CN-121974253-A - Method and computing equipment for accurately positioning tail end of mechanical arm in real time

Abstract

The application provides a method and computing equipment for accurately positioning the tail end of a mechanical arm in real time, wherein the method comprises the steps of collecting a real-time control signal sequence for the mechanical arm in real time and collecting real-time IMU data of the tail end of the mechanical arm at the same time; the real-time control signal sequence is input into a pre-constructed mapping model to obtain real-time IMU prediction data output by the mapping model, the real-time IMU prediction data are fused with corresponding real-time IMU data to obtain corrected IMU data, and the real-time position of the tail end of the mechanical arm in the motion process is calculated and output according to an initial datum point and the corrected IMU data. According to the technical scheme, the real-time accurate positioning of the tail end of the suspension arm in different scenes can be realized through model correction by combining control signals, inertial measurement data and static positioning data.

Inventors

• LIU BO
• WANG ZHIWEI

Assignees

• 北京智能人工科技有限公司

Dates

Publication Date

20260505

Application Date

20260126

Claims (10)

1. 1. A method for real-time accurate positioning of a robot arm end, the method comprising: Acquiring a real-time control signal sequence for the mechanical arm in real time, and simultaneously acquiring real-time IMU data of the tail end of the mechanical arm; inputting the real-time control signal sequence into a pre-constructed mapping model to obtain real-time IMU prediction data output by the mapping model; Fusing the real-time IMU prediction data with corresponding real-time IMU data to obtain corrected IMU data; and calculating and outputting the real-time position of the tail end of the mechanical arm in the motion process according to the initial datum point and the corrected IMU data.
2. 2. The method of claim 1, wherein the control signal sequence comprises at least one of signal strength, duration, and direction of motion information.
3. 3. The method of claim 1, wherein the IMU data comprises acceleration, angular velocity, and/or attitude angle data.
4. 4. The method of claim 1, wherein the pre-constructed mapping model employs an "encoder-decoder" architecture, wherein the encoder employs a Bi-directional long-short-term memory network Bi-LSTM and the decoder employs a long-short-term memory network LSTM.
5. 5. The method of claim 1, wherein the robotic arm comprises a crane boom.
6. 6. The method of claim 5, wherein acquiring real-time control signal sequences for the robotic arm in real-time while acquiring real-time IMU data for the robotic arm tip, comprises: analog control signals of left and right swinging, telescopic boom, lifting boom and lifting hook lifting of the boom corresponding to the four independent control rockers on the remote controller of the crane are collected and converted into control signal sequences of each shaft; and simultaneously acquiring real-time IMU data of the tail end of the suspension arm.
7. 7. The method of claim 6, wherein the pre-built mapping model is built by: Acquiring multi-source data in the crane operation process, constructing a plurality of groups of sample data comprising the control signal sequence, the IMU data corresponding to the time stamp and UWB static coordinates of a starting point and an ending point, and obtaining a sample data set; Preprocessing the control signal sequence and the IMU data in the sample data set, and completing IMU data calibration by combining the UWB static coordinates to construct an input-output matched data set; training a time sequence to a time sequence model by utilizing the constructed data set matched with the input and the output, and obtaining a mapping model which outputs a corresponding IMU data time sequence result according to the input control signal time sequence data.
8. 8. The method of claim 1, wherein fusing the real-time IMU prediction data with corresponding real-time IMU data comprises: And correcting accumulated errors in corresponding real-time IMU data by using the real-time IMU prediction data by adopting a Kalman filtering algorithm to obtain corrected IMU data.
9. 9. The method of claim 1, wherein calculating and outputting the real-time position of the robot tip during movement based on the initial fiducial point and the corrected IMU data comprises: Collecting a starting point UWB static coordinate of the tail end of the mechanical arm as the initial reference point of position calculation when each control instruction is started; performing secondary integration on the acceleration in the corrected IMU data to obtain real-time displacement of the tail end of the mechanical arm on each coordinate axis; Combining the attitude angle in the corrected IMU data, and carrying out coordinate conversion on the real-time displacement to obtain real-time displacement data under a world coordinate system; And superposing the UWB static coordinates of the initial datum point and the real-time displacement data to obtain the real-time position coordinates of the tail end of the mechanical arm in the motion process.
10. 10. A computing device, comprising: processor, and A memory storing a computer program which, when executed by the processor, causes the processor to perform the method of any one of claims 1-9.

Description

Method and computing equipment for accurately positioning tail end of mechanical arm in real time Technical Field The invention relates to the technical field of crane control and positioning, in particular to a method and computing equipment for accurately positioning the tail end of a mechanical arm in real time, which are particularly suitable for engineering operation scenes with high requirements on the position accuracy of the tail end of a suspension arm. Background The crane is used as core equipment in the fields of engineering construction, logistics hoisting and the like, and the real-time accurate positioning of the tail end of the suspension arm directly relates to the working efficiency, the construction safety and the working quality. In the scenes of precise hoisting, narrow space operation and the like, the positioning accuracy requirement on the tail end position of the suspension arm usually reaches the centimeter level, however, the traditional crane suspension arm tail end positioning method is difficult to meet the requirement. At present, the traditional real-time positioning method for the tail end of the crane boom mainly adopts an Inertial Measurement Unit (IMU) to position, and can acquire acceleration, angular speed and other data of the boom in real time and calculate the position through integration, but the IMU has the problem of accumulated errors, and errors can be continuously overlapped along with the extension of the operation time, so that the positioning precision is rapidly reduced, and the accurate positioning cannot be ensured for a long time. Or GPS/UWB positioning is adopted, but the GPS has weak or even lost signals in the indoor and severely shielded construction sites and other scenes, and continuous positioning can not be realized. While Ultra Wideband (UWB) technology has higher positioning accuracy in a short distance, only can acquire discrete coordinate points in a static or low-speed motion state, cannot provide real-time continuous positioning data in a motion process, and is easily influenced by factors such as on-site metal structures, electromagnetic interference and the like. And the traditional method has lag positioning response and unstable precision. Particularly in the load change or complex environment, the positioning accuracy is obviously reduced. Therefore, a technical scheme is needed, and real-time accurate positioning of the tail end of the suspension arm in different scenes can be realized through model correction by combining control signals, inertial measurement data and static positioning data. Disclosure of Invention The application aims to provide a method and computing equipment for real-time accurate positioning of a tail end of a mechanical arm, which can combine control signals, inertial measurement data and static positioning data to realize real-time accurate positioning of the tail end of a suspension arm in different scenes through model correction. According to an aspect of the present application, there is provided a method for real-time accurate positioning of a robot arm end, the method comprising: Acquiring a real-time control signal sequence for the mechanical arm in real time, and simultaneously acquiring real-time IMU data of the tail end of the mechanical arm; inputting the real-time control signal sequence into a pre-constructed mapping model to obtain real-time IMU prediction data output by the mapping model; Fusing the real-time IMU prediction data with corresponding real-time IMU data to obtain corrected IMU data; and calculating and outputting the real-time position of the tail end of the mechanical arm in the motion process according to the initial datum point and the corrected IMU data. According to some embodiments, the control signal sequence comprises at least one of signal strength, duration and direction of motion information. According to some embodiments, the IMU data includes acceleration, angular velocity, and/or attitude angle data. According to some embodiments, the pre-built mapping model employs an "encoder-decoder" architecture, wherein the encoder employs a Bi-directional long-short-term memory network Bi-LSTM and the decoder employs a long-short-term memory network LSTM. According to some embodiments, the robotic arm comprises a crane boom. According to some embodiments, acquiring real-time control signal sequences for the robotic arm in real-time while acquiring real-time IMU data for the robotic arm tip includes: analog control signals of left and right swinging, telescopic boom, lifting boom and lifting hook lifting of the boom corresponding to the four independent control rockers on the remote controller of the crane are collected and converted into control signal sequences of each shaft; and simultaneously acquiring real-time IMU data of the tail end of the suspension arm. According to some embodiments, the pre-built mapping model is built by: Acquiring multi-source data in the crane opera