Artificial intelligence experiment platform with somatosensory interaction function

Abstract

The invention discloses an artificial intelligence experiment platform with a somatosensory interaction function, and relates to the technical field of interaction demonstration. The pressure sensor array comprises a plurality of pressure sensors which are distributed in a rectangular array, and each pressure sensor has a unique position coordinate; the base is used for installing a plurality of lifting assemblies distributed in a rectangular array, the arrangement positions of the lifting assemblies are in one-to-one correspondence with the position coordinates of the pressure sensors, and each lifting assembly is provided with a lifting end capable of emitting light upwards; the tracing assembly is arranged above the lifting assembly array, and the tracing assembly is provided with a tracing end and is configured to reproduce a pressing track through movement; the visual sensing assembly is mounted on the tracing assembly; wherein the computer is configured to drive the corresponding lifting assembly to lift the lifting end according to information generated by pressing the pressure sensor by a user, and drive the tracing end to move so as to reproduce in space, so that effective interaction is realized.

Description

Technical Field

[0001] This invention relates to the field of interactive demonstration technology, and more specifically to an artificial intelligence experimental platform with motion-sensing interaction capabilities. Background Technology

[0002] With the deep integration of artificial intelligence and human-computer interaction technologies, existing technologies have revealed significant shortcomings in both interactive experience and algorithm verification. Current human-computer interaction paradigms heavily rely on two-dimensional planes such as screens, resulting in abstract user operations, simplistic feedback, a lack of immersive physical perception, and a sense of disconnect between humans and machines. Meanwhile, in the field of AI algorithm verification, existing methods are either trapped in the practical gap of pure software simulation, where the algorithm decision-making process is opaque; or constrained by the black box nature of physical entity testing, making it difficult to intuitively evaluate and efficiently optimize algorithm performance. Therefore, existing technologies generally suffer from a disconnect between interaction and verification.

[0003] Therefore, it is necessary to provide an artificial intelligence experimental platform with motion-sensing interaction capabilities to solve the above problems. Summary of the Invention

[0004] To address the above problems, the present invention provides the following technical solution: an artificial intelligence experimental platform with motion-sensing interaction function, comprising:

[0005] computer;

[0006] A pressure sensor array comprising multiple pressure sensors arranged in a rectangular array, wherein each pressure sensor has unique position coordinates;

[0007] The base is used to mount multiple lifting components arranged in a rectangular array. The arrangement of the lifting components corresponds one-to-one with the position coordinates of the pressure sensor. Each lifting component has a lifting end that can emit light upwards.

[0008] A tracking component is disposed above the lifting component array, the tracking component having a tracking end and being configured to reproduce the pressing trajectory by motion;

[0009] A visual sensing component is mounted on the tracking component;

[0010] The computer is configured as follows:

[0011] Based on the position coordinates and corresponding time information generated by the user pressing the pressure sensor, the lifting component corresponding to the position coordinate is driven to lift its lifting end, and the tracking end is driven to move according to the pressing time sequence to reproduce the pressing trajectory in space.

[0012] When the tracking end moves to a predetermined range above the lifting end of any lifting component during the movement, the lifting end is triggered to emit light.

[0013] Furthermore, as a preferred embodiment, the computer is also configured to:

[0014] Based on the intensity and timing information of the pressure signal generated by the user pressing the pressure sensor, the position coordinates and pressing time information of the pressed pressure sensor are obtained, the lifting component at the corresponding position is driven to lift, and the lifting end of the lifting component is divided into trajectory route point markers and target point markers according to the intensity of the pressure signal.

[0015] Based on the position coordinate sequence and its corresponding time information, the tracking end is driven to move sequentially to generate a tracking trajectory above the lifting component array;

[0016] When the tracking end moves to a predetermined range above the lifting end, the lifting end is triggered to emit light;

[0017] The visual sensing component is used to identify the spatial distribution of the light-emitting lifting end and obtain its optical distribution characteristics. The computer calculates and evaluates the complexity of the tracking trajectory based on the optical distribution characteristics.

[0018] Based on the evaluation results of the complexity, the built-in machine learning algorithm is invoked to optimize the path planning strategy of the path-finding component.

[0019] The evaluation, optimization, and path-following operations are performed continuously until the path-following end reaches the target point, at which point the calculated complexity is reduced to below a preset threshold.

[0020] Furthermore, preferably, the tracking component includes:

[0021] Three columns evenly spaced around the perimeter;

[0022] A sliding cylinder is movably fitted onto the column along the axial direction of the column;

[0023] Two hinged rods are symmetrically hinged to the slide cylinder;

[0024] The tracking seat is hinged to the ends of the plurality of hinged rods, forming the tracking end;

[0025] A magnetic block is disposed on the tracking base.

[0026] Furthermore, preferably, the lifting component includes:

[0027] The outer frame is fixed to the base;

[0028] Four corner panels arranged in a rectangular shape and fixed to the inner wall of the outer frame;

[0029] The lifting seat constitutes the lifting end, and its sidewall is slidably engaged with the corner plate.

[0030] The lifting assembly is connected to the lifting seat to drive its lifting movement.

[0031] Furthermore, as a preferred embodiment, a limiting rod is fixedly provided at the bottom of the lifting seat, and the limiting rod passes through a guide hole on the bottom plate of the outer frame;

[0032] The lifting seat is connected to a flexible power supply cable.

[0033] Furthermore, preferably, the lifting seat includes:

[0034] The top compartment contains a first light emitter and a second light emitter, each with a different emission spectrum.

[0035] The bottom position is located below the top position;

[0036] The first contact point, fixed to the top of the silo, is powered via a cable;

[0037] The second contact point, which is vertically slidably located in the bottom compartment, is connected to the first light-emitting element via a wire;

[0038] The first spring has its two ends abutting against the inner wall of the bottom compartment and the bottom of the second contact point, respectively.

[0039] When a magnetic force is applied to the tracked end, the second contact compresses the first spring and moves upward until it contacts the first contact to form the first light-emitting circuit.

[0040] Furthermore, preferably, the lifting seat also includes:

[0041] The third contact point, which is vertically slidably located in the bottom compartment, is connected to the second light-emitting body via a wire;

[0042] The second spring has its two ends abutting against the inner wall of the bottom compartment and the bottom of the third contact point, respectively.

[0043] The third contact compresses the second spring and moves upward when a magnetic force is applied to the tracked end, until it contacts the first contact to form a second light-emitting circuit.

[0044] Furthermore, preferably, the visual sensing component includes:

[0045] Mounting bracket for fixed tracking components;

[0046] The rotating arm that connects to the mounting base rotates about the first horizontal axis;

[0047] Rotary base that rotates around a vertical axis to connect to the rotating arm;

[0048] A vision sensor is rotatably connected to a rotating base about a second horizontal axis, wherein the second horizontal axis is perpendicular to the first horizontal axis.

[0049] Compared with existing technologies, the present invention provides an artificial intelligence experimental platform with motion-sensing interaction function, which has the following beneficial effects:

[0050] This invention breaks through the limitations of traditional screen interaction, transforming abstract digital trajectories into directly observable physical entities and dynamic lighting effects. After defining a task through a press operation, the user can observe in real time the complete process of the physical trajectory being drawn in space, including the physical dot matrix formed by the raised component and the dynamic luminous trajectory lines. This interaction paradigm, which directly maps user intent to physical processes, provides a strong sense of immersion and intuitiveness.

[0051] In this invention, the invisible algorithmic decision-making process within the machine is transformed into intuitive optical signals. Hybrid light identifies the correct path, while single light indicates the algorithm's autonomous adjustments. Combined with real-time capture of optical distribution features and complexity assessments by visual sensing components, a complete closed loop of visual feedback on algorithm performance is constructed. Attached Figure Description

[0052] Figure 1 A schematic diagram of the overall structure of an artificial intelligence experimental platform with motion-sensing interaction function;

[0053] Figure 2 A three-dimensional structural diagram of the tracking component, the lifting component, and the vision sensing component;

[0054] Figure 3 A three-dimensional structural diagram of the lifting component;

[0055] Figure 4 This is a cross-sectional structural diagram of the lifting seat;

[0056] Figure 5 A schematic diagram of the three-dimensional structure of the visual sensing component;

[0057] In the diagram: 1. Computer; 2. Pressure sensor; 3. Tracking assembly; 4. Base; 5. Lifting assembly; 6. Vision sensor assembly; 31. Tracking seat; 32. Magnetic block; 33. Column; 34. Slide cylinder; 35. Hinge rod; 51. Outer frame; 52. Corner plate; 53. Lifting seat; 54. Position sensor; 55. Limiting rod; 56. Lifting assembly; 57. Cable; 531. Bottom compartment; 532. Top compartment; 533. Light source; 534. First contact; 535. Second contact; 536. First spring; 537. Second contact; 538. Second spring; 61. Mounting base; 62. Rotating arm; 63. Rotating seat; 64. Vision sensor. Detailed Implementation

[0058] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0059] Example: In this embodiment of the invention, please refer to... Figures 1-5 An artificial intelligence experimental platform with motion-sensing interaction function is provided, including:

[0060] Computer 1;

[0061] A pressure sensor array comprising multiple pressure sensors 2 arranged in a rectangular array, wherein each pressure sensor 2 has a unique position coordinate;

[0062] The base 4 is used to install multiple lifting components 5 arranged in a rectangular array. The arrangement of the lifting components 5 corresponds one-to-one with the position coordinates of the pressure sensor 2. Each lifting component 5 has a lifting end that can emit light upwards.

[0063] Tracking component 3 is disposed above the array of lifting components 5. The tracking component 3 has a tracking end and is configured to reproduce the pressing trajectory through movement.

[0064] The visual sensing component 6 is mounted on the tracking component 3;

[0065] The computer 1 is configured as follows:

[0066] Based on the position coordinates and corresponding time information generated by the user pressing the pressure sensor 2, the lifting component 5 corresponding to the position coordinate is driven to lift its lifting end, and the tracking end is driven to move according to the pressing time sequence to reproduce the pressing trajectory in space.

[0067] When the tracking end moves to a predetermined range above the lifting end of any lifting component 5 during the movement, the lifting end is triggered to emit light.

[0068] The user presses on a pressure sensor array consisting of multiple pressure sensors 2. Each pressure sensor 2 has a unique position coordinate. Therefore, when the user presses, the computer 1 can not only capture the position coordinate, but also record the time information through the timing function of the computer 1.

[0069] After receiving the aforementioned position coordinate information, computer 1 will perform the first step of physical mapping, driving the lifting component 5 corresponding to the pressed position to move its lifting end upward. For example, if the user presses the pressure sensor 2 at coordinate (1.1), the lifting end of the lifting component 5 arranged at the same coordinate position on the base 4 will be driven to rise. After the user completes a series of presses, a physical dot matrix composed of multiple raised lifting ends will be formed above the array of lifting components 5. This dot matrix visually reproduces the geometry of the user's pressing trajectory.

[0070] Based on the formation of the physical dot matrix, computer 1 enters the second step of dynamic reproduction. It controls the tracking end of the tracking component 3 to move in space according to the recorded pressing time sequence. The tracking end will strictly follow the order in which the user pressed the buttons, and will sequentially visit and pass over the raised ends to reproduce the user's original pressing trajectory.

[0071] During the movement of the tracking end, when it moves to a predetermined range above any of the lifting components 5, the lifting component 5 will be triggered to emit light. This means that as the tracking end moves, an illuminated trajectory line will be dynamically drawn in space, marking the progress and path of the tracking in real time. This provides users with intuitive and vivid visual feedback.

[0072] In other words, by transforming abstract digital trajectories into directly observable physical entities and dynamic lighting effects, this platform provides a highly immersive interactive experience. Users no longer observe indirectly through a screen, but can directly see the process of a physical trajectory being drawn in space, which is of great value for path planning verification.

[0073] The computer 1 is further configured as follows:

[0074] Based on the intensity and timing information of the pressure signal generated by the user pressing the pressure sensor 2, the position coordinates and pressing time information of the pressed pressure sensor 2 are obtained, the lifting component 5 at the corresponding position is driven to lift, and the lifting end of the lifting component 5 is divided into trajectory route point markers and target point markers according to the intensity of the pressure signal.

[0075] Based on the position coordinate sequence and its corresponding time information, the tracking end is driven to move sequentially to generate a tracking trajectory above the lifting component 5 array;

[0076] When the tracking end moves to a predetermined range above the lifting end, the lifting end is triggered to emit light;

[0077] The visual sensing component 6 is used to identify the spatial distribution of the light-emitting lifting end and obtain its optical distribution characteristics. The computer 1 calculates and evaluates the complexity of the tracking trajectory based on the optical distribution characteristics.

[0078] Based on the evaluation results of the complexity, the built-in machine learning algorithm is invoked to optimize the path planning strategy of the path-following component 3.

[0079] The evaluation, optimization, and path-following operations are performed continuously until the path-following end reaches the target point, at which point the calculated complexity is reduced to below a preset threshold.

[0080] The difference from the basic solution lies in the introduction of pressure signal strength as a new information dimension in this embodiment. Computer 1 can distinguish whether the user presses lightly or heavily. Based on this, computer 1 classifies the points generated by the user's press: lightly pressed points are marked as trajectory route points, defining the general direction of the path; while heavy pressed points are marked as target points, clarifying the destination of the tracking task.

[0081] After obtaining the location and time sequence of the route points and target points, computer 1 drives the tracking end to perform a tracking movement on the physical point matrix composed of these lifting ends according to the initial path planning strategy, generating an initial tracking trajectory.

[0082] During the tracking process, the triggered lifting ends emit light, forming a dynamic optical trajectory. The vision sensing component 6, installed on the tracking component 3, identifies and captures the distribution of these emitting lifting ends in three-dimensional space in real time, i.e., the optical distribution characteristics. The computer then analyzes these characteristics and calculates a quantitative index, namely the complexity of the tracking trajectory. This complexity is not an abstract concept, but is calculated based on specific optical characteristics. For example, it can be a weighted value of physical parameters such as the total length of the trajectory, the degree of curvature change of the trajectory, or the smoothness of the trajectory path.

[0083] Computer 1 uses the complexity evaluated in the previous step as feedback and calls its built-in machine learning algorithm to optimize the path planning strategy for the next step. For example, this machine learning algorithm could be a user-written reinforcement learning model, where the state is the relative position of the current tracking device to each route point and the target point, the action is the direction and distance of the next movement, and the reward function is positively correlated with the degree of complexity reduction. The algorithm's goal is to learn a strategy that maximizes the cumulative reward. This process continues, forming a self-improving closed loop until the tracking device successfully reaches the target point, and the calculated trajectory complexity is below a preset threshold. This verifies the effectiveness of the user-written algorithm, and this verification can be observed in real time, enabling real-world interaction.

[0084] Furthermore, using pressure intensity to distinguish between route points and target points is an efficient human-computer interaction method. Users don't need complex programming or configuration; they can teach the system a task completely with just a single physical press. This interaction model, which integrates task definition, execution, and feedback, greatly enhances the user experience.

[0085] Furthermore, users can deploy, test, and compare various machine learning or path planning algorithms on this platform, observing their performance in the real physical world. The output of the physical entities provides intuitive feedback.

[0086] In this embodiment, the tracking component 3 includes:

[0087] Three circumferentially distributed columns 33;

[0088] The sliding cylinder 34 is movably sleeved on the column 33 along the axial direction of the column 33;

[0089] Two hinge rods 35 are symmetrically hinged to the slide cylinder 34;

[0090] Tracking seat 31 is hinged to the ends of the plurality of hinge rods 35, forming the tracking end;

[0091] A magnetic block 32 is disposed on the tracking base 31.

[0092] Three circumferentially spaced columns 33 form a stable support frame for the tracking component 3. The slide cylinder 34 is movably mounted on the columns 33 and is driven to move up and down by an external drive mechanism, such as a belt drive mechanism, a lead screw and nut pair mechanism, or an electric telescopic mechanism.

[0093] In this embodiment, the lifting component 5 includes:

[0094] The outer frame 51 is fixed to the base 4;

[0095] Four rectangular corner plates 52 are fixed to the inner wall of the outer frame 51;

[0096] The lifting seat 53 forms the lifting end, and its sidewall is slidably engaged with the corner plate 52.

[0097] The lifting assembly 56 is connected to the lifting seat 53 to drive its lifting movement. A position sensor 54 may also be provided on the outer frame 51 to sense whether the lifting seat 53 has moved into place.

[0098] The lifting assembly 5 is fixed to the base 4 by an outer frame 51. Four rectangular corner plates 52 are fixed to the inner wall of the outer frame 51. These four corner plates 52 together form a vertical guide frame. The side wall of the lifting seat 53 is in sliding engagement with these four corner plates. When the lifting seat 53 is subjected to a driving force, its side wall will slide along the surface of the corner plates. Since the corner plates 52 are fixed and parallel to each other, this sliding engagement restricts the degree of freedom of movement of the lifting seat 53, allowing it to only move linearly along the vertical direction defined by the corner plates 52, and preventing horizontal offset or rotation.

[0099] When the lifting assembly 56 is activated, it transmits power to the lifting seat, forcing the lifting seat 53 to move smoothly up and down along the vertical track set by the corner plate. The lifting assembly 56 can be an electric telescopic rod, a cylinder, a hydraulic cylinder, etc.

[0100] Furthermore, a limiting rod 55 is fixedly provided at the bottom of the lifting seat 53, and the limiting rod 55 passes through the guide hole on the bottom plate of the outer frame 51;

[0101] The lifting seat 53 is connected to a flexible power supply cable 57.

[0102] The lifting seat 53 includes:

[0103] Top compartment 532 contains a first light emitter and a second light emitter, each with a different emission spectrum;

[0104] The bottom position 531 is positioned below the top position 532;

[0105] The first contact 534, fixed to the top compartment 532, is powered via cable 57;

[0106] The second contact 535, which is vertically slidably located in the bottom compartment 531, is connected to the first light-emitting body via a wire;

[0107] The first spring 536 has its two ends abutting against the inner wall of the bottom chamber 531 and the bottom of the second contact point 535, respectively.

[0108] When the tracking end is magnetically applied, the second contact 535 compresses the first spring 536 and moves upward until it contacts the first contact 534 to form the first light-emitting circuit.

[0109] The lifting seat 53 also includes:

[0110] The third contact 537, which is vertically slidably located in the bottom compartment 531, is connected to the second light-emitting body via a wire;

[0111] The second spring 538 has its two ends abutting against the inner wall of the bottom chamber 531 and the bottom of the third contact point 537, respectively.

[0112] The third contact 537 compresses the second spring 538 and moves upward when a magnetic force is applied to the tracked end, until it contacts the first contact 534 to form a second light-emitting circuit.

[0113] In other words, in this embodiment, inside the lifting seat 53, the second contact 535 and the third contact 537, which form two light-emitting circuits, have different initial heights when not triggered. Specifically, the initial position of the third contact 537 is designed to be higher than that of the second contact 535.

[0114] Based on the aforementioned mechanical differences, the response of the lifting seat 53 to a constant magnetic field exhibits two distinct distance thresholds:

[0115] Long-distance threshold (triggered single light): When the tracking end is at a preset tracking height and passes a non-raised lifting seat 53, the distance between them is relatively far. At this distance, the strength of the constant magnetic field is sufficient to overcome the elastic force of the second spring 538 above the third contact 537, which is initially positioned higher and closer to the magnetic source, causing it to move upward and close the second light-emitting circuit, generating a single light. However, the magnetic field strength is insufficient to overcome the elastic force of the first spring 536 above the second contact 535, which is initially positioned lower and farther away.

[0116] Near-distance threshold (triggers mixed light): When the lifting seat 53 is raised by user command, the distance between its entirety and the tracking end is significantly reduced. At this close distance, the strength of the constant magnetic field is sufficient to overcome the elastic force of the two springs (first spring 536 and second spring 538) simultaneously, causing the third contact 537 and the second contact 535 to both move upward and close their respective light-emitting circuits, thereby generating mixed light.

[0117] It is this distance-sensitive dual-threshold triggering mechanism, based on preset mechanical differences, that gives the lifting seat 53 the ability to encode information.

[0118] First tracking (correct reproduction): When the tracking end passes the lifting seat 53 raised by the user for the first time, the distance between the lifting seat 53 and the tracking end is reduced, and the constant magnetic field strength is sufficient to overcome the elastic force of the two springs at the same time, triggering two light-emitting circuits, causing all accessed lifting ends to emit mixed light (two kinds of light).

[0119] Subsequent tracking (autonomous adjustment): In the subsequent tracking optimization process, when the computer 1 guides the tracking end past some lifting seats 53 that have not been lifted by the user, due to the large distance between them, the constant magnetic field strength can only overcome the elastic force of the second spring 538 above the initially higher position 537, causing it to move upward to close the second light-emitting circuit and generate a single light.

[0120] In other words, one of the core innovations of the artificial intelligence experimental platform proposed in this invention lies in constructing an immersive human-computer interaction paradigm. This interaction paradigm is not a traditional command-response model, but a dialogic interaction based on physical entity mapping and information-encoded light effects. It transforms the user's abstract intentions into perceptible physical processes and visualizes the invisible algorithmic decision-making processes within the machine through intuitive optical signals, thereby achieving two-way, deep communication between humans and machines at both the physical and informational levels.

[0121] In this embodiment, the visual sensing component 6 includes:

[0122] Mounting base 61 for fixed tracking component 3;

[0123] Rotate the connecting arm 62 of the mounting base 61 about the first horizontal axis (X direction);

[0124] Rotate the pivot 63 connecting the rotating arm 62 about the vertical axis (Y direction);

[0125] A vision sensor 64 is rotatably connected to a rotating base 63 about a second horizontal axis (Z-axis), wherein the second horizontal axis is perpendicular to the first horizontal axis.

[0126] In other words, the visual sensing component 6 achieves multi-angle adjustment of the visual sensor 64 through a three-axis series rotating structure.

[0127] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An artificial intelligence experiment platform with somatosensory interaction function, characterized in that, The application relates to a computer (1) and a pressure sensor array comprising a plurality of pressure sensors (2) arranged in a rectangular array, wherein each pressure sensor (2) has a unique position coordinate; a base (4) for mounting a plurality of lifting assemblies (5) arranged in a rectangular array, the arrangement position of the lifting assemblies (5) corresponding to the position coordinates of the pressure sensors (2) one by one, the lifting assemblies (5) having upwardly emitting lifting ends; a tracing assembly (3) arranged above the array of the lifting assemblies (5), the tracing assembly (3) having a tracing end and being configured to reproduce a pressing track by movement; a visual sensing assembly (6) mounted on the tracing assembly (3); wherein the computer (1) is configured to: drive the lifting assembly (5) corresponding to the position coordinate to lift its lifting end according to the position coordinate and corresponding time information generated by a user pressing the pressure sensor (2), and drive the tracing end to move according to the time sequence of pressing to reproduce the pressing track in space; trigger the lifting end to emit light when the tracing end moves to a predetermined range above the lifting end of any lifting assembly (5) during movement. The computer (1) is further configured to: obtain the position coordinate of the pressed pressure sensor (2) and the pressing time information according to the received pressing signal strength and occurrence time sequence information generated by the user pressing the pressure sensor (2), drive the lifting assembly (5) at the corresponding position to lift, and distinguish the lifting end of the lifting assembly (5) into track route point markers and target point markers according to the pressing signal strength; drive the tracing end to move in sequence to generate a tracing track above the array of the lifting assemblies (5) based on the position coordinate sequence and the corresponding time information; trigger the lifting end to emit light when the tracing end moves to a predetermined range above the lifting end; the visual sensing assembly (6) is used for identifying the distribution of the light-emitting lifting end in space, obtaining the optical distribution characteristics, the computer (1) calculates and evaluates the complexity of the tracing track based on the optical distribution characteristics; based on the evaluation result of the complexity, the built-in machine learning algorithm is called to optimize the tracing path planning strategy of the tracing assembly (3); the operations of the evaluation, optimization and tracing according to the updated path are continuously performed until the calculated complexity is lower than a preset threshold when the tracing end reaches the target point. The tracing assembly (3) comprises: three circumferentially distributed and spaced-apart columns (33); a sliding cylinder (34) movably sleeved on the columns (33) in the axial direction of the columns (33); two hinged rods (35) symmetrically hinged on the sliding cylinder (34); a tracing seat (31) hinged on the ends of the hinged rods (35) and constituting the tracing end; and a magnetic block (32) arranged on the tracing seat (31). The lifting assembly (5) comprises: an outer frame (51) fixed to the base (4); four corner plates (52) arranged in a rectangular manner and fixed to the inner wall of the outer frame (51); ​ ​ ​ ​ ​ 2. The artificial intelligence experiment platform according to claim 1, characterized in that, ​ ​ ​ ​ ​ ​ ​ 3.The artificial intelligence experiment platform of claim 1, wherein, ​ ​ ​ ​ ​ ​ 4.The artificial intelligence experiment platform of claim 1, wherein, ​ ​ ​ A lifting seat (53) is arranged at the lifting end and is in sliding connection with the corner plate (52); A lifting assembly (56) is in driving connection with the lifting seat (53) to drive the lifting movement of the lifting seat (53). 5.The artificial intelligence experiment platform of claim 4, wherein: A limiting rod (55) is arranged at the bottom of the lifting seat (53) and penetrates through a guide hole in the bottom plate of the outer frame (51); The lifting seat (53) is connected with a flexible power supply cable (57). 6.The artificial intelligence experiment platform of claim 4, wherein, The lifting seat (53) comprises: A top compartment (532) in which a first light emitter and a second light emitter with different light spectra are arranged; A bottom compartment (531) arranged below the top compartment (532); A first contact (534) fixed to the top compartment (532) and supplied with power through the cable (57); A second contact (535) vertically slidingly arranged in the bottom compartment (531) and connected with the first light emitter through a wire; A first spring (536) with two ends abutting against the inner wall of the bottom compartment (531) and the bottom of the second contact (535), respectively; When the second contact (535) is subjected to a magnetic force from the tracing end, the first spring (536) is compressed and the second contact (535) moves upward until it contacts the first contact (534) to form a first light emitting circuit.

7. The artificial intelligence experiment platform of claim 6, wherein, The lifting seat (53) further comprises: A third contact (537) vertically slidingly arranged in the bottom compartment (531) and connected with the second light emitter through a wire; A second spring (538) with two ends abutting against the inner wall of the bottom compartment (531) and the bottom of the third contact (537), respectively; When the third contact (537) is subjected to a magnetic force from the tracing end, the second spring (538) is compressed and the third contact (537) moves upward until it contacts the first contact (534) to form a second light emitting circuit. 8.The artificial intelligence experiment platform of claim 1, wherein, The visual sensing assembly (6) comprises: A mounting seat (61) fixed to the tracing assembly (3); A rotating arm (62) rotatably connected with the mounting seat (61) about a first horizontal axis; A rotating seat (63) rotatably connected with the rotating arm (62) about a vertical axis; A visual sensor (64) rotatably connected with the rotating seat (63) about a second horizontal axis, wherein the second horizontal axis is perpendicular to the first horizontal axis.

Images