CN114718546A - A Novel Spatially Distributed IMU-Based Pose Adjustment Method for Anti-scour Drilling Robot - Google Patents

Abstract

The invention discloses a position and posture adjusting method of an anti-impact drilling robot with a novel spatial distribution IMU (inertial measurement Unit), which comprises the anti-impact drilling robot, an inertia unit group, a regular octahedral structure frame and an inertia unit. The invention has the beneficial effects that: six group data that will gather fuse the processing to leading-in circulation neural network trains, obtains the prediction model, and the model that can make the training out like this is more reliable, and the displacement and the gesture of prediction scour protection drilling robot are more accurate and carry out real-time error compensation, thereby realize the automatically regulated of scour protection drilling robot displacement and gesture, improve drilling efficiency, realize the automation of release process.

Description

A Novel Spatially Distributed IMU-Based Pose Adjustment Method for Anti-scour Drilling Robot

technical field

The invention relates to a method for adjusting the position and attitude of an anti-scouring drilling robot, in particular to a method for automatically adjusting the position and attitude of an anti-scouring drilling robot based on a novel spatially distributed inertial unit (IMU), which belongs to the technical field of coal mining.

Background technique

Coal is one of the main energy sources used by human beings. my country has a large amount of coal storage, but in the process of mining, it will cause many hazards due to rock bursts. Moreover, with the increase in the depth and intensity of coal mining in my country, the frequency and frequency of rock bursts will increase. The damage strength is also increasing, so the anti-scouring drilling robot is required to carry out the construction of the anti-scouring pressure relief hole to ensure the safety of mining. The traditional anti-scour drilling robot usually adjusts its attitude at a fixed position, and at the same time manually observes the operation of the frame, resulting in low drilling efficiency. In order to improve the drilling efficiency and realize the automation of the pressure relief process, the precise displacement and posture of the anti-punch drilling robot is the basis for the automatic adjustment of the posture and posture.

Using the inertial unit (IMU) to solve the displacement and attitude of the object is a commonly used method. The inertial unit is a device that measures the three-axis attitude angle (or angular rate) and acceleration of the object, and its data update rate is high. The short-term accuracy and stability are good, but due to the integration of the obtained information, the positioning error increases with time, the long-term accuracy is poor, and the effect of a single inertial unit is not good.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a new method for adjusting the position and attitude of an anti-scour drilling robot with a novel spatially distributed IMU in order to solve the above problems.

The present invention achieves the above objects through the following technical solutions: a novel method for adjusting the position and attitude of an anti-scour drilling robot with a novel spatially distributed IMU, comprising:

An anti-scouring drilling robot, which is used for drilling construction of anti-scouring and pressure relief holes in coal mines, and adopts a drilling robot with a model of ZYWL-4000Y;

An inertial unit group, which is fixedly installed on the anti-scouring drilling robot for predicting the displacement and attitude of the anti-scouring drilling robot and performing real-time error compensation, and includes a regular octahedron structure frame, which is fixedly arranged on the regular eight The inertial unit on the surface structure frame, and the inertial unit is selected from the BW-IMU400 high-performance inertial measurement sensor of Wuxi Beiwei Sensing Technology Co., Ltd., including MEMS accelerometer and gyroscope;

Wherein, the inertial unit 22 measures the attitude parameters (roll angle, pitch angle, angular velocity, acceleration) of the moving carrier. Attitude and angular velocity deviations are optimally estimated by a 6-state Kalman filter with appropriate gain, which is suitable for inclination measurement in motion or vibration. Through various compensations such as nonlinear compensation, quadrature compensation, temperature compensation and drift compensation, the error source can be greatly eliminated and the measurement accuracy level can be improved. At the same time, BW-IMU400 has a digital interface, which can be easily integrated into the user's system to facilitate the collection of experimental data, so that the displacement and attitude information of the anti-scouring drilling robot can be collected in real time, and the data can be fed back to the upper computer. Follow up.

Its adjustment method includes the following steps:

Step 1: Collect the inertial unit data of the inertial unit group at a certain moment for multiple times to obtain multiple sets of data, input a large amount of experimental data to obtain a trained recurrent neural network model, and set an error threshold, and repeat the experiment multiple times. In order to collect a large amount of experimental data, prepare for training the recurrent neural network model;

Step 2: After the trained model is obtained, at time t, the collected six-way data is fused and imported into the trained model to obtain an error value. At time t, a new set of data is collected for the purpose of Verify the accuracy of the recurrent neural network model trained in step 1;

Step 3. Compare the obtained error value with the set error threshold. If the error value is not greater than the set error threshold, continue to collect the data at the next time t+1, and make an error with the error generated at the time t. Accumulate, and then compare it with the set error threshold. If it is still not greater than the error threshold, continue to collect the data at the next time t+2 for error accumulation, and repeat this cycle until the accumulated error generated is greater than the set error threshold, then end the process. After the adjustment is made, the next round of automatic adjustment of the position and posture is carried out. Since the error of the anti-scouring drilling robot is generated at any time during the working process, the setting of the error threshold is that when the error accumulation is too large, the robot must be adjusted. Automatic adjustment to compensate for the resulting error.

As a further solution of the present invention, the inertial units are respectively located at the six vertices of the regular octahedral structure frame, and can collect six sets of data at the same time.

As a further solution of the present invention: the distances from the six inertial units to the center position of the regular octahedral structure frame are the same, and are set at 1:1:1:1:1:1, so that the measurement data of each inertial unit has the same proportion, Therefore, according to the ratio of 1:1:1:1:1:1, the six-channel data is fused and processed to make the measurement more accurate.

As a further solution of the present invention, the inertial unit group is located at the center of gravity of the anti-scour drilling robot.

As a further scheme of the present invention: in the first step, the data is imported into the cyclic neural network for training to obtain a prediction model, which can make the trained model more reliable and predict the displacement and attitude of the anti-scouring drilling robot. It is more accurate and performs real-time error compensation, so as to realize the automatic adjustment of the displacement and attitude of the anti-scouring drilling robot, improve the drilling efficiency, and realize the automation of the pressure relief process.

The beneficial effect of the invention is that: the collected six groups of data are fused, and imported into a cyclic neural network for training to obtain a prediction model, which can make the trained model more reliable, and predict the displacement and The posture is more accurate and real-time error compensation is performed, so as to realize the automatic adjustment of the displacement and posture of the anti-punch drilling robot, improve the drilling efficiency, and realize the automation of the pressure relief process.

Description of drawings

1 is a schematic structural diagram of an anti-scouring drilling robot of the present invention;

2 is a schematic diagram of the distribution structure of the inertial unit of the present invention;

FIG. 3 is a schematic diagram of the automatic adjustment process of the position and posture of the anti-scouring drilling robot of the present invention.

In the picture: 1. Anti-scour drilling robot, 2. Inertial unit group, 21. Regular octahedral structure frame, 22. Inertial unit.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Example 1

Please refer to Figures 1~3, a new type of spatially distributed IMU's anti-punch drilling robot pose adjustment method, including

A punch-proof drilling robot 1, which is used for drilling construction of punch-proof and pressure relief holes in a coal mine, and adopts a drilling robot with a model of ZYWL-4000Y;

The inertial unit group 2 is fixedly installed on the anti-scour drilling robot 1 for predicting the displacement and attitude of the anti-scouring drilling robot 1 and performing real-time error compensation, and it includes a regular octahedral structure frame 21, a fixed arrangement The inertial unit 22 on the regular octahedral structure frame 21 is selected from the BW-IMU400 high-performance inertial measurement sensor of Wuxi Beiwei Sensing Technology Co., Ltd., including a MEMS accelerometer and a gyroscope.

In the embodiment of the present invention, the inertial units 22 are respectively located at six vertices of the regular octahedral structure frame 21, and can collect six sets of data at the same time.

In the embodiment of the present invention, the distances from the six inertial units 22 to the center position of the regular octahedral structure frame 21 are the same, and are set at 1:1:1:1:1:1, so that the measurement data of each inertial unit 22 accounts for The same, so the six-channel data is processed in a ratio of 1:1:1:1:1:1 to make the measurement more accurate.

In the embodiment of the present invention, the inertial unit group 2 is located at the center of gravity of the anti-scouring drilling robot 1 .

Embodiment 2

Please refer to Figure 3, a new type of spatially distributed IMU anti-scour drilling robot pose adjustment method, the adjustment method includes the following steps:

Step 1: Collect the data of the inertial unit 22 of the inertial unit group 2 at a certain moment for multiple times to obtain multiple sets of data, input a large amount of experimental data to obtain a trained recurrent neural network model, and set an error threshold, repeat multiple times The experiment is to collect a large amount of experimental data and prepare for training the recurrent neural network model, because theoretically enough samples can train a good recurrent neural network model;

Step 2: After the trained model is obtained, at time t, the collected six-way data is fused and imported into the trained model to obtain an error value. At time t, a new set of data is collected for the purpose of To verify the accuracy of the cyclic neural network model trained in step 1, it is impossible to have no error, so there is an error value;

Step 3. Compare the obtained error value with the set error threshold. If the error value is not greater than the set error threshold, continue to collect the data at the next time t+1, and make an error with the error generated at the time t. Accumulate, and then compare it with the set error threshold. If it is still not greater than the error threshold, continue to collect the data at the next time t+2 for error accumulation, and repeat this cycle until the accumulated error is greater than the set error threshold, then end the process. After the adjustment is made, the next round of automatic adjustment of the position and posture is carried out. Since the error of the anti-scouring drilling robot is generated at any time during the working process, the setting of the error threshold is that when the error accumulation is too large, the robot must be adjusted. Automatic adjustment to compensate for the resulting error. The purpose is to make the displacement and attitude deviation of the anti-scouring drilling robot can be adjusted automatically during the working process to eliminate the generated errors.

In the embodiment of the present invention, in the first step, the data is imported into the cyclic neural network for training to obtain the prediction model, which can make the trained model more reliable, and predict the displacement and posture of the anti-scouring drilling robot more accurately. Accurate and real-time error compensation, so as to realize the automatic adjustment of the displacement and attitude of the anti-collision drilling robot, improve the drilling efficiency, and realize the automation of the pressure relief process.

Embodiment 3

The selected inertial unit (IMU) is BW-IMU400, which can measure the attitude parameters of the motion carrier (roll angle, pitch angle, angular velocity, acceleration). 0x05 and 0x06, all the signal lines of the six inertial sensing units that constitute the inertial unit group are connected to the converter, and the converter is connected to the host computer through the USB interface to realize data transmission.

Collect the data of the inertial unit through the host computer, and fuse the collected data, import the fused data into the calculation algorithm for pose calculation, output the displacement and attitude parameters of the anti-scouring drilling robot, and save the experimental data . Repeat this process to obtain a large amount of experimental data.

The obtained experimental data is imported into the cyclic neural network for training, and the well-trained neural network is used to predict the displacement and attitude error of the anti-scouring drilling robot output by the above solution algorithm. The calculation error of the displacement and attitude of the drilling robot is compensated to obtain more accurate displacement and attitude parameters of the anti-scouring drilling robot, so as to perform automatic adjustment of the position and attitude.

Design multiple inertial units to collect data and perform fusion processing, collect a large amount of experimental data, and import the data into a recurrent neural network for training to obtain a prediction model. Substitute the new data into the obtained prediction model, and perform real-time error compensation on the displacement and attitude of the anti-scouring drilling robot, so as to realize the automatic adjustment of the displacement and attitude of the anti-scouring drilling robot.

It will be apparent to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, but that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments are to be regarded in all respects as illustrative and not restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, which are therefore intended to fall within the scope of the appended claims. All changes within the meaning and range of the equivalents of , are included in the present invention. Any reference signs in the claims shall not be construed as limiting the involved claim.

In addition, it should be understood that although this specification is described in terms of embodiments, not each embodiment only includes an independent technical solution, and this description in the specification is only for the sake of clarity, and those skilled in the art should take the specification as a whole , the technical solutions in each embodiment can also be appropriately combined to form other implementations that can be understood by those skilled in the art.

Claims (5)

1. a kind of anti-scour drilling robot position and attitude adjustment method of novel spatial distribution IMU, is characterized in that: comprising: An anti-scour drilling robot (1), which is used for drilling construction of anti-scour and pressure relief holes in a coal mine; An inertial unit group (2), which is fixedly installed on the anti-scour drilling robot (1) for predicting the displacement and attitude of the anti-scouring drilling robot (1) and performs real-time error compensation, and includes a regular octahedron a structural frame (21), an inertial unit (22) fixedly arranged on the regular octahedral structural frame (21), and the inertial unit (22) includes a MEMS accelerometer and a gyroscope; Its adjustment method includes the following steps: Step 1: Collect the data of the inertial unit (22) of the inertial unit group (2) at a certain moment for many times to obtain multiple sets of data, input a large amount of experimental data to obtain a trained recurrent neural network model, and set an error threshold , the experiment is repeated many times in order to collect a large amount of experimental data and prepare for training the recurrent neural network model; Step 2. After the trained model is obtained, at time t, the collected six-way data are fused and imported into the trained model to obtain an error value. At time t, a new set of data is collected for the purpose of Verify the accuracy of the recurrent neural network model trained in step 1; Step 3. Compare the obtained error value with the set error threshold. If the error value is not greater than the set error threshold, continue to collect the data at the next time t+1, and make an error with the error generated at the time t. Accumulate, and then compare it with the set error threshold. If it is still not greater than the error threshold, continue to collect the data at the next time t+2 for error accumulation, and repeat this cycle until the accumulated error is greater than the set error threshold, then end the process. After the adjustment is made, the next round of automatic adjustment of the pose will be carried out. Since the error of the anti-scouring drilling robot is generated at any time during the working process, the setting of the error threshold is that when the error accumulation is too large, the robot will be adjusted. Automatic adjustment to compensate for the resulting error. 2 . The method for adjusting the position and attitude of a new type of space-distributed IMU for anti-scouring drilling robots according to claim 1 , wherein the inertial units ( 22 ) are respectively located at the six vertices of the regular octahedral structure frame ( 21 ). 3 . place. 3. The method for adjusting the position and attitude of a new type of space-distributed IMU for an anti-scouring drilling robot according to claim 2, wherein the six inertial units (22) are located at the center of the regular octahedral structure frame (21). The same distances are set for 1:1:1:1:1:1. 4. The method for adjusting the position and attitude of a new type of space-distributed IMU for an anti-scouring drilling robot according to claim 1, wherein the inertial unit group (2) is located at the center of gravity of the anti-scouring drilling robot (1). place. 5 . The method for adjusting the position and attitude of a new type of space-distributed IMU anti-scour drilling robot according to claim 1 , wherein in the first step, data is imported into a recurrent neural network for training to obtain a prediction model. 6 .

Images