Two-dimensional and three-dimensional thundersight fusion twisting target torque-turn number-length synchronization method

Abstract

The invention discloses a two-dimensional and three-dimensional thundersight fusion twisting target torque-turn number-length synchronization method, which comprises the following steps of S1, constructing a three-dimensional model of a male joint and a female joint of a drill rod, and supporting dynamic loading of parameters of drill rods with different specifications; the female joint model is rendered by adopting a semitransparent material, the transparency is set to be 30%-70%, and an internal thread structure is visible; the part, entering the female joint, of the male joint is marked with a highlight color, and the meshing length is displayed in real time; s2, respectively carrying out precession turn number calculation, meshing length calculation, torque synchronous display, parameter error correction, data filtering processing, direction perception and symbol processing and multi-period accumulation calculation; s3, carrying out multi-view linkage visualization processing; and S4, realizing low-delay communication between the sensor data and the front-end three-dimensional model by adopting a WebSocket or MQTT protocol. According to the method, real-time synchronous display of the torque, the number of precession turns and the meshing length is achieved, and the operation precision and safety are improved.

Description

Technical Field

[0001] The present invention relates to the technical fields of emergency rescue, oil drilling engineering, visual monitoring of mechanical connections, industrial automation, and digital twin technology, and particularly relates to a method for synchronizing two-dimensional and three-dimensional radar-vision fusion for torsional target torque, number of turns, and length. Background Art

[0002] In fields such as oil drilling and mechanical assembly, the screw connection of drill pipes is a core link to ensure the safety of downhole operations; traditional technologies mainly monitor the screw connection process in the following ways:

[0003] Judgment based on manual experience: relying on operators to judge the torsional torque and meshing state through sound and touch, which is easily affected by subjective factors, has low accuracy, and poses safety hazards;

[0004] Two-dimensional data monitoring system: using sensors to collect data such as torque and rotation angle, and displaying them in the form of line charts or digital meters, but lacking spatial visualization ability and unable to intuitively reflect the dynamic meshing process of male and female thread connections;

[0005] Static three-dimensional model display: some systems use CAD software to construct a three-dimensional model of the drill pipe, but it is only limited to static display, unable to be linked with real-time sensor data, and the visibility problem of the internal structure of the female joint is not solved.

[0006] However, the current technology has the following key defects:

[0007] (1) Insufficient visualization dimension:

[0008] Existing two-dimensional curves, for example, torque-time curves, cannot express the spatial movement trajectory of thread screwing in, and it is difficult for operators to judge the actual screwing-in depth of the male thread; for example, whether it reaches the meshing length required by API standards; the occlusion of the internal structure of the female joint causes the phenomenon of "blind screwing", which may lead to problems such as thread misalignment, over-tightening, or incomplete meshing, resulting in equipment damage or even blowout accidents.

[0009] (2) Data fragmentation and asynchrony:

[0010] Parameters such as torque, number of turns, and length are usually displayed separately on different interfaces, lacking the ability of synchronous correlation analysis. For example, when the torque is abnormal, it is difficult to quickly locate whether it is caused by insufficient number of turns or deviation of the meshing length; the data refresh rate of existing systems is low, usually <5Hz, making it difficult to match high-speed screwing scenarios, such as automated drilling rigs, resulting in display delay and operation lag.

[0011] (3) Bottleneck of semi-transparent rendering technology:

[0012] Although traditional industrial software (such as Unity and Unreal Engine) supports translucent materials, it relies on high-performance GPUs and is difficult to integrate into the Web, which limits the lightweight deployment of on-site equipment.

[0013] Dynamic semi-transparent rendering is prone to problems such as model penetration and light refraction distortion (for example, the overlapping part between the internal thread of the female connector and the male buckle is blurred), which affects operational judgment.

[0014] Therefore, providing a method that can realize real-time synchronous display of torque, number of precession turns and engagement length and improve operation accuracy and safety is the key to the technical solution of the present invention. Summary of the invention

[0015] In view of this, the present invention provides a method for preparing tofu.

[0016] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0017] A method for synchronizing a twisting target torque, number of turns and length by integrating two-dimensional and three-dimensional radar vision includes the following steps:

[0018] Step S1: 3D model construction and display

[0019] Use the WebGL engine to build a 3D model of the male and female joints of the drill pipe, and support dynamic loading of parameters of drill pipes of different specifications;

[0020] The female connector model is rendered with a semi-transparent material, with the transparency set to 30%-70%, so that the internal thread structure is visible;

[0021] The part where the male connector enters the female connector is highlighted in a bright color, showing the engagement length in real time;

[0022] Step S2: Dynamic monitoring and calculation of parameters

[0023] It can calculate the number of precession circles, meshing length, torque synchronization display, parameter error correction, data filtering, direction perception and symbol processing, and multi-cycle accumulation calculation respectively;

[0024] Step S3: Multi-view linkage visualization processing

[0025] Step S4: Data synchronization and communication

[0026] Use WebSocket or MQTT protocol to achieve low-latency communication between sensor data and front-end 3D models.

[0027] Preferably, in step S2, the number of precession turns is calculated by obtaining the rotation angle θ (unit: degree) through a rotation angle sensor or a motor encoder, and combining the thread pitch P (unit: mm / turn) to calculate the actual number of precession turns N;

[0028] The calculation formula is:

[0029]

[0030] Wherein, the unit of the rotation angle θ is degree, and the unit of the thread pitch P is millimeter / turn;

[0031] When calculating the total feed length L, the calculation formula is:

[0032]

[0033] Wherein, the unit of the total feed length L is millimeter;

[0034] When the rotation angle θ changes with time, the number of feed turns is dynamically updated to:

[0035]

[0036] Preferably, in the step S2, the engagement length calculation is based on the product of the number of feed turns N and the pitch P, and the depth L of the male thread entering the female thread is dynamically updated;

[0037] The calculation formula is:

[0038]

[0039] When the number of feed turns N changes with time, the engagement length is dynamically updated to:

[0040]

[0041] Preferably, in the step S2, the torque is synchronously displayed by integrating the torque sensor data and mapping it to the 3D model and curve in real time; the calculation formula of the torque is:

[0042]

[0043] Wherein,

[0044] V is the output voltage of the sensor, and the unit is volt;

[0045] S is the sensitivity of the sensor, and the unit is volt / newton meter;

[0046] T is the torque, and the unit is newton meter;

[0047] The torque data can be plotted as a curve in real time through the time series T(t):

[0048] The formula is:

[0049]

[0050] Preferably, in the step S2, the parameter error correction includes precession cycle number correction, engagement length correction, and torque zero drift correction; among them,

[0051] The calculation formula for precession cycle number correction is:

[0052]

[0053] where, Δθ is the error of the angle sensor, unit: degree; N′ is the corrected cycle number;

[0054] The calculation formula for engagement length correction is:

[0055]

[0056] where, ΔP is the pitch mechanical error, unit: mm / rev; L′ is the corrected length;

[0057] The calculation formula for torque zero drift correction is:

[0058]

[0059] where, ΔV is the zero drift voltage of the sensor, unit: volt; T′ is the corrected torque.

[0060] Preferably, in the step 2, the data filtering process includes moving average filtering and low-pass filtering; among them,

[0061] The moving average filtering process takes the average value of continuous n sampling points:

[0062] Precession cycle number filtering:

[0063]

[0064] Torque filtering:

[0065]

[0066] The low-pass filtering process uses a first-order low-pass filter to suppress high-frequency noise:

[0067] y k =α·x k +(1-α)·y k-1 ;

[0068] where, α is the filtering coefficient, and the value range is 0 < α < 1;

[0069] x k is the current input data;

[0070] y k is the current output data.

[0071] Preferably, in step 2, when direction sensing and symbol processing support forward and reverse rotation, a rotation direction parameter D ∈ {+1, -1} needs to be introduced;

[0072] Correction of the symbol of the number of precession turns:

[0073]

[0074] Correction of the symbol of the engagement length:

[0075]

[0076] Preferably, in step 2, when the angle sensor is an incremental encoder during multi-period cumulative calculation, the angle of multiple turns needs to be accumulated:

[0077] θ total =θ current + 360°·k;

[0078] Total precession length:

[0079]

[0080] Preferably, in step S1, the material of the female joint uses the MeshPhongMaterial material of Threejs; the material of the male joint uses the MeshLambertMaterial material.

[0081] Preferably, in step S3, the specific processing method is as follows:

[0082] Front view: The 3D model dynamically displays the precession process, and the female joint is semi-transparent to show the engagement state;

[0083] Side view: Draw a torque-turns-time curve for data playback and analysis;

[0084] Parameter panel: Real-time display of the current torque value, cumulative number of turns and engagement length.

[0085] The present invention has achieved the following technical effects compared with the prior art:

[0086] (1) By adjusting the transparency of the material of the female joint and using the MeshPhongMaterial material of Three.js, the female joint is set to a semi-transparent state with the transparency controlled between 30% - 70%. Operators can observe the screwing-in depth of the male thread and the thread engagement state in real time without disassembling the drill pipe, effectively avoiding the hidden dangers of over-tightening or non-compliance;

[0087] (2) During the rendering process of the present invention, for the display of the translucent material of the female connector and the highlighted part of the male connector, the rendering order and fusion method are dynamically adjusted to prevent problems such as model penetration and light refraction distortion, ensuring that the model can clearly and accurately display the screwing process from different perspectives and lighting conditions, and improving the reliability of operation judgment;

[0088] (3) During the screwing process of the drill pipe, the changes in key parameters such as torque, number of turns, and length are closely related. Through the WebSocket or MQTT protocol, these parameters collected by the sensor are transmitted to the front end in real time. The front end uses the Three.js and React frameworks to dynamically map these data to the 3D model and curve graph, realizing real-time linkage between parameters;

[0089] (4) The multi-parameter synchronous mapping of the present invention not only facilitates the operator to monitor the screwing process in real time, but also provides comprehensive data support for subsequent data analysis and process optimization; through the comprehensive analysis of parameters such as torque, number of turns, and length, the mechanical characteristics and thread wear laws during the screwing process can be deeply studied, providing a basis for formulating more reasonable screwing process parameters and improving the connection quality and service life of the drill pipe;

[0090] (5) Based on the WebGL technology, the present invention can be deployed across platforms and supports access on various terminals such as PCs, tablets, and AR devices; the operator does not need to install specific software or plugins, and only needs to use a standard web browser, which greatly improves the usability and convenience. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] Figure 1 is a flowchart of a two-dimensional and three-dimensional radar-vision fusion twisting target torque - number of turns - length synchronization method of the present invention;

[0092] Figure 2 is a multi-view linkage visualization processing flowchart of a two-dimensional and three-dimensional radar-vision fusion twisting target torque - number of turns - length synchronization method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0093] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0094] As Figure 1 shown, the present invention discloses a two-dimensional and three-dimensional radar-vision fusion twisting target torque - number of turns - length synchronization method, including the following steps:

[0095] Step S1: 3D Model Construction and Display

[0096] Use the WebGL engine to construct a 3D model of the drill pipe male and female connectors, supporting dynamic loading of different specifications of drill pipe parameters;

[0097] Render the female connector model with a translucent material, set the transparency to 30%-70%, and the internal thread structure is visible;

[0098] The part of the male connector entering the female connector is marked with a highlighted color, and the meshing length is displayed in real time;

[0099] Among them, the material of the female connector uses the MeshPhongMaterial material of Threejs; the material of the male connector uses the MeshLambertMaterial material;

[0100] Step S2: Parameter Dynamic Monitoring and Calculation

[0101] Perform calculations for the number of rotation-in circles, meshing length calculation, torque synchronous display, parameter error correction, data filtering processing, direction perception and symbol processing, and multi-cycle cumulative calculation respectively;

[0102] The calculation of the number of rotation-in circles is to obtain the rotation angle θ (unit: degree) through a rotation angle sensor or motor encoder, and combine it with the thread pitch P (unit: millimeter / circle) to calculate the actual number of rotation-in circles N;

[0103] The calculation formula is:

[0104]

[0105] Among them, the unit of the rotation angle θ is degree, and the unit of the thread pitch P is millimeter / circle;

[0106] The calculation formula for calculating the total rotation-in length L is:

[0107]

[0108] Among them, the unit of the total rotation-in length L is millimeter;

[0109] When the rotation angle θ changes with time, the number of rotation-in circles is dynamically updated as:

[0110]

[0111] The meshing length calculation is based on the product of the number of rotation-in circles N and the pitch P, and dynamically updates the depth L of the male thread entering the female thread;

[0112] The calculation formula is:

[0113]

[0114] When the number of precession cycles N changes with time, the engagement length is dynamically updated as follows:

[0115]

[0116] Torque synchronous display is achieved by integrating torque sensor data and mapping it to a 3D model and curve in real time; the formula for torque is:

[0117]

[0118] Where,

[0119] V is the sensor output voltage, unit: volt);

[0120] S is the sensor sensitivity, unit: volt / Newton meter);

[0121] T is the torque, unit: Newton meter;

[0122] Torque data can be plotted as a curve in real time through the time series T(t):

[0123] The formula is:

[0124]

[0125] Parameter error correction includes precession cycle number correction, engagement length correction, and torque zero drift correction; among them,

[0126] The calculation formula for precession cycle number correction is:

[0127]

[0128] Where, Δθ is the error of the angle sensor, unit: degree; N′ is the corrected number of cycles;

[0129] The calculation formula for engagement length correction is:

[0130]

[0131] Where, ΔP is the pitch mechanical error, unit: mm / cycle; L′ is the corrected length;

[0132] The calculation formula for torque zero drift correction is:

[0133]

[0134] Where, ΔV is the sensor zero drift voltage, unit: volt; T′ is the corrected torque;

[0135] Data filtering processing includes moving average filtering and low-pass filtering; among them,

[0136] Moving average filtering takes the average value of n consecutive sampling points:

[0137] Precession revolution number filtering:

[0138]

[0139] Torque filtering:

[0140]

[0141] Low-pass filtering is performed using a first-order low-pass filter to suppress high-frequency noise:

[0142] y k = α · x k +(1 - α) · y k-1 ;

[0143] where α is the filtering coefficient, and the value range is 0 < α < 1;

[0144] x k is the current input data;

[0145] y k is the current output data;

[0146] When supporting forward and reverse rotation during direction sensing and symbol processing, a rotation direction parameter D ∈ {+1, -1} needs to be introduced;

[0147] Precession revolution number symbol correction:

[0148]

[0149] Engagement length symbol correction:

[0150]

[0151] When the angle sensor is an incremental encoder during multi-period cumulative calculation, the multi-turn angle needs to be accumulated:

[0152] θ total = θ current + 360° · k;

[0153] Total precession length:

[0154]

[0155] Step S3: Multi-view linkage visualization processing

[0156] The specific processing method is as follows:

[0157] Main view: The three-dimensional model dynamically displays the precession process, and the female joint is semi-transparently displayed in the engagement state;

[0158] Sub-view: Draw the torque-revolution number-time curve for data playback and analysis;

[0159] Parameter panel: Displays the current torque value, cumulative number of turns, and engagement length in real time;

[0160] As Figure 2 shown is the multi-view linkage visualization processing flowchart of the present invention;

[0161] Step S4: Data synchronization and communication

[0162] Adopt the WebSocket or MQTT protocol to achieve low-latency communication between sensor data and the front-end 3D model.

[0163] Example 1:

[0164] 3D model creation and import

[0165] Modeling tool selection: According to the complexity and accuracy requirements of the drill pipe model, select a suitable 3D modeling software; for drill pipes with complex shapes and fine thread structures, unity is an open-source and powerful choice; if engineering accuracy and parametric design are emphasized, SolidWorks is more suitable.

[0166] Model creation: In the modeling software, create male and female joint models according to the actual size and structure of the drill pipe. Accurately model parameters such as the tooth profile, pitch, and diameter of the thread to ensure the accuracy of the model; to improve rendering efficiency, optimize the model, reduce unnecessary faces, and retain key details at the same time.

[0167] Model export and import: Export the created model in STL format, which has good compatibility and is suitable for network transmission and rendering; use the STL loader provided by WebGL to import the model into the front-end project, and during the import process, set the unit scale of the model to be consistent with the actual size.

[0168] Material and texture settings

[0169] Female joint material: To achieve the semi-transparent effect of the female joint, use the MeshPhongMaterial material of Three.js; set the color to 0x6699ff, the transparency opacity to 0.5, and turn on the transparent property. This material can produce a high-gloss effect under light, making the model more three-dimensional.

[0170] Male joint material: The male joint uses an opaque material, and the color is distinguished from that of the female joint, such as set to a metallic gray; use the MeshLambertMaterial material, which has a soft response to light and can clearly display the details of the model.

[0171] Texture Mapping: If textures such as metal textures and wear marks need to be added to the model surface, use UV mapping technology; create UV coordinates for the model and map the texture image to the model surface to enhance visual realism.

[0172] Model Assembly and Scene Building

[0173] Model Assembly: In the WebGL scene, place the male and female connector models according to the actual assembly relationship, adjust their positions and rotation angles to ensure that the male connector aligns with the entrance of the female connector in the initial state.

[0174] Scene Building: Create a scene containing auxiliary elements such as coordinate axes and grids to help operators better understand the spatial position of the model. Add appropriate light sources such as point lights and directional lights to illuminate the model and highlight its three-dimensional effect.

[0175] Sensor Data Acquisition and Processing

[0176] Data Acquisition: Obtain the real-time rotation angle θ during the drill pipe's progressive rotation through a rotation angle sensor or motor encoder; at the same time, a torque sensor collects torque data T; these sensors usually output analog or digital signals and need to be converted into a computer-readable digital format through a data acquisition card or module.

[0177] Data Preprocessing: Perform preprocessing operations such as filtering and denoising on the acquired raw data; adopt a moving average filtering algorithm, take n consecutive data points for averaging to reduce random noise in the data.

[0178] For example, for the rotation angle data, calculate θ′ = (θ1 + θ2 +... + θ n ) / n, where θ′ is the filtered angle value.

[0179] Data Binding and Model Update

[0180] Rotation Animation Binding: Drive the rotation of the male connector model according to the processed rotation angle data. In WebGL, by setting the rotation property of the model, make it rotate by the corresponding angle around the Z-axis. Every time it rotates 360°, trigger the accumulation of the revolution counter, and at the same time calculate the actual progressive revolution number N = θ / (360° × P) according to the thread pitch P, and update the engagement length L = N × P.

[0181] Translation Animation Binding: As the male connector rotates, drive the male connector model to move in the negative Z-axis direction according to the change of the engagement length L. By updating the position property of the model, achieve the visual effect of the male connector gradually entering the female connector.

[0182] Torque Data Binding: Map the torque sensor data T onto a 3D model and a curve graph; in the 3D model, visually represent the magnitude of the torque by changing visual attributes such as the color depth and brightness of the male connector or the connection part.

[0183] For example, the greater the torque, the darker or brighter the color. At the same time, plot the torque-time curve in the curve graph to show the changing trend of the torque in real time.

[0184] Animation Control and Synchronization

[0185] Animation Control: Use the animation system of WebGL, such as AnimationMixer and Clock, to control the rotation and displacement animations of the model; set an appropriate animation time step to ensure smooth and natural animations; through control functions such as pause, resume, and reset, it is convenient for operators to observe the screwing process at different stages.

[0186] Data and Animation Synchronization: Ensure that the update of sensor data is synchronized with the playback of model animations; adopt the requestAnimationFrame mechanism. At each frame update, obtain the latest sensor data and update the model state and animation position accordingly; avoid visual errors caused by out-of-sync data updates and animation playback.

[0187] Front-end Architecture and Implementation

[0188] Framework Selection: The front-end is developed using the React framework in combination with the WebGL library; React provides an efficient component-based development mode, which is convenient for building complex user interfaces; WebGL focuses on 3D graphics rendering, and the combination of the two can achieve high-performance visualization applications.

[0189] Back-end Architecture and Implementation

[0190] Back-end Technology Selection: The back-end can be developed using Python or Node.js; Python has rich scientific computing and data processing libraries, such as NumPy, Pandas, etc., which are suitable for complex calculations on sensor data; Node.js has efficient network communication capabilities and can quickly handle data transmission between the front-end and sensors.

[0191] Data Processing and Storage: Receive real-time data from sensors, and perform further processing and analysis; if historical data needs to be stored and queried, an appropriate database can be selected, such as MongoDB, MySQL, etc.; store the data in a structured manner for subsequent data mining and process optimization.

[0192] Interface Design: Design standardized RESTful API or WebSocket interfaces to achieve data interaction between the front end and the back end; the front end obtains sensor data, sends control instructions, etc. by calling these interfaces.

[0193] Deployment and Access

[0194] Server Deployment: Deploy the front-end and back-end codes to the server, and cloud servers or local servers can be selected; configure the network environment, security policies, etc. of the server to ensure stable operation and data security.

[0195] Through real-time monitoring of key parameters such as torque, number of turns, and meshing length, operators can precisely control the screwing process of the drill pipe, effectively avoiding downhole accidents caused by improper screwing, such as blowouts and stuck pipes.

[0196] Through visual monitoring of the screwing process, maintenance personnel can comprehensively and intuitively understand the wear condition and internal structure status of the drill pipe. This not only helps to timely detect potential fault hazards but also provides a strong basis for formulating reasonable maintenance plans. In the training area of the workshop, new employees can use this method to observe the standard screwing process to deepen their understanding of the drill pipe connection technology. At the same time, it can also record various data during the maintenance process for subsequent quality traceability and process optimization. For example, by analyzing historical data, workshop technicians can determine the best screwing parameters for different types of drill pipes to improve maintenance efficiency and quality.

[0197] The above is only a preferred embodiment of the present invention, and it does not impose any limitation on the technical scope of the present invention. Therefore, any minor modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A method for synchronizing torque - number of turns - length of a two - and three - dimensional radar - vision fusion twisting target, characterized in that, It includes the following steps: Step S1: 3D model construction and display Use the WebGL engine to construct a 3D model of the drill pipe male and female joints, supporting dynamic loading of drill pipe parameters of different specifications; The female joint model is rendered with a translucent material, and the transparency is set to 30%-70%, with the internal thread structure visible; The part of the male joint entering the female joint is marked with a highlighted color, and the meshing length is displayed in real time; Step S2: Parameter dynamic monitoring and calculation Perform calculations of the number of rotation-in circles, meshing length calculation, torque synchronous display, parameter error correction, data filtering processing, direction perception and symbol processing, and multi-cycle cumulative calculation respectively; Step S3: Multi-view linkage visualization processing Step S4: Data synchronization and communication Use the WebSocket or MQTT protocol to achieve low-latency communication between sensor data and the front-end 3D model.

2. A two - three - dimensional radar - vision fusion twisting target torque - rotation number - length synchronization method according to claim 1, characterized in that In the step S2, the calculation of the number of rotation-in circles is to obtain the rotation angle θ (unit: degree) through a rotation angle sensor or a motor encoder, and combine it with the thread pitch P (unit: mm / circle) to calculate the actual number of rotation-in circles N; The calculation formula is: where the unit of the rotation angle θ is degree, and the unit of the thread pitch P is mm / circle; The calculation formula when calculating the total rotation-in length L is: where the unit of the total rotation-in length L is mm; When the rotation angle θ changes with time, the number of rotation-in circles is dynamically updated to:

3. A two- and three-dimensional radar-vision fusion method for synchronizing the torque, number of turns, and length of a twisted target, according to claim 1, characterized in that In the step S2, the meshing length calculation is based on the product of the number of rotation-in circles N and the pitch P, and dynamically updates the depth L of the male thread entering the female thread; The calculation formula is: When the number of rotation-in circles N changes with time, the meshing length is dynamically updated to:

4. A method for synchronizing torque - number of turns - length of a two - three - dimensional radar - vision fusion twisting target according to claim 1, characterized in that, In the step S2, the torque synchronous display is to integrate the torque sensor data and map it to the 3D model and the curve graph in real time; the calculation formula of the torque is: where, V is the output voltage of the sensor, unit: volt; S is the sensor sensitivity, unit: volt / N·m; T is the torque, unit: N·m; The torque data can be plotted as a curve in real time through the time series T(t): The formula is:

5. A two- and three-dimensional radar-vision fusion twisting target torque-turns-length synchronization method according to claim 1, characterized in that In the step S2, the parameter error correction includes the correction of the number of rotation-in circles, the correction of the meshing length, and the zero-drift correction of the torque; among them, The calculation formula for the correction of the number of rotation-in circles is: where Δθ is the error of the angle sensor, unit: degree; N′ is the number of circles after correction; The calculation formula for the correction of the meshing length is: where ΔP is the mechanical error of the pitch, unit: mm / circle; L′ is the length after correction; The calculation formula for the zero-drift correction of the torque is: where ΔV is the zero-drift voltage of the sensor, unit: volt; T′ is the torque after correction.

6. A two - three - dimensional radar - vision fusion method for synchronizing the torque - rotation - number - length of a twisting target according to claim 1, characterized in that In the step 2, the data filtering processing includes moving average filtering and low-pass filtering; among them, The moving average filtering process takes the average value of n consecutive sampling points: Filtering of the number of rotation-in circles: Filtering of the torque: The low-pass filtering process uses a first-order low-pass filter to suppress high-frequency noise: y k = α · x k + (1 - α) · y k-1 ; where α is the filtering coefficient, and the value range is 0 < α < 1; x k is the current input data; y k is the current output data.

7. A two- and three-dimensional radar-vision fusion method for synchronizing the torque, number of turns, and length of a twisted target, according to claim 1, characterized in that In the step 2, when the direction perception and symbol processing support forward and reverse rotation, a rotation direction parameter D ∈ {+1, -1} needs to be introduced; Symbol correction of the number of rotation-in circles: Symbol correction of the meshing length:

8. A two- and three-dimensional radar-vision fusion method for synchronizing the torque, number of turns, and length of a twisted target according to claim 1, characterized in that In step 2, when performing multi-cycle cumulative calculation, if the angle sensor is an incremental encoder, it is necessary to accumulate the angles of multiple turns: θ total = θ current + 360°·k; Total precession length:

9. A two- and three-dimensional radar-vision fusion method for synchronizing the torque, number of turns, and length of a twisting target according to claim 1, characterized in that In step S1, the material of the female joint uses the MeshPhongMaterial material of Three.js; the material of the male joint uses the MeshLambertMaterial material.

10. A two - three - dimensional radar - vision fusion method for synchronizing the torque - number of turns - length of a twisting target according to claim 1, characterized in that In step S3, the specific processing method is: Front view: The three-dimensional model dynamically displays the precession process, and the female joint is semi-transparent to show the meshing state; Side view: Draw the torque-turns-time curve for data playback and analysis; Parameter panel: Real-time display of the current torque value, cumulative turns, and meshing length.

Images