure can use a preset assembly template or be customized, allowing for free setting of the assembly position and shape of the microstructure.

[0034] (2) Generate the transport path and dynamically adjust it based on feedback: After obtaining the spatial coordinates of the moving target, the positions of the identified microstructures are acquired in real time. For example... Figure 3 The spatial coordinates of the microstructure identified by the image are (x1, y1, z1), and the target point of one of the microstructures in the assembly is located at (x2, y2, z2). First, using these two positions as the starting and ending coordinates, n movement steps are calculated according to the acoustic manipulation movement path. The number of steps n is calculated based on the distance between the two points / step length. After obtaining the focal position corresponding to each step, the ultrasonic phase information during ultrasonic emission is calculated using the time-inversion IB algorithm. Then, the phase relationship of the ultrasound is converted into the delay parameters corresponding to each channel. Thus, n sets of delay parameter data for m channels (m is the total number of emission channels) are calculated. During pulse emission, the pulse delay of each step corresponding to all channels is shown in Table 1. The temporal relationship of the emission delay before and after pulse emission is as follows: Figure 4 As shown.

[0035] Table 1

[0036] Based on the real-time position updates of the manipulated microstructures, the control path is corrected. If the currently manipulated microstructure cell deviates from the set control path, a corrected control path is calculated based on the current deviation position and the nearest point in the path. The control delay parameters of the corrected path are then sent down to control the microstructure back to the control path, and it continues to move according to the path.

[0037] (3) Generate matching control parameters based on the spatial layout of the target and the required sound field shape: Based on the spatial layout of the target and the desired sound field shape, matching control parameters are generated. Combining the hardware characteristics of the acoustic drive module (such as the number of transducers, center frequency range, and phase resolution), basic drive parameters such as center frequency (Fc), pulse repetition frequency (PRF), pulse duty cycle, and voltage excitation amplitude (V) are set. Simultaneously, based on the preset target position of the microstructure to be controlled in three-dimensional space, relative delay time and phase adjustment parameters for each drive channel are generated to construct the desired focused, standing wave, or rotating sound field shape. Furthermore, multiple sound field modes are supported, such as single-point focusing, trajectory tracking, and multi-point synchronous control. By combining parameters, sound field structures that meet different control requirements are flexibly generated, enabling precise transport and three-dimensional positioning of the microstructure. All generated parameters are sent to the acoustic drive module in the form of control commands to achieve real-time sound field control.

[0038] Step S2 involves acquiring real-time image information of the microstructure and identifying its position, shape, and state in three-dimensional space. The image recognition results are used as input to the control logic and for feedback control. Specifically, this includes: (1) Acquire real-time images, including: confocal microscopy and inverted microscopy: Using microscopic imaging devices, such as inverted microscopes and confocal imaging modules, images of the microstructure manipulation area are acquired in real time to obtain image information of the particles in the current experiment. The acquired real-time images are displayed in the UI interface to facilitate observation of the movement of the microstructure.

[0039] (2) Using image recognition algorithms to identify and detect microstructures: Image recognition algorithms, such as lightweight recognition models based on convolutional neural networks (CNN) or YOLO, are used to extract microstructure target information from the imaging field of view, including its current center coordinates and edge morphology. The recognition model is pre-trained using a labeled dataset to improve the accuracy of microstructure recognition. After training, the data is imported and run for real-time image recognition and inference. The spatial coordinate data of the recognized microstructures is standardized to provide a transport path for generating microstructures. In addition to spatial coordinate information, the detected microstructure cell diameter and edge morphology information are also output.

[0040] (3) Feed back the positional deviation to the layout planning module in real time to participate in path correction: During the acoustic manipulation and handling of microstructures, images of the current field of view are continuously acquired, the actual position of the detected microstructures is updated in real time, and compared with the planned path. If an offset occurs, an error feedback signal is generated to indicate that the current sample position has been offset, and the parameters of the sound field are adjusted to achieve closed-loop control of the movement path.

[0041] Step S3: Receive sound field parameters and drive the ultrasonic transducer array to generate a three-dimensional sound field, achieving three-dimensional non-contact capture and transport of the microstructure, thus completing the three-dimensional structure assembly. Specifically: (1) Generate excitation signals and control the ultrasonic transducer based on acoustic control parameters: Based on acoustic control parameters such as frequency, amplitude, phase, and pulse repetition frequency, corresponding driving voltage pulses are generated. Independent or synchronous excitation of multiple array elements is supported to ensure the accuracy of the sound field in three-dimensional space. Based on the acoustic control parameters, the ultrasonic transducer is controlled to achieve the movement of the sound field focal point and the switching of the sound field shape. Figure 5 , Figure 6 ), to be suitable for a variety of acoustic control scenarios.

[0042] (2) Provide a two-way communication interface with the control terminal module: Communication is achieved through wired or wireless data transmission, allowing for real-time parameter adjustments and data readback to ensure the accuracy of acoustic control parameters.

[0043] (3) Amplify the power of the generated excitation signal and perform output matching: The generated excitation signal is amplified to provide sufficient energy to excite the front-end transducer. At the same time, impedance matching is used to reduce reflections during transmission, thereby reducing energy loss on the line.

[0044] In other embodiments of the present invention: In addition to using a phased array transducer system to generate a sound field, this invention can also use array-type or single transducer combined with scanning, or use volume waves and surface waves to generate a sound field and thus manipulate particles. The images of this invention can be recognized using a recognition method based on a convolutional neural network (CNN) deep learning model, or using an AI image recognition method based on a Transformer architecture; The manipulation strategy of this invention can be used not only for pre-planning structural parameters, but also for optimizing the generation of microstructure manipulation paths by means of reinforcement learning model feedback; The microstructures mentioned in this invention include, but are not limited to, cells, organoids, microspheres, biological tissue fragments, microrobots, microdevices, and other microscale targets that can be spatially manipulated in three-dimensional space.

[0045] This invention improves the three-dimensional spatial freedom and customizability of microstructure assembly, and combines artificial intelligence to automatically identify microstructures in the field of view for manipulation, providing automatic path planning and feedback correction for acoustic manipulation systems. It is suitable for bioengineering applications such as constructing microstructure arrays and tissue splicing models.

[0046] This invention studies a three-dimensional acoustic field manipulation system based on acoustic tweezers and an automatic identification and assembly process for microstructures. It optimizes acoustic field control strategies and assembly path planning to achieve the automatic transport and combination of multiple microstructures in three-dimensional space. By studying the application of multi-target image recognition and deep learning models in microstructure recognition, the system's ability to recognize the position, shape, and arrangement of microstructures is improved, thereby automating target matching and path generation. Furthermore, by studying three-dimensional programmable acoustic field control methods and transducer array layout, the acoustic field distribution and manipulation stability are optimized, enabling non-contact movement, positioning, and assembly operations of microstructures, thus improving system operating efficiency and adaptability.

[0047] Although the present invention has been described with reference to the present preferred embodiments, those skilled in the art should understand that the above preferred embodiments are only used to illustrate the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An automated assembly system for microstructures, characterized in that, The system includes a control terminal module, an imaging recognition module, and an acoustic drive module, wherein: The control terminal module is used to construct a three-dimensional assembly layout model of the target microstructure, and calculate the transport target position and transport path of the microstructure based on the initial position information of the microstructure obtained by the imaging recognition module, while generating sound field control parameters. The imaging recognition module is used to acquire real-time image information of microstructures, identify their position, shape and state in three-dimensional space, use the image recognition results as input to control logic, and use them for feedback control. The acoustic drive module is used to receive sound field parameters and drive the ultrasonic transducer array to generate a three-dimensional sound field, thereby realizing three-dimensional non-contact capture and handling of microstructures.

2. The system as described in claim 1, characterized in that, The control terminal module is specifically used for: Perform microstructure assembly structural modeling; Generate a transport path and dynamically adjust it based on feedback from the imaging recognition module; Based on the spatial layout of the target and the required sound field shape, control parameters that match the acoustic drive module are generated.

3. The system as described in claim 2, characterized in that, The aforementioned microstructure assembly structure modeling includes: Before performing the microstructure assembly operation, a physical 3D model of the expected assembly form is completed in the control terminal module, and the spatial information of the assembly is obtained from the expected microstructure model. The spatial information of the assembly mainly includes: the number of microstructures and the three-dimensional spatial coordinates (x, y, z) of each microstructure. After modeling is completed, the expected coordinate information of the microstructure is used as the target point for acoustic manipulation, and the sound field of the target point and the corresponding emission delay parameters are calculated.

4. The system as described in claim 3, characterized in that, The generated transport path includes: The spatial coordinates of the microstructure identified in the image are (x1, y1, z1), and the target point of one of the microstructures in the assembly is located at (x2, y2, z2). First, using the above two positions as the starting and ending coordinates, n movement steps are calculated according to the acoustic control movement path. The number of steps n is calculated based on the distance between the two points / step length. After obtaining the focal position corresponding to each step, the ultrasonic phase information during ultrasonic emission is calculated using the time inversion IB algorithm. Then, the phase relationship of the ultrasound is converted into the delay parameters corresponding to each channel, thereby calculating n sets of delay parameter data for m channels; m is the total number of emission channels.

5. The system as described in claim 4, characterized in that, The imaging recognition module is specifically used for: Acquire real-time images; Image recognition algorithms are used to identify and detect microstructures; It provides real-time feedback on positional deviations and participates in path correction.

6. The system as described in claim 5, characterized in that, The aforementioned image recognition algorithm for identifying and detecting microstructures includes: Image recognition algorithms are used to extract microstructure target information in the imaging field of view, including its current center position coordinates and edge morphology; the recognition model is pre-trained using a labeled dataset for image recognition training; after training, the data is imported and run for real-time image recognition and inference.

7. The system as described in claim 6, characterized in that, The real-time feedback of position deviation, participating in path correction, includes: The system continuously acquires images of the current field of view, updates the actual position of detected microstructures in real time, and compares them with the planned path. If an offset occurs, an error feedback signal is generated to indicate to the control terminal module that the current sample position has shifted. The control terminal module then adjusts the sound field parameters to achieve closed-loop control of the movement path.

8. The system as described in claim 7, characterized in that, The acoustic driving module is specifically used for: Based on acoustic control parameters, an excitation signal is generated and the ultrasonic transducer is controlled. Provides a two-way communication interface with the control terminal module; The generated excitation signal is amplified and output matched.

9. The system as described in claim 8, characterized in that, The process of generating excitation signals and controlling the ultrasonic transducer based on acoustic control parameters includes: Based on the acoustic control parameters sent by the control terminal module, a corresponding driving voltage pulse is generated; based on the acoustic control parameters, the acoustic drive module controls the ultrasonic transducer to realize the position movement of the sound field focal point and the switching of the sound field shape.

10. A method for automatically assembling microstructures, characterized in that, The method includes: Step S1: Construct a three-dimensional assembly layout model of the target microstructure, and calculate the transport target position and transport path of the microstructure based on the obtained initial position information of the microstructure, while generating sound field control parameters. Step S2: Acquire real-time image information of the microstructure, identify its position, shape and state in three-dimensional space; use the image recognition results as input to the control logic and for feedback control; Step S3: Receive sound field parameters and drive the ultrasonic transducer array to generate a three-dimensional sound field, thereby achieving three-dimensional non-contact capture and transport of the microstructure and completing the three-dimensional structure assembly.

Images