A multi-sensor adaptive calibration method, system and medium for unmanned vehicles

Abstract

The present invention relates to a multi-sensor adaptive calibration method, system, and medium for unmanned vehicles. The method comprises: W1. When the vehicle is traveling on a road, image data information of the road is acquired in real time based on an onboard camera, point cloud data information of the road is acquired in real time based on an onboard laser radar, and a topological model of the environment surrounding the vehicle is constructed. The environmental data surrounding the vehicle is topologically analyzed to obtain topological data information of the vehicle's surrounding environment; W2. Based on the topological data information of the vehicle's surrounding environment, a local feature matching algorithm based on WLD is used to extract features from the data of the vehicle's surrounding environment, and a feature matrix of the vehicle's surrounding environment is constructed. The present invention not only eliminates the dependence on calibration targets, reduces calibration cost and complexity, and improves calibration flexibility and adaptability, but also the calibration process can be performed in real time while the vehicle is traveling without interrupting travel, thereby improving calibration efficiency and the real-time perception capability of the autonomous vehicle.

Description

Technical Field

[0001] The present invention relates to the field of unmanned driving technology, and in particular to a multi-sensor adaptive calibration method, system and medium for an unmanned vehicle. Background Art

[0002] Autonomous driving technology relies heavily on the coordinated operation of multiple sensors, including cameras, millimeter-wave radar, and lidar. Traditional sensor calibration methods often rely on calibration targets and fixed calibration sites. This not only increases the complexity and cost of the calibration process but also limits the choice of calibration scenarios. Furthermore, the accuracy of the calibration targets is directly affected by their precision and stability. Summary of the Invention

[0003] In view of the above problems, the present invention provides a multi-sensor adaptive calibration method, system and medium for unmanned vehicles, which not only gets rid of the dependence on calibration targets, reduces calibration costs and complexity, and improves calibration flexibility and adaptability, but also the calibration process can be carried out in real time during vehicle driving without interrupting driving, thereby improving calibration efficiency and the real-time perception capability of autonomous driving vehicles.

[0004] In order to achieve the above-mentioned and other related purposes, the present invention provides the following technical solutions:

[0005] A multi-sensor adaptive calibration method for an unmanned vehicle, the method comprising:

[0006] W1. The vehicle, while driving on the road, acquires real-time road image data using the onboard camera and point cloud data using the onboard lidar. A topological model of the vehicle's surrounding environment is constructed and topological data is performed on the surrounding environment to obtain topological data of the vehicle's surrounding environment.

[0007] W2. Based on the topological data information of the vehicle surroundings, the data of the vehicle surroundings is extracted using a local feature matching algorithm based on WLD, and a feature matrix of the vehicle surroundings is constructed to obtain feature matrix data information of the vehicle surroundings;

[0008] W3. Based on the characteristic matrix data information of the vehicle surrounding environment, an adaptive calibration algorithm is used to calculate the relative posture relationship between the sensors to obtain the relative posture relationship data information between the sensors;

[0009] W4. Based on the relative posture relationship data information between the sensors, the internal and external parameters of the sensors are optimized and calibrated using a BA algorithm based on an adaptive adjustment factor to obtain the internal and external parameter data information of the calibrated sensors.

[0010] Furthermore, in step W1, the construction of the environmental topology model around the vehicle and the characterization of the environmental data around the vehicle include:

[0011] W11. Based on the image data information of the road and the point cloud data information of the road, construct a fusion function Q of the road point cloud and the image,

[0012]

[0013] Where x is the image data information of the road, y is the point cloud data information of the road, α1, α2 and α3 are the feature fusion factors of the image and point cloud. The image and point cloud of the road are fused to obtain the fused data information of the point cloud and image of the road;

[0014] W12. Based on the fusion data information of the road point cloud and image, establish the environment topology function R around the vehicle,

[0015]

[0016] Among them, z is the fusion data information of the road point cloud and image, β1, β2 and β3 are the topological factors of the vehicle's surrounding environment;

[0017] W13. Based on the environmental topology function R around the vehicle, perform topology on the environmental data around the vehicle to obtain topology data information of the environment around the vehicle.

[0018] Furthermore, the topological factors β1, β2 and β3 of the vehicle surrounding environment are,

[0019]

[0020]

[0021]

[0022] Among them, z is the fusion data information of the road point cloud and image.

[0023] Furthermore, in step W2, the feature extraction of the vehicle surrounding environment data using the WLD-based local feature matching algorithm includes:

[0024] W21. Based on the topological data information of the vehicle surroundings, establishing a WLD local partitioning function U of the vehicle surroundings,

[0025]

[0026] Where a is the topological data information of the vehicle's surrounding environment, δ and γ are the local partitioning weight coefficients of WLD, and the vehicle's surrounding environment is divided to obtain the divided vehicle surrounding environment data information;

[0027] W22. Based on the divided vehicle surrounding environment data information, establish a differential excitation operator function P of the vehicle surrounding environment,

[0028]

[0029] Among them, x i is the data information of the vehicle surrounding environment after division, x0 is the mean data information of the vehicle surrounding environment after division, η i is the differential constant parameter of the vehicle surrounding environment after division, n is the sample size, and the differential excitation operator of the vehicle surrounding environment is characterized to obtain the differential excitation operator data information of the vehicle surrounding environment;

[0030] W23. Based on the differential excitation operator data information of the vehicle surrounding environment, establish a local feature matching function S of the vehicle surrounding environment,

[0031]

[0032] Among them, b is the differential excitation operator data information of the vehicle's surrounding environment, λ1, λ2 and λ3 are the feature matching factors of the vehicle's surrounding environment, feature extraction is performed on the data of the vehicle's surrounding environment, and a feature matrix of the vehicle's surrounding environment is constructed to obtain the feature matrix data information of the vehicle's surrounding environment.

[0033] Furthermore, the constraints of the vehicle surrounding environment feature matching factors λ1, λ2 and λ3 are as follows:

[0034]

[0035] Furthermore, the mean data information x0 of the vehicle surrounding environment after the division is,

[0036]

[0037]

[0038] Among them, x i is the divided vehicle surrounding environment data information, n is the sample capacity, μ i is the gain constant parameter of the divided vehicle environment.

[0039] Furthermore, in step W3, the use of the adaptive calibration algorithm to calculate the relative posture relationship between the sensors includes:

[0040] W31. Based on the characteristic matrix data information of the vehicle surrounding environment, establish a geometric relationship function F between the sensor and the vehicle surrounding environment,

[0041]

[0042] Where c is the characteristic matrix data information of the vehicle's surrounding environment, θ1 and θ2 are the geometric adjustment factors of the sensor and the vehicle's surrounding environment. The geometric relationship between the sensor and the vehicle's surrounding environment is calculated to obtain the geometric relationship data information between the sensor and the vehicle's surrounding environment;

[0043] W32. Based on the geometric relationship data between the sensor and the vehicle's surrounding environment, establish a relative pose function G between the sensors,

[0044]

[0045] Among them, r is the geometric relationship data information between the sensor and the vehicle's surrounding environment, ρ1, ρ2 and ρ3 are the relative posture error factors between sensors;

[0046] W33. Based on the relative posture function G between the sensors, the relative posture relationship between the sensors is calculated to obtain data information on the relative posture relationship between the sensors.

[0047] Furthermore, in step W4, the optimization and calibration of the internal and external parameters of the sensor using the BA algorithm based on the adaptive adjustment factor includes:

[0048] W41. Based on the relative position relationship data information between the sensors, construct the internal and external parameter matrix of the sensor, and initialize the bat population to obtain the initialized bat population data information;

[0049] W42. Based on the initialized bat population data information, establish the fitness function L of the population individuals,

[0050]

[0051] Among them, g is the initialized bat population data information, σ1 and σ2 are the fitness determining factors of the bat population individuals, and the fitness values ​​of the population individuals are calculated to obtain the fitness value data information of the population individuals;

[0052] W43. Based on the fitness value data information of the population individuals, establish the objective function V,

[0053] Among them, h is the fitness value data information of the population individual, ω1, ω2 and ω3 are the target optimization adaptive adjustment factors, and the internal and external parameters of the sensor are optimized and calibrated to obtain the internal and external parameter data information of the calibrated sensor.

[0054] In order to achieve the above-mentioned and other related objectives, the present invention further provides a system for implementing any one of the multi-sensor adaptive calibration methods for an unmanned vehicle, the system comprising:

[0055] A sensor suite, including cameras and lidar, to sense the vehicle's surroundings;

[0056] Data processing module, used to receive the raw data from each sensor and perform preprocessing, feature extraction and data analysis;

[0057] Adaptive calibration algorithm module, used to use adaptive calibration algorithm to characterize the geometric relationship between natural features in the environment and sensors, and automatically calculate the relative position and posture relationship between sensors;

[0058] The control module is used to adjust the operating parameters of the sensor according to the calibration results to ensure data synchronization and consistency between sensors.

[0059] In order to achieve the above-mentioned objectives and other related objectives, the present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the multi-sensor adaptive calibration methods for unmanned vehicles.

[0060] The present invention has the following positive effects:

[0061] 1. The present invention constructs a topological model of the environment around the vehicle, performs topology on the environmental data around the vehicle, and uses a local feature matching algorithm based on WLD to extract features from the data of the environment around the vehicle. This not only eliminates the dependence on calibration targets, reduces calibration costs and complexity, and improves calibration flexibility and adaptability, but also can adapt to sensors of different models and different precisions, as well as complex and changeable environmental conditions, thereby improving the robustness and reliability of the autonomous driving system.

[0062] 2. The present invention uses an adaptive calibration algorithm to calculate the relative posture relationship between sensors, and combines it with a BA algorithm based on an adaptive adjustment factor to optimize and calibrate the internal and external parameters of the sensors. Not only can the calibration process be performed in real time while the vehicle is driving without interrupting driving, thereby improving the calibration efficiency and the real-time perception capability of the autonomous driving vehicle, but the adaptive calibration algorithm and complex optimization technology also ensure the accuracy and stability of the calibration results. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 Schematic diagram of the method flow of the present invention;

[0064] Figure 2 A schematic diagram of the process of constructing an environmental topology model around a vehicle according to the present invention;

[0065] Figure 3 Schematic diagram of the flow of the WLD-based local feature matching algorithm of the present invention;

[0066] Figure 4 Schematic diagram of the process of the adaptive calibration algorithm of the present invention;

[0067] Figure 5 Schematic diagram of the process of the BA algorithm based on the adaptive adjustment factor of the present invention;

[0068] Figure 6 Schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0069] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0070] Example 1: Figure 1 As shown, a multi-sensor adaptive calibration method for an unmanned vehicle, the method comprising:

[0071] W1. The vehicle, while driving on the road, acquires real-time road image data using the onboard camera and point cloud data using the onboard lidar. A topological model of the vehicle's surrounding environment is constructed and topological data is performed on the surrounding environment to obtain topological data of the vehicle's surrounding environment.

[0072] W2. Based on the topological data information of the vehicle surroundings, the data of the vehicle surroundings is extracted using a local feature matching algorithm based on WLD, and a feature matrix of the vehicle surroundings is constructed to obtain feature matrix data information of the vehicle surroundings;

[0073] W3. Based on the characteristic matrix data information of the vehicle surrounding environment, an adaptive calibration algorithm is used to calculate the relative posture relationship between the sensors to obtain the relative posture relationship data information between the sensors;

[0074] W4. Based on the relative posture relationship data information between the sensors, the internal and external parameters of the sensors are optimized and calibrated using a BA algorithm based on an adaptive adjustment factor to obtain the internal and external parameter data information of the calibrated sensors.

[0075] In this embodiment, if Figure 2 As shown, in step W1, the construction of the environmental topology model around the vehicle and the characterization of the environmental data around the vehicle include:

[0076] W11. Based on the image data information of the road and the point cloud data information of the road, construct a fusion function Q of the road point cloud and the image,

[0077]

[0078] Where x is the image data information of the road, y is the point cloud data information of the road, α1, α2 and α3 are the feature fusion factors of the image and point cloud. The image and point cloud of the road are fused to obtain the fused data information of the point cloud and image of the road;

[0079] W12. Based on the fusion data information of the road point cloud and image, establish the environment topology function R around the vehicle,

[0080]

[0081] Among them, z is the fusion data information of the road point cloud and image, β1, β2 and β3 are the topological factors of the vehicle's surrounding environment;

[0082] W13. Based on the environmental topology function R around the vehicle, perform topology on the environmental data around the vehicle to obtain topology data information of the environment around the vehicle.

[0083] In this embodiment, the topological factors β1, β2 and β3 of the vehicle surrounding environment are:

[0084]

[0085]

[0086] Among them, z is the fusion data information of the road point cloud and image.

[0087] In this embodiment, if Figure 3 As shown, in step W2, the feature extraction of the vehicle surrounding environment data using the WLD-based local feature matching algorithm includes:

[0088] W21. Based on the topological data information of the vehicle surroundings, establishing a WLD local partitioning function U of the vehicle surroundings,

[0089]

[0090] Where a is the topological data information of the vehicle's surrounding environment, δ and γ are the local partitioning weight coefficients of WLD, and the vehicle's surrounding environment is divided to obtain the divided vehicle surrounding environment data information;

[0091] W22. Based on the divided vehicle surrounding environment data information, establish a differential excitation operator function P of the vehicle surrounding environment,

[0092]

[0093] Among them, x i is the data information of the vehicle surrounding environment after division, x0 is the mean data information of the vehicle surrounding environment after division, η i is the differential constant parameter of the vehicle surrounding environment after division, n is the sample size, and the differential excitation operator of the vehicle surrounding environment is characterized to obtain the differential excitation operator data information of the vehicle surrounding environment;

[0094] W23. Based on the differential excitation operator data information of the vehicle surrounding environment, establish a local feature matching function S of the vehicle surrounding environment,

[0095]

[0096] Among them, b is the differential excitation operator data information of the vehicle's surrounding environment, λ1, λ2 and λ3 are the feature matching factors of the vehicle's surrounding environment, feature extraction is performed on the data of the vehicle's surrounding environment, and a feature matrix of the vehicle's surrounding environment is constructed to obtain the feature matrix data information of the vehicle's surrounding environment.

[0097] In this embodiment, the constraints of the vehicle surrounding environment feature matching factors λ1, λ2 and λ3 are:

[0098]

[0099] In this embodiment, the mean data information x0 of the vehicle surrounding environment after division is,

[0100]

[0101] Among them, x i is the divided vehicle surrounding environment data information, n is the sample capacity, μ i is the gain constant parameter of the divided vehicle environment.

[0102] Example 2: Based on the multi-sensor adaptive calibration method for an unmanned vehicle in Example 1, the present invention is further illustrated and described below.

[0103] like Figure 1 As shown, a multi-sensor adaptive calibration method for an unmanned vehicle, the method comprising:

[0104] W1. The vehicle, while driving on the road, acquires real-time road image data using the onboard camera and point cloud data using the onboard lidar. A topological model of the vehicle's surrounding environment is constructed and topological data is performed on the surrounding environment to obtain topological data of the vehicle's surrounding environment.

[0105] W2. Based on the topological data information of the vehicle surroundings, the data of the vehicle surroundings is extracted using a local feature matching algorithm based on WLD, and a feature matrix of the vehicle surroundings is constructed to obtain feature matrix data information of the vehicle surroundings;

[0106] W3. Based on the characteristic matrix data information of the vehicle surrounding environment, an adaptive calibration algorithm is used to calculate the relative posture relationship between the sensors to obtain the relative posture relationship data information between the sensors;

[0107] W4. Based on the relative posture relationship data information between the sensors, the internal and external parameters of the sensors are optimized and calibrated using a BA algorithm based on an adaptive adjustment factor to obtain the internal and external parameter data information of the calibrated sensors.

[0108] In this embodiment, if Figure 4 As shown, in step W3, the use of the adaptive calibration algorithm to calculate the relative posture relationship between the sensors includes:

[0109] W31. Based on the characteristic matrix data information of the vehicle surrounding environment, establish a geometric relationship function F between the sensor and the vehicle surrounding environment,

[0110]

[0111] Where c is the characteristic matrix data information of the vehicle's surrounding environment, θ1 and θ2 are the geometric adjustment factors of the sensor and the vehicle's surrounding environment. The geometric relationship between the sensor and the vehicle's surrounding environment is calculated to obtain the geometric relationship data information between the sensor and the vehicle's surrounding environment;

[0112] W32. Based on the geometric relationship data between the sensor and the vehicle's surrounding environment, establish a relative pose function G between the sensors,

[0113]

[0114] Among them, r is the geometric relationship data information between the sensor and the vehicle's surrounding environment, ρ1, ρ2 and ρ3 are the relative posture error factors between sensors;

[0115] W33. Based on the relative posture function G between the sensors, the relative posture relationship between the sensors is calculated to obtain data information on the relative posture relationship between the sensors.

[0116] In this embodiment, if Figure 5 As shown, in step W4, the optimization and calibration of the internal and external parameters of the sensor using the BA algorithm based on the adaptive adjustment factor includes:

[0117] W41. Based on the relative position relationship data information between the sensors, construct the internal and external parameter matrix of the sensor, and initialize the bat population to obtain the initialized bat population data information;

[0118] W42. Based on the initialized bat population data information, establish the fitness function L of the population individuals,

[0119]

[0120] Among them, g is the initialized bat population data information, σ1 and σ2 are the fitness determining factors of the bat population individuals, and the fitness values ​​of the population individuals are calculated to obtain the fitness value data information of the population individuals;

[0121] W43. Based on the fitness value data information of the population individuals, establish the objective function V,

[0122] Among them, h is the fitness value data information of the population individual, ω1, ω2 and ω3 are the target optimization adaptive adjustment factors, and the internal and external parameters of the sensor are optimized and calibrated to obtain the internal and external parameter data information of the calibrated sensor.

[0123] In this embodiment, if Figure 6 As shown, the present invention provides a system for implementing any one of the multi-sensor adaptive calibration methods for an unmanned vehicle, the system comprising:

[0124] A sensor suite, including cameras and lidar, to sense the vehicle's surroundings;

[0125] Data processing module, used to receive the raw data from each sensor and perform preprocessing, feature extraction and data analysis;

[0126] Adaptive calibration algorithm module, used to use adaptive calibration algorithm to characterize the geometric relationship between natural features in the environment and sensors, and automatically calculate the relative position and posture relationship between sensors;

[0127] The control module is used to adjust the operating parameters of the sensor according to the calibration results to ensure data synchronization and consistency between sensors.

[0128] In this embodiment, the present invention provides a computer-readable storage medium storing a computer program programmed or configured to execute any one of the multi-sensor adaptive calibration methods for an unmanned vehicle.

[0129] Any reference to memory, storage, database or other media used in the embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0130] In summary, the present invention not only gets rid of the dependence on calibration targets, reduces calibration costs and complexity, and improves calibration flexibility and adaptability, but also the calibration process can be carried out in real time while the vehicle is driving without interrupting driving, thereby improving calibration efficiency and the real-time perception capability of autonomous driving vehicles.

[0131] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A multi-sensor adaptive calibration method for an unmanned vehicle, characterized in that: The method comprises: W1. The vehicle, while driving on the road, acquires real-time road image data using the onboard camera and point cloud data using the onboard lidar. A topological model of the vehicle's surrounding environment is constructed and topological data is performed on the surrounding environment to obtain topological data of the vehicle's surrounding environment. W2. Based on the topological data information of the vehicle surroundings, the data of the vehicle surroundings is extracted using a local feature matching algorithm based on WLD, and a feature matrix of the vehicle surroundings is constructed to obtain feature matrix data information of the vehicle surroundings; W3. Based on the characteristic matrix data information of the vehicle surrounding environment, an adaptive calibration algorithm is used to calculate the relative posture relationship between the sensors to obtain the relative posture relationship data information between the sensors; W4. Based on the relative posture relationship data information between the sensors, the internal and external parameters of the sensors are optimized and calibrated using a BA algorithm based on an adaptive adjustment factor to obtain the internal and external parameter data information of the calibrated sensors.

2. The multi-sensor adaptive calibration method for an unmanned vehicle according to claim 1, characterized in that: In step W1, the construction of the environmental topology model around the vehicle and the characterization of the environmental data around the vehicle include: W11. Based on the image data information of the road and the point cloud data information of the road, construct a fusion function Q of the road point cloud and the image, Where x is the image data information of the road, y is the point cloud data information of the road, α1, α2 and α3 are the feature fusion factors of the image and point cloud. The image and point cloud of the road are fused to obtain the fused data information of the point cloud and image of the road; W12. Based on the fusion data information of the road point cloud and image, establish the environment topology function R around the vehicle, Among them, z is the fusion data information of the road point cloud and image, β1, β2 and β3 are the topological factors of the vehicle's surrounding environment; W13. Based on the environmental topology function R around the vehicle, perform topology on the environmental data around the vehicle to obtain topology data information of the environment around the vehicle.

3. The multi-sensor adaptive calibration method for an unmanned vehicle according to claim 2, characterized in that: The topological factors β1, β2 and β3 of the vehicle surrounding environment are, Among them, z is the fusion data information of the road point cloud and image.

4. The multi-sensor adaptive calibration method for an unmanned vehicle according to claim 1, characterized in that: In step W2, the feature extraction of the vehicle surrounding environment data using the WLD-based local feature matching algorithm includes: W21. Based on the topological data information of the vehicle surroundings, establishing a WLD local partitioning function U of the vehicle surroundings, Where a is the topological data information of the vehicle's surrounding environment, δ and γ are the local partitioning weight coefficients of WLD, and the vehicle's surrounding environment is divided to obtain the divided vehicle surrounding environment data information; W22. Based on the divided vehicle surrounding environment data information, establish a differential excitation operator function P of the vehicle surrounding environment, Among them, x i is the data information of the vehicle surrounding environment after division, x0 is the mean data information of the vehicle surrounding environment after division, η i is the differential constant parameter of the vehicle surrounding environment after division, n is the sample size, and the differential excitation operator of the vehicle surrounding environment is characterized to obtain the differential excitation operator data information of the vehicle surrounding environment; W23. Based on the differential excitation operator data information of the vehicle surrounding environment, establish a local feature matching function S of the vehicle surrounding environment, Among them, b is the differential excitation operator data information of the vehicle's surrounding environment, λ1, λ2 and λ3 are the feature matching factors of the vehicle's surrounding environment, feature extraction is performed on the data of the vehicle's surrounding environment, and a feature matrix of the vehicle's surrounding environment is constructed to obtain the feature matrix data information of the vehicle's surrounding environment.

5. The multi-sensor adaptive calibration method for an unmanned vehicle according to claim 4, characterized in that: The constraints of the vehicle surrounding environment feature matching factors λ1, λ2 and λ3 are:

6. The multi-sensor adaptive calibration method for an unmanned vehicle according to claim 4, characterized in that: The mean data information x0 of the vehicle surrounding environment after the division is, Among them, x i is the divided vehicle surrounding environment data information, n is the sample capacity, μ i is the gain constant parameter of the divided vehicle environment.

7. The multi-sensor adaptive calibration method for an unmanned vehicle according to claim 1, characterized in that: In step W3, the use of the adaptive calibration algorithm to calculate the relative posture relationship between the sensors includes: W31. Based on the characteristic matrix data information of the vehicle surrounding environment, establish a geometric relationship function F between the sensor and the vehicle surrounding environment, Where c is the characteristic matrix data information of the vehicle's surrounding environment, θ1 and θ2 are the geometric adjustment factors of the sensor and the vehicle's surrounding environment. The geometric relationship between the sensor and the vehicle's surrounding environment is calculated to obtain the geometric relationship data information between the sensor and the vehicle's surrounding environment; W32. Based on the geometric relationship data between the sensor and the vehicle's surrounding environment, establish a relative pose function G between the sensors, Among them, r is the geometric relationship data information between the sensor and the vehicle's surrounding environment, ρ1, ρ2 and ρ3 are the relative posture error factors between sensors; W33. Based on the relative posture function G between the sensors, the relative posture relationship between the sensors is calculated to obtain data information on the relative posture relationship between the sensors.

8. The multi-sensor adaptive calibration method for an unmanned vehicle according to claim 1, characterized in that: In step W4, the optimization and calibration of the internal and external parameters of the sensor using the BA algorithm based on the adaptive adjustment factor includes: W41. Based on the relative position relationship data information between the sensors, construct the internal and external parameter matrix of the sensor, and initialize the bat population to obtain the initialized bat population data information; W42. Based on the initialized bat population data information, establish the fitness function L of the population individuals, Among them, g is the initialized bat population data information, σ1 and σ2 are the fitness determining factors of the bat population individuals, and the fitness values ​​of the population individuals are calculated to obtain the fitness value data information of the population individuals; W43. Based on the fitness value data information of the population individuals, establish the objective function V, Among them, h is the fitness value data information of the population individual, ω1, ω2 and ω3 are the adaptive adjustment factors of the target optimization, and the internal and external parameters of the sensor are optimized and calibrated to obtain the internal and external parameter data information of the calibrated sensor.

9. A system for implementing the multi-sensor adaptive calibration method for an unmanned vehicle according to any one of claims 1 to 8, characterized in that: The system comprises: A sensor suite, including cameras and lidar, to sense the vehicle's surroundings; Data processing module, used to receive the raw data from each sensor and perform preprocessing, feature extraction and data analysis; Adaptive calibration algorithm module, used to use adaptive calibration algorithm to characterize the geometric relationship between natural features in the environment and sensors, and automatically calculate the relative position and posture relationship between sensors; The control module is used to adjust the operating parameters of the sensor according to the calibration results to ensure data synchronization and consistency between sensors.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program that is programmed or configured to execute the multi-sensor adaptive calibration method for an unmanned vehicle according to any one of claims 1 to 8.

Images