Underwater unmanned underwater vehicle recovery method based on multi-element detector

By combining multi-element detectors and neural network models, the problem of limited recovery distance for underwater unmanned submersibles has been solved, achieving high-precision optical guidance and anti-interference capabilities, and improving the recovery success rate.

CN118936509APending Publication Date: 2024-11-12XIAN INST OF OPTICS & PRECISION MECHANICS CHINESE ACAD OF SCI
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202410988100.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-23
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Existing underwater unmanned vehicle recovery technologies have limited operating range and low recovery success rates.

Method used

An optical guidance method based on multiple detectors is adopted. By encoding the LED light source and training the neural network model, the angular relationship between the multiple photodetectors and the LED light source is measured to achieve high-precision guidance between the UUV and the recovery cage.

Benefits of technology

It improves the success rate of UUV recovery, overcomes the shortcomings of acoustic ranging, enhances ranging accuracy and anti-interference ability, and simplifies underwater ranging methods.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118936509A_ABST
    Figure CN118936509A_ABST
Patent Text Reader

Abstract

The invention discloses an unmanned underwater vehicle (UUV) recovery method based on a multi-element detector, and solves the problems of limited operating distance and low UUV recovery success rate of the existing underwater recovery technical method. According to the invention, based on the optical guidance mode of the multi-element photoelectric detector, the angle relation between the multi-element photoelectric detector assembly and the LED light source is measured, the neural network model training is carried out, the relation between the angle, the side length and the like and the distance and azimuth information is obtained, and the distance and azimuth information between the UUV and the UUV recovery cage is accurately measured. The high-precision guidance of the UUV recycling process is ensured; meanwhile, the LED light source is coded, interference of external pulses and induction signals can be effectively avoided, and the UUV recycling success rate is increased. The method can also be used for other measurement equipment with high distance measurement precision requirements.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to an underwater recovery technology method, and in particular to an underwater unmanned submersible recovery method based on a multi-element detector. Background Art

[0003] Through the autonomous recovery of underwater unmanned submersibles, energy replenishment, data transmission, and task arrangement of underwater unmanned mobile platforms can be realized, so as to achieve longer concealed working time, higher data detection capabilities, and a larger range of action. Therefore, how to effectively guide underwater unmanned platforms so that they can complete autonomous recovery is the key and forefront of the development of ocean exploration and unmanned platform technology.

[0004] Common underwater recovery technology methods mainly include acoustic guidance technology and optical guidance technology. Acoustic guidance technology mainly uses ultra-short baseline hydroacoustic positioning system (USBL) and ultra-short baseline and long baseline combination to carry out research and experiments on long-distance acoustic guidance. It has the advantages of long guidance distance and omnidirectional guidance. However, the non-uniformity and dynamic variability of the marine environment medium lead to significant nonlinear effects of long-distance sound wave transmission routes, thus deviating from the linear propagation assumption of the signal of the ultra-short baseline positioning principle; the slow update frequency, environmental sound field and platform noise all limit the positioning accuracy of the ultra-short baseline, making it difficult to further improve the performance of long-distance hydroacoustic positioning. Optical guidance technology senses the light source signal of the recovery station through a multi-element photoelectric detector to realize the posture measurement of UUV (underwater unmanned submersible). Its working distance is limited, but the guidance accuracy is high, the ability to resist external interference is strong, and the robustness is strong. It is suitable for short-distance high-precision guidance and recovery. Summary of the invention

[0005] The purpose of the present invention is to provide an underwater unmanned submersible recovery method based on a multi-element detector to solve the technical problems of the limited action range of existing underwater recovery technical methods and the low recovery success rate of UUVs.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A method for recovering an underwater unmanned submersible based on a multi-element detector is special in that it includes the following steps:

[0008] Step 1, dividing the LED light sources into L1 group light sources and L2 group light sources;

[0009] The L1 group of light sources includes n or more light sources, n≥3, which are installed at the mouth of the UUV recovery cage to form a polygon; the side lengths of the polygon are m1, m2, ..., mn, and are used for the angle and distance measurement of the multi-detector of the UUV;

[0010] The L2 group of light sources includes one light source, which is installed at the center of the tail of the UUV recovery cage and forms a polyhedron with the L1 group of light sources for close-range field of view switching and angle measurement of the multi-detector;

[0011] Step 2: Encode the LED light sources at different installation positions, and the LED light sources light up and turn off at high frequencies in a preset frequency hopping manner according to the codes;

[0012] Step 3: The multi-element detector collects and decodes the light information emitted by the LED light source, and then calculates the position information of the UUV relative to the UUV recovery cage where the LED light source is located, so as to adjust the position of the UUV so that it can enter the UUV recovery cage smoothly.

[0013] Furthermore, the step 3 specifically includes:

[0014] When the L1 group of light sources is within the field of view of the UUV's multi-detector, the multi-detector collects and decodes the light information emitted by the L1 group of light sources, calculates the horizontal and vertical angle information between the UUV and the L1 group of light sources through signal processing, and then obtains the distance and angle information between the UUV and the UUV recovery cage where the L1 group of light sources are located through the trained neural network model based on the horizontal and vertical angle information and the polygon side lengths m1, m2, ..., mn. The UUV adjusts its direction according to the calculated angle and distance information and enters the UUV recovery cage.

[0015] When the UUV is so close to the recycling station that the L1 group of light sources is out of the field of view of the multi-detector, the angle information between the UUV and the L2 group of light sources is calculated by the signal processing module. The UUV adjusts its direction according to the calculated angle information and enters the UUV recycling cage.

[0016] Furthermore, the encoding in step 2 specifically includes:

[0017] Perform m pseudo-random coding on the light sources of group L1 and group L2. Suppose the initial state of m pseudo-random coding is a 14 , a 13 , ...a1, a0 is 000111101011001, each shift, the input of the first end on the left side of the register A total of 32767 states are generated, and n+1 states are randomly selected as the codes of n+1 LED light sources.

[0018] Furthermore, the decoding in step 3 is specifically:

[0019] The autocorrelation algorithm is used to analyze the optical information received and collected by the multi-element photodetector assembly. If there are multiple non-zero values, return to step 2 to change the preset frequency hopping rule to avoid interference / induction of the data.

[0020] Furthermore, the calculation of the posture information corresponding to the LED light source in step 3 specifically includes:

[0021] When the light information of the light source of group L1 is used for calculation, steps 3.1 to 3.3 are adopted;

[0022] When the light information of the light source of group L2 is used for calculation, steps 3.1 to 3.2 are adopted;

[0023] The steps 3.1, 3.2 and 3.3 are as follows:

[0024] 3.1) Calculate the horizontal and vertical angles between each LED light source and the UUV θ;

[0025]

[0026] Among them, K x , K y is the calibration factor, m x 、m y is the compensation value, A1, A2, A3, A4 are the voltage signal values ​​of the four pixels of the multi-element photodetector;

[0027] 3.2) Calculate the offset position Dirx and Diry between the multi-element detector and the LED light source:

[0028]

[0029] Dirx>0 means that the LED light source is on the right side of the center point of the multi-element detector, that is, the positive direction of the X axis, and Diry>0 means that the LED light source is on the lower side of the center point of the multi-element detector, that is, the negative direction of the Z axis;

[0030] 3.3) Use the neural network model to train the relationship between input and output. The input is the horizontal and vertical angles between the multi-detector and the L1 group of light sources, as well as the side length of the polygon formed by the L1 group of light sources. The output is the yaw angle, pitch angle, and distance information between the multi-detector and the UUV recovery cage. Use the trained neural network model to calculate the position information corresponding to the LED light source.

[0031] Furthermore, the calculation of the posture information corresponding to the LED light source in step 3 also includes:

[0032] When the UUV does not obtain the light information of all LED light sources, the UUV is controlled to approach the central axis of the recovery cage according to the obtained light information through the calculation results of steps 3.1 to 3.3 until the light information of all LED light sources can be received, and then the distance information is calculated.

[0033] Beneficial effects of the present invention:

[0034] 1. The present invention is based on the optical guidance method of the multi-element photoelectric detector. By measuring the angular relationship between the multi-element photoelectric detector assembly and the LED light source, the neural network model is trained to obtain the relationship between the angle, side length, etc. and the distance and orientation information, and the distance and orientation information between the UUV and the UUV recovery cage are accurately measured to ensure high-precision guidance of the UUV recovery process; at the same time, the present invention encodes the LED light source, which can effectively avoid the interference of external pulses and induced signals, and improve the success rate of UUV recovery. This method can also be used for other measuring equipment with high ranging accuracy requirements.

[0035] 2. The present invention uses optical guidance to encode the light source, including but not limited to pseudo-random coding, and controls the high-frequency on and off time of the LED light source by frequency hopping, which can effectively avoid the influence of external interference pulses and induced pulses and improve data validity.

[0036] 3. The present invention utilizes neural network model training to accurately obtain the relationship between angle and posture and improve measurement accuracy.

[0037] 4. The present invention overcomes the shortcomings of the traditional distance measurement method using sound waves as carriers, solves the problems of high power consumption and large volume of sound waves, simplifies the underwater distance measurement method, and improves the distance measurement accuracy;

[0038] 5. The present invention increases the field of view of the photoelectric detector by providing two groups of LED light sources. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a system structure diagram of an embodiment of an underwater unmanned submersible recovery method based on a multi-element detector of the present invention;

[0040] Figure 2 Schematic diagram of the angle of the multi-element photoelectric detector of the UUV and the LED light source on the UUV recovery cage in the embodiment of the present invention;

[0041] Figure 3 yes Figure 2 The right side structural diagram of the LED light source;

[0042] Figure 4 Schematic diagram of m pseudo-random coding in an embodiment of the present invention;

[0043] Figure 5 is a graph of an autocorrelation function in the absence of interference / induction pulses in an embodiment of the present invention;

[0044] Figure 6 is a graph of an autocorrelation function in the presence of an interference / induced pulse in an embodiment of the present invention;

[0045] Figure 7is a schematic diagram of frequency hopping in an embodiment of the present invention;

[0046] Figure 8 is the angle between the LED light source and the UUV in the horizontal and vertical directions in the embodiment of the present invention Schematic diagram of θ (Led' is the projection of Led). DETAILED DESCRIPTION

[0047] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] An embodiment of the present invention provides a method for recovering an underwater unmanned submersible based on a multi-element detector, such as Figure 1 , Figure 2 As shown, the recycling method comprises the following steps:

[0049] Step 1, dividing the LED light sources into L1 group light sources and L2 group light sources;

[0050] The L1 group of light sources contains more than n light sources, n ≥ 3, which are installed at the mouth of the UUV recovery cage with a radius of r to form a polygon; the polygon side lengths are m1, m2, ..., mn, which are used for the UUV's multi-detector angle and distance measurement; the multi-detector is installed at the front end position inside the UUV to receive the LED light source coding information.

[0051] The L2 light source group includes one light source, which is installed at the center of the tail of the UUV recovery cage and forms a polyhedron with the L1 light source group for close-range field of view switching and angle measurement of the multi-detector;

[0052] In this embodiment, Figure 2 , 3 As shown, the light source group L1 includes n=3 light sources, and the lengths of the polygon formed are m1, m2, and m3;

[0053] Step 2, encode the LED light sources at different installation positions, including but not limited to m pseudo-random codes, and control the LED light sources to light up and turn off at high frequency in a preset frequency hopping manner according to the codes by the driving circuit;

[0054] The encoding specifically includes: performing m pseudo-random encoding on the light sources of group L1 and group L2, such as Figure 4 As shown, let m pseudo-randomly encode the initial state a 14 , a 13 , ...a1, a0 is 000111101011001, each shift, the input of the first end on the left side of the register (m is a variable, not a fixed value, such as a 14 、a 13 This formula is used for calculation), and a total of 32767 states are generated. Randomly select n+1 states as the codes of n+1 LED light sources. In this embodiment, randomly select 4 states as the codes of 4 LEDs.

[0055] Step 3: The multi-element detector collects light information emitted by the LED light source, which is decoded by the decoding module. The decoding process uses an autocorrelation algorithm (including but not limited to), and then calculates the position information of the UUV relative to the UUV recovery cage where the LED light source is located, so as to adjust the UUV so that it can enter the UUV recovery cage smoothly.

[0056] Specifically include:

[0057] When the L1 group of light sources is within the field of view of the UUV's multi-detector, the multi-detector collects and decodes the light information emitted by the L1 group of light sources, and calculates the horizontal and vertical angle information between the UUV and the L1 group of light sources through the signal processing module. Then, based on the horizontal and vertical angle information and the polygon side lengths m1, m2, ..., mn, the distance and angle information between the UUV and the UUV recycling cage where the L1 group of light sources is located is obtained through the trained neural network model. The UUV adjusts its direction according to the calculated angle and distance information and enters the recycling station;

[0058] When the UUV is so close to the recycling station that the L1 group of light sources is out of the field of view of the multi-element detector, the angle information between the UUV and the L2 group of light sources is calculated by the signal processing module, and the UUV adjusts its direction according to the calculated angle information and enters the recycling station.

[0059] The decoding is:

[0060] The MCU processing unit uses an autocorrelation algorithm to analyze the light information received and collected by the multi-element photodetector assembly. If there are multiple non-zero values, it returns to step 2 to change the preset frequency hopping rules to avoid interference / induction of the data.

[0061] Specifically, Figure 5 As shown in FIG. 1 , if there is no interference / induced pulse, the autocorrelation characteristics of the received information and the four states of the transmission are good, and the corresponding relationship between the LED and the received information can be directly analyzed according to the degree of autocorrelation between the received information and the four states; Figure 6 As shown in Figure 1, in the presence of interference / induced pulses, the autocorrelation function deteriorates and there are multiple non-zero values, proving that this set of data is affected by interference / induced pulses, such as Figure 7 As shown, return to step 2 to change the preset frequency hopping rule to avoid the influence of interference / induction.

[0062] Calculating the posture information corresponding to the LED light source specifically includes:

[0063] When the light information of the light source of group L1 is used for calculation, steps 3.1 to 3.3 are adopted;

[0064] When the light information of the light source of group L2 is used for calculation, steps 3.1 to 3.2 are adopted;

[0065] The steps 3.1, 3.2 and 3.3 are as follows:

[0066] 3.1) According to the signal size of the four quadrants of the multi-element detector, combined Figure 8 As shown, calculate the angle between each LED light source and the UUV in the horizontal and vertical directions θ;

[0067]

[0068] Among them, K x , K y is the calibration factor, m x 、m y is the compensation value, A1, A2, A3, A4 are the voltage signal values ​​of the four pixels of the multi-element photodetector;

[0069] 3.2) Calculate the offset position Dirx and Diry between the multi-element detector and the LED light source:

[0070]

[0071] Dirx>0 means that the LED light source is on the right side of the center point of the multi-element detector, that is, the positive direction of the X axis, and Diry>0 means that the LED light source is on the lower side of the center point of the multi-element detector, that is, the negative direction of the Z axis;

[0072] 3.3) Use the neural network model to train the relationship between input and output. The input is the horizontal and vertical angles between the multi-detector and the L1 group light source, as well as the side length of the polygon formed by the L1 group light source. The output is the yaw angle, pitch angle, and distance information between the multi-detector and the UUV recovery cage. Neural network training needs to be carried out after data collection in the laboratory.

[0073] The pose information corresponding to the LED light source is also calculated:

[0074] When the UUV does not obtain the light information of all LED light sources, the UUV is controlled to approach the central axis of the recovery cage according to the obtained light information through the calculation results of steps 3.1 to 3.3 until the light information of all LED light sources can be received, and then the distance information is calculated.

[0075] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for recovering an underwater unmanned submersible based on a multi-element detector. It is characterized in that The following steps are involved: Step 1: Divide the LED light source into L 1 Group light source, L 2 Group light source; The L 1 The group light source includes n or more light sources, n≥3, which are installed at the mouth of the UUV recovery cage to form a polygon; the side lengths of the polygon are m1, m2, ..., mn, which are used for the angle and distance measurement of the multi-detector of the UUV; The L 2 The light source group includes 1 light source, which is installed at the center of the tail of the UUV recovery cage. 1 The light sources are grouped to form a polyhedron, which is used for close-range field of view switching and angle measurement of multi-element detectors; Step 2: Encode the LED light sources at different installation positions, and the LED light sources light up and turn off at high frequencies in a preset frequency hopping manner according to the codes; Step 3: The multi-element detector collects and decodes the light information emitted by the LED light source, and then calculates the position information of the UUV relative to the UUV recovery cage where the LED light source is located, so as to adjust the position of the UUV so that it can enter the UUV recovery cage smoothly.

2. The underwater unmanned submersible recovery method based on a multi-element detector according to claim 1, It is characterized in that The step 3 specifically includes: When L 1 When the light source is within the field of view of the multi-detector of the UUV, the multi-detector collects L 1 The light information emitted by the group light source is decoded and the UUV and L are calculated through signal processing. 1 The horizontal and vertical angle information between the light sources of the group is obtained, and then based on the horizontal and vertical angle information and the polygon side lengths m1, m2, ..., mn, the UUV and L are obtained through the trained neural network model. 1 The distance and angle information of the UUV recovery cage where the light source is located is obtained. The UUV adjusts its direction according to the calculated angle and distance information and enters the UUV recovery cage. When the UUV is close to the recycling station, 1 When the light source of the group leaves the field of view of the multi-element detector, the signal processing module calculates the UUV and L 2 The angle information between the group light sources is calculated, and the UUV adjusts its direction according to the calculated angle information and enters the UUV recovery cage.

3. The underwater unmanned submersible recovery method based on a multi-element detector according to claim 1 or 2, It is characterized in that The coding in step 2 specifically includes: For L 1 Group light source and L 2 The group of light sources performs m pseudo-random coding, and the initial state of m pseudo-random coding is a 14 , a 13 , …a 1 , a 0 000111101011001. Each time the register is shifted, the input at the first end on the left side is A total of 32767 states are generated, and n+1 states are randomly selected as the codes of n+1 LED light sources.

4. The underwater unmanned submersible recovery method based on a multi-element detector according to claim 3, It is characterized in that The decoding in step 3 is specifically: The autocorrelation algorithm is used to analyze the optical information received and collected by the multi-element photodetector assembly. If there are multiple non-zero values, return to step 2 to change the preset frequency hopping rule to avoid interference / induction of the data.

5. The underwater unmanned submersible recovery method based on a multi-element detector according to claim 1, It is characterized in that Calculating the posture information corresponding to the LED light source in step 3 specifically includes: When using L 1 When calculating the light information of the group light source, use steps 3.1 to 3.3; When using L 2 When calculating the light information of the group light source, use steps 3.1 to 3.2; The steps 3.1, 3.2 and 3.3 are as follows: 3.1) Calculate the horizontal and vertical angles between each LED light source and the UUV θ; Among them, K x , K y is the calibration factor, m x 、m y is the compensation value, A 1 , A 2 , A 3 , A 4 is the voltage signal value of the four pixels of the multi-element photodetector; 3.2) Calculate the offset position Dirx and Diry between the multi-element detector and the LED light source: Dirx>0 means that the LED light source is on the right side of the center point of the multi-element detector, that is, the positive direction of the X axis, and Diry>0 means that the LED light source is on the lower side of the center point of the multi-element detector, that is, the negative direction of the Z axis; 3.3) Use the neural network model to train the relationship between input and output. The input is the multi-detector and L 1 The horizontal and vertical angles of the light source group, and L 1 The side length of the polygon formed by the group light source, the output is the yaw angle, pitch angle, and distance information of the multi-element detector and the UUV recovery cage; the trained neural network model is used to calculate the posture information corresponding to the LED light source.

6. The underwater unmanned submersible recovery method based on a multi-element detector according to claim 5, It is characterized in that Calculating the posture information corresponding to the LED light source in step 3 also includes: When the UUV does not obtain the light information of all LED light sources, the UUV is controlled to approach the central axis of the recovery cage according to the obtained light information through the calculation results of steps 3.1 to 3.3 until the light information of all LED light sources can be received, and then the distance information is calculated.

Citation Information

Cited By

  • Dynamic four-quadrant underwater high-speed guiding method

    CN120740610A

  • A dynamic four-quadrant underwater high-rate guidance method

    CN120740610B

  • Anti-interference underwater guidance system and method based on area array detector

    CN121069310A