Unmanned vehicle underground positioning method, system, equipment and medium

Abstract

The present invention discloses an underground positioning method, system, equipment and medium for an unmanned vehicle. The method comprises: constructing an environmental map of an underground tunnel; acquiring first environmental point cloud data, and performing target recognition and posture transformation to obtain first positioning information of a current vehicle; using the environmental map of the underground tunnel as an original point cloud, the second environmental point cloud data as a target point cloud, and the first positioning information of the current vehicle as an NDT prediction value, performing NDT solution, and outputting second positioning information of the current vehicle; acquiring a quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and outputting the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain an underground positioning result of the unmanned vehicle; the present invention utilizes the first and second positioning information to realize dual positioning of the vehicle, with high positioning accuracy and low safety.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned driving technology, and in particular relates to an underground positioning method, system, equipment and medium for an unmanned vehicle. Background Art

[0002] At present, unmanned vehicles in complex scenes mostly use real-time dynamic differential (RTK) combined navigation, ultra-wideband technology (UWB) and laser radar in conjunction with high-precision maps for positioning; however, in the complex environment underground, there is RTK signal obstruction, and its positioning result only relies on the integral calculation of the wheel speed meter and IMU, which can easily cause large cumulative drift; UWB technology has low positioning accuracy, and the UWB system occupies a high bandwidth, which can easily interfere with other wireless communication systems underground; at the same time, matching positioning based on high-precision maps requires a given initial position and posture for initialization, but in reality the unmanned vehicle may be at any position on the high-precision map, and the initial value of the point cloud and map registration cannot be provided at the beginning of positioning, making positioning difficult and the positioning result less accurate. Summary of the invention

[0003] In view of the technical problems existing in the prior art, the present invention provides a method, system, equipment and medium for positioning an unmanned vehicle underground to solve the technical problems that the existing unmanned vehicles underground have large positioning errors, low safety and are very likely to interfere with other wireless communication systems underground.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] The present invention provides an underground positioning method for an unmanned vehicle, the method comprising:

[0006] Construct environmental maps of underground tunnels;

[0007] Acquire first environmental point cloud data; wherein the first environmental point cloud data is an environmental point cloud collected by a roadside laser radar at the current moment; the roadside laser radars are arranged at intervals in an underground tunnel;

[0008] According to the first environmental point cloud data, a preset target detection algorithm is used to perform target recognition to obtain the position and posture information of the current vehicle relative to the roadside laser radar;

[0009] According to the position and posture information of the current vehicle relative to the roadside laser radar and the environment map of the roadside laser radar in the underground tunnel, the current vehicle is transformed in position and posture to obtain the first positioning information of the current vehicle;

[0010] The environment map of the underground tunnel is used as the original point cloud, the second environment point cloud data is used as the target point cloud, and the first positioning information of the current vehicle is used as the NDT prediction value to perform NDT solution, and output the second positioning information of the current vehicle; wherein the second environment point cloud data is the environment point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle;

[0011] Obtain a quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and output the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle.

[0012] Furthermore, the process of constructing the environmental map of the underground tunnel is specifically as follows: constructing the environmental map of the underground tunnel using the real-time positioning and three-dimensional mapping modules.

[0013] Furthermore, the environmental map of the underground tunnel is three-dimensional point cloud information of the environment of the underground tunnel, including position information and posture information of the coordinate origin of the environmental map of the underground tunnel in the world coordinate system; wherein the posture information includes roll angle, pitch angle and roll angle.

[0014] Furthermore, the spacing distance between adjacent roadside laser radars is less than or equal to 200m.

[0015] Furthermore, the preset target detection algorithm is a Pointpillars target detection algorithm.

[0016] Furthermore, the method further includes taking the environment map of the underground tunnel as the original point cloud, the second environment point cloud data as the target point cloud, and the first positioning information of the current vehicle as the NDT prediction value, performing NDT solution, and outputting the second positioning information of the current vehicle before performing NDT initialization;

[0017] The NDT initialization steps are as follows:

[0018] The environmental map of the underground tunnel is input as the original point cloud, and the first environmental point cloud data is input as the target point cloud, and NDT solution is performed. If the NDT solution converges, the first environmental point cloud data at the current moment is successfully aligned with the environmental map of the underground tunnel, that is, the NDT initialization is completed; otherwise, the position and posture information of the current vehicle relative to the roadside laser radar is re-input, and the NDT solution is performed until convergence is successful, and the NDT initialization is completed.

[0019] Further, a process of obtaining a quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, outputting the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference, and obtaining the underground positioning result of the unmanned vehicle is specifically as follows:

[0020] Compare the quantized difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle. If the quantized difference is greater than a set threshold, output the first positioning information of the current vehicle as the underground positioning result of the unmanned vehicle; otherwise, output the second positioning information of the current vehicle as the underground positioning result of the unmanned vehicle.

[0021] The present invention also provides an unmanned vehicle underground positioning system, comprising:

[0022] Map module, used to build environmental maps of underground tunnels;

[0023] A point cloud data module is used to obtain first environmental point cloud data; wherein the first environmental point cloud data is an environmental point cloud collected by a roadside laser radar at the current moment; the roadside laser radars are arranged at intervals in an underground tunnel;

[0024] A target recognition module is used to perform target recognition based on the first environmental point cloud data using a preset target detection algorithm to obtain the position and posture information of the current vehicle relative to the roadside laser radar;

[0025] A first positioning information module is used to transform the posture of the current vehicle according to the position and posture information of the current vehicle relative to the roadside laser radar and the environment map of the roadside laser radar in the underground tunnel to obtain the first positioning information of the current vehicle;

[0026] A second positioning information module is used to use the environment map of the underground tunnel as an original point cloud, the second environment point cloud data as a target point cloud, and the first positioning information of the current vehicle as an NDT prediction value to perform NDT solution, and output the second positioning information of the current vehicle; wherein the second environment point cloud data is an environment point cloud collected by a vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle;

[0027] The result output module is used to obtain the quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and output the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle.

[0028] The present invention also provides an unmanned vehicle underground positioning device, comprising:

[0029] Memory for storing computer programs;

[0030] A processor is used to implement the steps of an unmanned vehicle underground positioning method as described in any one of claims 1 to 7 when executing the computer program.

[0031] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the unmanned vehicle underground positioning method are implemented.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] The present invention provides an underground positioning method and system for an unmanned vehicle, which utilizes the first environmental point cloud data collected by a roadside laser radar to identify and locate the current vehicle, realize real-time monitoring of the vehicle's position information, improve the positioning accuracy and provide real-time data for ground monitoring; combine the first positioning information with the second positioning information collected by the vehicle-side unit, obtain the second positioning information through NDT alignment, and realize dual positioning of the underground vehicle, with high positioning accuracy and low safety; at the same time, it can avoid interference with other wireless communication systems underground.

[0034] Furthermore, combined with the ranging range of the roadside laser radar, the interval distance between adjacent roadside laser radars is set to be less than or equal to 200m, ensuring that the roadside laser radar can accurately obtain the corresponding point cloud feature information, thereby improving the accuracy of vehicle positioning.

[0035] Furthermore, the Pointpillars target detection algorithm is used to identify the vehicle target. The algorithm process is simple and the result has high accuracy.

[0036] Furthermore, through the NDT initialization operation, it is ensured that the matching is iterated on the basis of the initial position and attitude, so that a more accurate position and attitude can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The figure is a flow chart of the underground positioning method of the unmanned vehicle described in the embodiment. DETAILED DESCRIPTION

[0038] In order to make the technical problems, technical solutions and beneficial effects solved by the present invention more clearly understood, the present invention is further described in detail in the following specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0039] As attached Figure 1 As shown, the present invention provides an underground positioning method for an unmanned vehicle, comprising the following steps:

[0040] Step 1: construct an environmental map by loading an instant positioning and three-dimensional mapping module to obtain an environmental map of the underground tunnel; wherein the environmental map of the underground tunnel is the three-dimensional point cloud information of the environment of the underground tunnel, including the position information and attitude information of the coordinate origin of the environmental map of the underground tunnel in the world coordinate system; wherein the attitude information includes a roll angle, a pitch angle and a roll angle.

[0041] Step 2: Use a roadside laser radar to obtain first environmental point cloud data; wherein the roadside laser radars are arranged at intervals in an underground tunnel, and the interval distance between adjacent roadside laser radars is less than or equal to 200m; the first environmental point cloud data is the environmental point cloud collected by the first roadside laser radar at the current moment.

[0042] Step 3: Based on the first environmental point cloud data, a preset target detection algorithm is used to perform target recognition to obtain the position and posture information of the current vehicle relative to the roadside laser radar; wherein the preset target detection algorithm adopts the Pointpillars target detection algorithm.

[0043] Step 4: Input the environmental map of the underground tunnel as the original point cloud, and the first environmental point cloud data as the target point cloud, and perform NDT solution. If the NDT solution converges, the first environmental point cloud data at the current moment is successfully aligned with the environmental map of the underground tunnel, that is, the NDT initialization is completed; otherwise, re-input the current vehicle's position and posture information relative to the roadside laser radar, and perform NDT solution until convergence is successful, and the NDT initialization is completed.

[0044] Step 5: Use the environmental map of the underground tunnel as the original point cloud, the second environmental point cloud data as the target point cloud, and the first positioning information of the current vehicle as the NDT prediction value to perform NDT solution and output the second positioning information of the current vehicle; wherein the second environmental point cloud data is the environmental point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle.

[0045] Step 7: obtaining a quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and outputting the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle; the specific process includes the following steps:

[0046] Compare the quantized difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle. If the quantized difference is greater than a set threshold, output the first positioning information of the current vehicle as the underground positioning result of the unmanned vehicle; otherwise, output the second positioning information of the current vehicle as the underground positioning result of the unmanned vehicle.

[0047] In the present invention, the NDT: Normal Distribution Transformation, a point cloud registration algorithm that uses the spatial distribution of laser point clouds to solve the spatial transformation of two point clouds; NDT solution requires three inputs: the original point cloud, the target point cloud, and the predicted value of the point cloud transformation. This predicted value is mainly used as the initial value of iterative optimization, and a more accurate point cloud spatial transformation is output after the optimization converges.

[0048] The present invention also provides an underground positioning system for an unmanned vehicle, including a map module, a point cloud data module, a target recognition module, a first positioning information module, a second positioning information module and a result output module; the map module is used to construct an environmental map of the underground tunnel; the point cloud data module is used to obtain first environmental point cloud data; wherein the first environmental point cloud data is an environmental point cloud collected by a roadside laser radar at the current moment; the roadside laser radars are arranged at intervals in the underground tunnel; the target recognition module is used to perform target recognition based on the first environmental point cloud data and a preset target detection algorithm to obtain the position and posture information of the current vehicle relative to the roadside laser radar; the first positioning information module is used to obtain the position and posture information of the current vehicle relative to the roadside laser radar, and the roadside laser radar in the underground tunnel The environment map is used to transform the posture of the current vehicle to obtain the first positioning information of the current vehicle; the second positioning information module is used to use the environment map of the underground tunnel as the original point cloud, the second environment point cloud data as the target point cloud, and the first positioning information of the current vehicle as the NDT prediction value to perform NDT solution, and output the second positioning information of the current vehicle; wherein the second environment point cloud data is the environment point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle; the result output module is used to obtain the quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and output the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle.

[0049] The present invention also provides an unmanned vehicle underground positioning device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the unmanned vehicle underground positioning method when executing the computer program.

[0050] When the processor executes the computer program, the steps of the above-mentioned unmanned vehicle underground positioning method are implemented, for example: constructing an environmental map of the underground tunnel; obtaining first environmental point cloud data; wherein the first environmental point cloud data is an environmental point cloud collected by a roadside laser radar at the current moment; the roadside laser radars are arranged at intervals in the underground tunnel; according to the first environmental point cloud data, a preset target detection algorithm is used to perform target recognition to obtain the position and posture information of the current vehicle relative to the roadside laser radar; according to the position and posture information of the current vehicle relative to the roadside laser radar, and the environmental map of the roadside laser radar in the underground tunnel, the current vehicle is transformed in position and posture to obtain the current vehicle The first positioning information of the vehicle; using the environmental map of the underground tunnel as the original point cloud, the second environmental point cloud data as the target point cloud, and the first positioning information of the current vehicle as the NDT prediction value, NDT solution is performed, and the second positioning information of the current vehicle is output; wherein the second environmental point cloud data is the environmental point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle; obtaining the quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and outputting the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference, to obtain the underground positioning result of the unmanned vehicle.

[0051] Alternatively, when the processor executes the computer program, the functions of each module in the above system are implemented, for example: a map module is used to construct an environmental map of the underground tunnel; a point cloud data module is used to obtain first environmental point cloud data; wherein the first environmental point cloud data is an environmental point cloud collected by a roadside laser radar at the current moment; the roadside laser radars are arranged at intervals in the underground tunnel; a target recognition module is used to perform target recognition based on the first environmental point cloud data using a preset target detection algorithm to obtain the position and posture information of the current vehicle relative to the roadside laser radar; a first positioning information module is used to perform position and posture detection of the current vehicle based on the position and posture information of the current vehicle relative to the roadside laser radar, and the environmental map of the roadside laser radar in the underground tunnel. The first positioning information of the current vehicle is obtained by transformation; the second positioning information module is used to use the environmental map of the underground tunnel as the original point cloud, the second environmental point cloud data as the target point cloud, and the first positioning information of the current vehicle as the NDT prediction value to perform NDT solution, and output the second positioning information of the current vehicle; wherein the second environmental point cloud data is the environmental point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle; the result output module is used to obtain the quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and output the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle.

[0052] Exemplarily, the computer program can be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments that can complete preset functions, and the instruction segments are used to describe the execution process of the computer program in the unmanned vehicle underground positioning device. For example, the computer program can be divided into a map module, a point cloud data module, a target recognition module, a first positioning information module, a second positioning information module and a result output module; the specific functions of each module are as follows: a map module, which is used to construct an environmental map of the underground tunnel; a point cloud data module, which is used to obtain the first environmental point cloud data; wherein the first environmental point cloud data is the environmental point cloud collected by the roadside laser radar at the current moment; the roadside laser radar is arranged at intervals in the underground tunnel; a target recognition module, which is used to perform target recognition based on the first environmental point cloud data and a preset target detection algorithm to obtain the current vehicle's position and posture information relative to the roadside laser radar; the first positioning information module is used to obtain the position and posture information of the current vehicle relative to the roadside laser radar, and the roadside laser radar in the underground tunnel. The environmental map of the underground tunnel is used to transform the posture of the current vehicle to obtain the first positioning information of the current vehicle; the second positioning information module is used to use the environmental map of the underground tunnel as the original point cloud, the second environmental point cloud data as the target point cloud, and the first positioning information of the current vehicle as the NDT prediction value to perform NDT solution, and output the second positioning information of the current vehicle; wherein the second environmental point cloud data is the environmental point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle; the result output module is used to obtain the quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and output the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle.

[0053] The unmanned vehicle underground positioning device can be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The unmanned vehicle underground positioning device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the above are merely examples of unmanned vehicle underground positioning devices and do not constitute a limitation on unmanned vehicle underground positioning devices. More components than the above may be included, or some components may be combined, or different components may be included. For example, the unmanned vehicle underground positioning device may also include input and output devices, network access devices, buses, etc.

[0054] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The processor is the control center of the unmanned vehicle underground positioning device, and uses various interfaces and lines to connect the various parts of the entire unmanned vehicle underground positioning device.

[0055] The memory can be used to store the computer program and / or module, and the processor implements various functions of the unmanned vehicle underground positioning device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory.

[0056] The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDigital, SD) card, a flash card (FlashCard), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0057] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the unmanned vehicle underground positioning method are implemented.

[0058] If the module / unit integrated in the unmanned vehicle underground positioning system is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0059] Based on such understanding, the present invention implements all or part of the process in the above-mentioned unmanned vehicle underground positioning method, and can also be completed by instructing related hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned unmanned vehicle underground positioning method can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or preset intermediate form, etc.

[0060] The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0061] It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0062] Example

[0063] As attached Figure 1 As shown, this embodiment provides an unmanned vehicle underground positioning method, comprising the following steps:

[0064] Step 1: Load the environment map constructed by the real-time positioning and mapping module.

[0065] The environmental map is used for NDT matching and positioning with the real-time laser point cloud; the environmental map is the environmental three-dimensional point cloud information of the underground tunnel, including the position information and posture information of the origin of the environmental map coordinate system in the world coordinate system; wherein the posture information includes roll angle, pitch angle and roll angle; and checks the working status of the sensor interface of the current vehicle and obtains the data of each sensor and the vehicle posture information, etc.

[0066] Step 2: In the underground tunnel, roadside laser radars are installed at preset positions within a range of 200m; wherein the ranging range of the roadside laser radar is 200m; the roadside laser radar is used to obtain the first environmental point cloud data; and the second environmental point cloud data obtained by the vehicle-mounted laser radar, dual detection of the vehicle's position for more accurate positioning.

[0067] Step 3: Based on the first environmental point cloud data, use the pointpillars target detection algorithm to perform target recognition to obtain the position and posture information of the current vehicle relative to the roadside laser radar; wherein the pointpillars target detection algorithm uses the first environmental point cloud data as input, and through target recognition, the position and posture information of the current vehicle relative to the roadside laser radar can be obtained.

[0068] Step 4: According to the position and posture information of the current vehicle relative to the roadside laser radar and the environmental map of the roadside laser radar in the underground tunnel, the posture of the current vehicle is transformed to obtain the first positioning information of the current vehicle.

[0069] Step 5: Input the environment map of the underground tunnel as the original point cloud, and the first environment point cloud data as the target point cloud, and perform NDT solution. If the NDT solution converges, the first environment point cloud data at the current moment is successfully aligned with the environment map of the underground tunnel, that is, the NDT initialization is completed; otherwise, re-input the current vehicle's position and posture information relative to the roadside laser radar, and perform NDT solution until convergence is successful, and the NDT initialization is completed.

[0070] Step 6: After the NDT matching positioning is initialized successfully, the environmental map of the underground tunnel is used as the original point cloud, the second environmental point cloud data is used as the target point cloud, and the first positioning information of the current vehicle is used as the NDT prediction value to perform NDT solution and output the second positioning information of the current vehicle; wherein, the second environmental point cloud data is the environmental point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle.

[0071] Step 7: Compare the quantitative differences between the second positioning information and the first positioning information. Since the positioning is more accurate in a shorter time, if the quantitative difference is greater than the set threshold, it can be considered that the registration is degraded, and the first positioning information is output as the final positioning information; if the quantitative difference is less than the set threshold, it can be considered that the laser point cloud registration is accurate. Point cloud registration has high accuracy because each point is aligned with the point in the map, which corrects the first positioning information as the initial value. Therefore, the second positioning information is output as the final positioning information.

[0072] The description of the relevant parts of the unmanned vehicle underground positioning system, equipment and computer-readable storage medium provided in this embodiment can be found in the detailed description of the corresponding parts of an unmanned vehicle underground positioning method described in this embodiment, and will not be repeated here.

[0073] The unmanned vehicle underground positioning method and system described in the present invention utilizes the first environmental point cloud data collected by the road test laser radar to identify and locate the vehicle underground, monitors the vehicle's position information at all times, improves the positioning accuracy, provides real-time data for ground monitoring, and provides safety protection for personnel traveling underground; compared with the UWB base station, it has the advantages of low cost and small workload; the final positioning result is output by utilizing the quantitative difference between the first positioning information and the second positioning information, thereby realizing the dual positioning of the underground vehicle. If the NDT positioning matching is stuck or disconnected, the roadside laser radar can also obtain the vehicle's position to prevent accidents.

[0074] In the present invention, the first positioning information is sent to the NDT to allow it to obtain an initial position and attitude, completing the initialization of the NDT positioning model, and subsequent matching is iterated based on the initial position and attitude to obtain a more accurate position and attitude.

[0075] The above embodiment is only one of the implementation methods that can realize the technical solution of the present invention. The scope of protection claimed by the present invention is not limited only to this embodiment, but also includes changes, replacements and other implementation methods that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed by the present invention.

Claims

1. A method for positioning an unmanned vehicle underground, characterized in that: Methods include: Constructing an environment map of the underground tunnel; the environment map of the underground tunnel is three-dimensional point cloud information of the environment of the underground tunnel, including position information and attitude information of the coordinate origin of the environment map of the underground tunnel in the world coordinate system; wherein the attitude information includes a roll angle, a pitch angle and a roll angle; Acquire first environmental point cloud data; wherein the first environmental point cloud data is an environmental point cloud collected by a roadside laser radar at the current moment; the roadside laser radars are arranged at intervals in an underground tunnel; According to the first environmental point cloud data, a preset target detection algorithm is used to perform target recognition to obtain the position and posture information of the current vehicle relative to the roadside laser radar; the preset target detection algorithm is a Pointpillars target detection algorithm; According to the position and posture information of the current vehicle relative to the roadside laser radar and the environment map of the roadside laser radar in the underground tunnel, the current vehicle is transformed in position and posture to obtain the first positioning information of the current vehicle; The environment map of the underground tunnel is used as the original point cloud, the second environment point cloud data is used as the target point cloud, and the first positioning information of the current vehicle is used as the NDT prediction value to perform NDT solution, and output the second positioning information of the current vehicle; wherein the second environment point cloud data is the environment point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle; Obtaining a quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and outputting the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain an underground positioning result of the unmanned vehicle; The environment map of the underground tunnel is used as the original point cloud, the second environment point cloud data is used as the target point cloud, and the first positioning information of the current vehicle is used as the NDT prediction value to perform NDT solution. Before outputting the second positioning information of the current vehicle, an NDT initialization step is also included; The NDT initialization steps are as follows: The environment map of the underground tunnel is input as the original point cloud, and the first environment point cloud data is input as the target point cloud, and NDT solution is performed. If the NDT solution converges, the first environment point cloud data at the current moment is successfully registered with the environment map of the underground tunnel, that is, the NDT initialization is completed; otherwise, the position and posture information of the current vehicle relative to the roadside laser radar is re-input, and the NDT solution is performed until convergence is successful, and the NDT initialization is completed; The process of obtaining the quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and outputting the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle is as follows: Compare the quantized difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle. If the quantized difference is greater than a set threshold, output the first positioning information of the current vehicle as the underground positioning result of the unmanned vehicle; otherwise, output the second positioning information of the current vehicle as the underground positioning result of the unmanned vehicle.

2. The method for underground positioning of an unmanned vehicle according to claim 1, characterized in that: The process of constructing the environmental map of the underground tunnel is specifically: using the real-time positioning and three-dimensional mapping module to construct the environmental map of the underground tunnel.

3. The method for underground positioning of an unmanned vehicle according to claim 1, characterized in that: The distance between adjacent roadside laser radars is less than or equal to 200m.

4. An unmanned vehicle underground positioning system, characterized in that: include: A map module is used to construct an environment map of the underground tunnel; the environment map of the underground tunnel is three-dimensional point cloud information of the environment of the underground tunnel, including position information and attitude information of the coordinate origin of the environment map of the underground tunnel in the world coordinate system; wherein the attitude information includes a roll angle, a pitch angle and a roll angle; A point cloud data module is used to obtain first environmental point cloud data; wherein the first environmental point cloud data is an environmental point cloud collected by a roadside laser radar at the current moment; the roadside laser radars are arranged at intervals in an underground tunnel; A target recognition module is used to perform target recognition based on the first environmental point cloud data using a preset target detection algorithm to obtain the position and posture information of the current vehicle relative to the roadside laser radar; the preset target detection algorithm is a Pointpillars target detection algorithm; A first positioning information module is used to transform the posture of the current vehicle according to the position and posture information of the current vehicle relative to the roadside laser radar and the environment map of the roadside laser radar in the underground tunnel to obtain the first positioning information of the current vehicle; A second positioning information module is used to use the environment map of the underground tunnel as the original point cloud, the second environment point cloud data as the target point cloud, and the first positioning information of the current vehicle as the NDT prediction value to perform NDT solution and output the second positioning information of the current vehicle; wherein the second environment point cloud data is the environment point cloud collected by the vehicle-side laser radar at the current moment, and the vehicle-side laser radar is installed on the current vehicle; A result output module is used to obtain a quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and output the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle; The environment map of the underground tunnel is used as the original point cloud, the second environment point cloud data is used as the target point cloud, and the first positioning information of the current vehicle is used as the NDT prediction value to perform NDT solution. Before outputting the second positioning information of the current vehicle, an NDT initialization step is also included; The NDT initialization steps are as follows: The environment map of the underground tunnel is input as the original point cloud, and the first environment point cloud data is input as the target point cloud, and NDT solution is performed. If the NDT solution converges, the first environment point cloud data at the current moment is successfully registered with the environment map of the underground tunnel, that is, the NDT initialization is completed; otherwise, the position and posture information of the current vehicle relative to the roadside laser radar is re-input, and the NDT solution is performed until convergence is successful, and the NDT initialization is completed; The process of obtaining the quantitative difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle, and outputting the first positioning information of the current vehicle or the second positioning information of the current vehicle according to the quantitative difference to obtain the underground positioning result of the unmanned vehicle is as follows: Compare the quantized difference between the first positioning information of the current vehicle and the second positioning information of the current vehicle. If the quantized difference is greater than a set threshold, output the first positioning information of the current vehicle as the underground positioning result of the unmanned vehicle; otherwise, output the second positioning information of the current vehicle as the underground positioning result of the unmanned vehicle.

5. An unmanned vehicle underground positioning device, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of an unmanned vehicle underground positioning method as described in any one of claims 1 to 3 when executing the computer program.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, the steps of the unmanned vehicle underground positioning method as described in any one of claims 1-3 are implemented.

Images