Positioning method, device and equipment based on environment features and readable storage medium

Abstract

The invention discloses an environment feature-based positioning method, device and equipment and a readable storage medium, and the method comprises the steps: carrying out the matching of environment feature information collected by a sensor with a pre-stored prior map, and employing a Gaussian-Newton algorithm to optimize the matching process, so as to find a map subset which is most consistent with the current environment feature; and calculating a transformation matrix between the point pairs according to the historical pose truth value and the feature point coordinate truth value of the vehicle provided by the matched map subset, and solving the pose of the current vehicle. The Gaussian-Newton algorithm can be adopted to enable the matched map subsets to maintain the precision, reduce the calculation amount, improve the operation speed, continuously reduce the spatial position difference or distance between the feature points in the real-time sensor data and the corresponding feature points in the map subsets, and solve the current pose. And the precision and reliability of the positioning system are improved.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving, and specifically to a method, device, equipment and readable storage medium based on environmental feature positioning. Background Art

[0002] Positioning technology is the core of realizing applications such as autonomous driving and navigation, and can be mainly divided into the following categories: Signal-based positioning: such as GPS (Global Positioning System), the global positioning system determines the position by receiving satellite signals, which belongs to global positioning. The global positioning system consists of a group of satellites evenly distributed in the earth's orbit. The ground receiving equipment receives these satellite signals and performs time measurement and distance calculation to ultimately determine the precise geographic location of the receiver. Positioning based on track estimation: such as mileage estimation, which obtains acceleration and angular velocity information through inertial sensors to infer the current position and heading, which belongs to local relative positioning.

[0003] In the related technologies, for signal-based positioning, satellite signals have their own inherent disadvantages. For example, in tunnels, satellite signals will be blocked and cannot be used normally, or because of the multipath effect, the performance of satellite positioning will be affected to a certain extent; positioning based on track estimation can maintain accuracy for a short period of time, but if used for a long time, the track estimation will have the problem of accumulated deviations, which will cause positioning drift. Summary of the invention

[0004] The present application provides a method, apparatus, device and computer-readable storage medium for positioning based on environmental features, which can use the Gauss-Newton algorithm to ensure that the matched map subset can maintain accuracy while reducing the amount of calculation and improving the operation speed, continuously reducing the spatial position difference or distance between the feature points in the real-time sensor data and the corresponding feature points in the map subset, solving the current posture, and thus improving the accuracy and reliability of the positioning system.

[0005] In a first aspect, an embodiment of the present application provides a positioning method based on environmental features, and the positioning method based on environmental features includes:

[0006] Match the environmental feature information collected by the sensor with the pre-stored prior map, and use the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environmental features;

[0007] According to the historical true values ​​of the vehicle's posture and the true values ​​of the feature point coordinates provided by the matched map subset, the transformation matrix between the point pairs is calculated to solve the current vehicle's posture.

[0008] In combination with the first aspect, in one implementation, matching the environmental feature information collected by the sensor with the pre-stored priori map, and optimizing the matching process using the Gauss-Newton algorithm to find a map subset that best matches the current environmental features, includes:

[0009] Matching the environmental feature information collected by the sensor with the pre-stored priori map, and calculating the fitting value of the initial matching result;

[0010] Determine whether the fitting value is lower than the preset fitting value. If so, switch to the next sensor, match the environmental feature information collected by the next sensor with the pre-stored prior map, and calculate the fitting value of the re-matching result;

[0011] Repeat the step of determining the fitting value until a sensor whose fitting value is not less than the preset fitting value is found;

[0012] For sensors whose fitting values ​​are not lower than the preset fitting values, the Gauss-Newton algorithm is used to optimize the matching process to find the map subset that best matches the current environment characteristics.

[0013] In combination with the first aspect, in one embodiment, before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes:

[0014] The environmental feature information collected by the sensor is obtained, and the collected environmental feature information is preprocessed using a voxel filtering algorithm to reduce the density of the point cloud data and retain the point cloud data representing the environmental features.

[0015] In combination with the first aspect, in one embodiment, before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes:

[0016] Obtain the environmental feature information collected by the sensor, use the random filtering algorithm to pre-process the collected environmental feature information, and reduce the density of point cloud data.

[0017] In combination with the first aspect, in one embodiment, before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes:

[0018] The environmental feature information collected by the sensor is obtained, and the Euclidean distance point cloud filtering method is used to remove the data that is not within the set threshold range in the environmental feature information collected by the sensor.

[0019] In combination with the first aspect, in one embodiment, before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes:

[0020] Obtain environmental feature information collected by lidar or visual camera.

[0021] In a second aspect, an embodiment of the present application provides a positioning device based on environmental features, the positioning device based on environmental features comprising:

[0022] A map subset matching module is used to match the environmental feature information collected by the sensor with the pre-stored prior map, and use the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environmental features;

[0023] The current posture calculation module is used to calculate the transformation matrix between point pairs based on the historical posture true value of the vehicle and the true value of the feature point coordinates provided by the matched map subset, and solve the current vehicle posture.

[0024] In combination with the second aspect, in one embodiment, the current posture calculation module is also used to match the environmental feature information collected by the sensor with a pre-stored prior map, and calculate the fitting value of the initial matching result; determine whether the fitting value is lower than a preset fitting value, and if so, switch to the next sensor, and match the environmental feature information collected by the next sensor with the pre-stored prior map, and calculate the fitting value of the re-matching result; repeat the step of determining the fitting value until a sensor whose fitting value is not lower than the preset fitting value is found; for sensors whose fitting value is not lower than the preset fitting value, use the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environmental features.

[0025] In the third aspect, an embodiment of the present application provides an environmental feature-based positioning device, which includes a processor, a memory, and an environmental feature-based positioning program stored in the memory and executable by the processor, wherein when the environmental feature-based positioning program is executed by the processor, the steps of the environmental feature-based positioning method described in some of the above embodiments are implemented.

[0026] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a positioning program based on environmental features is stored. When the positioning program based on environmental features is executed by a processor, the steps of the positioning method based on environmental features as described in some of the above embodiments are implemented.

[0027] The beneficial effects brought by the technical solution provided in the embodiments of the present application include:

[0028] By matching the environmental feature information collected by the sensor with the pre-stored prior map, the Gauss-Newton algorithm is used to optimize the matching process to find the map subset that best matches the current environmental features; based on the historical true values ​​of the vehicle's posture and the true values ​​of the feature point coordinates provided by the matched map subset, the transformation matrix between the point pairs is calculated to solve the current vehicle's posture. Among them, the Gauss-Newton algorithm is used to ensure that the matched map subset can maintain accuracy while reducing the amount of calculation and improving the operation speed, continuously reducing the spatial position difference or distance between the feature points in the real-time sensor data and the corresponding feature points in the map subset, solving the current posture, and thus improving the accuracy and reliability of the positioning system. It avoids the problems of poor satellite signals and accumulated deviations in track estimation methods in traditional global positioning system methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flow chart of an embodiment of a positioning method based on environmental features of the present application;

[0030] Figure 2 This is a schematic diagram of the hardware structure of the environmental feature-based positioning device involved in the embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0032] First, some technical terms in the present application are explained to facilitate those skilled in the art to understand the present application.

[0033] Environmental characteristic information: Environmental characteristic information refers to the data captured by sensors from the surrounding environment that can be used to describe and identify environmental characteristics.

[0034] Prior map: A pre-stored prior map is a map containing environmental information, which provides prior knowledge about static and dynamic elements such as roads, traffic signs, obstacles, etc. This map is usually generated through pre-data collection and processing, and is used to assist the positioning, navigation, and decision-making process of autonomous vehicles.

[0035] Map subset: A map subset refers to a portion of map area or content selected from a pre-stored prior map.

[0036] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0037] In a first aspect, an embodiment of the present application provides a positioning method based on environmental features.

[0038] In one embodiment, referring to Figure 1 , Figure 1 This is a flow chart of an embodiment of the method for positioning based on environmental features of this application. Figure 1 As shown, the positioning method based on environmental features includes:

[0039] S100: matching the environmental feature information collected by the sensor with the pre-stored prior map, and optimizing the matching process using the Gauss-Newton algorithm to find a map subset that best matches the current environmental features;

[0040] S200: Calculate the transformation matrix between point pairs based on the historical true values ​​of the vehicle's posture and the true values ​​of the feature point coordinates provided by the matched map subset, and solve the current posture of the vehicle.

[0041] For example, for the problem of matching LiDAR (laser radar) point cloud frames with maps, the Gauss-Newton iterative algorithm is used. This algorithm is mainly used to solve nonlinear least squares problems, that is, to minimize the sum of squares of errors. The following is the specific process and formula of the Gauss-Newton algorithm:

[0042] Initialization: Set the initial value X0 of the pose (including position and attitude), and set the number of iterations k = 0. Set the convergence conditions, such as the maximum number of iterations or the error threshold.

[0043] Calculate the error: For each LiDAR point p i , calculate its corresponding point q in the map i , calculation error e i =q i -p i .

[0044] Construct the Jacobian matrix: For each error e i , calculate its Jacobian matrix J about the pose x i , all J i Stacked into the total Jacobian matrix J.

[0045] Calculate the iteration step: Calculate the stack of error vectors e = [e1, e2, ..., e n ] T , use the Gauss-Newton formula to calculate the iteration step length Δx=(J T J) -1 J Te. Among them, Δx: the iterative update of the pose, is a vector that represents the direction and distance to be moved in the pose space. J: Jacobian matrix, is a matrix, each row of which corresponds to the partial derivative of an error term with respect to the pose. The Jacobian matrix describes how the error changes with the change of pose. J T : The transpose of the Jacobian matrix, that is, the rows and columns of the Jacobian matrix are swapped. T J: The transpose of the Jacobian matrix multiplied by itself, the result is a square matrix. (J T J) -1 : The inverse matrix of the product of the transpose of the Jacobian matrix and itself. This inverse matrix is ​​used in the Gauss-Newton algorithm to calculate the update direction of the pose. e: Error vector, which is a vector that represents the difference between the predicted point cloud data and the actual point cloud data under the current pose. J T e: The product of the transpose of the Jacobian matrix and the error vector. The result is a vector that represents the direction in which we should move in pose space to reduce the error.

[0046] Update pose: Update pose x k+1 =x k +Δx.

[0047] Check convergence: If the convergence condition is met (for example, |Δx| is less than a certain threshold, or the maximum number of iterations is reached), stop the iteration. Otherwise, increase the number of iterations k=k+1 and return to the step to calculate the error. Among them, |Δx| represents its modulus.

[0048] In this embodiment, the Gauss-Newton algorithm is an iterative optimization algorithm for solving nonlinear least squares problems. In the process of matching the environmental feature information collected by the sensor with the pre-stored prior map, the Gauss-Newton algorithm can minimize the sum of squares of the residual vector by iteratively updating the posture parameters, thereby finding the map subset that best matches the current environmental features. According to the historical posture true value of the vehicle provided by the matched map subset and the true value of the feature point coordinates, the transformation matrix between the point pairs is calculated, and the current vehicle posture is solved. The Gauss-Newton algorithm continuously reduces the spatial position difference or distance between the feature points in the real-time sensor data and the corresponding feature points in the map subset, solves the current posture, and thus improves the accuracy and reliability of the positioning system. It avoids the problems of poor satellite signals and accumulated deviations in the track estimation method in the traditional global positioning system method.

[0049] Furthermore, in one embodiment, in S100, the following steps are included:

[0050] S101: Matching the environmental feature information collected by the sensor with the pre-stored priori map, and calculating the fitting value of the initial matching result;

[0051] S102: Determine whether the fitting value is lower than a preset fitting value. If so, switch to the next sensor, match the environmental feature information collected by the next sensor with the pre-stored priori map, and calculate the fitting value of the re-matching result;

[0052] S103: Repeat the step of determining the fitting value until a sensor whose fitting value is not less than a preset fitting value is found;

[0053] S104: For sensors whose fitting values ​​are not less than a preset fitting value, the Gauss-Newton algorithm is used to optimize the matching process to find a map subset that best matches the current environment characteristics.

[0054] In this embodiment, the strategy of switching sensors is intended to improve positioning accuracy. The specific strategy is as follows:

[0055] Initial matching and evaluation: First, the environmental feature information collected by the sensor is matched with the pre-stored prior map, and the fitting value of the initial matching result is calculated. This step is to evaluate the accuracy of the positioning information provided by the current sensor.

[0056] Fitting value judgment: If the fitting value of the initial matching result is lower than the preset fitting value threshold, it means that the positioning information of the current sensor may not be accurate enough. At this time, the strategy will switch to the next sensor.

[0057] Sensor switching and rematching: After switching the sensor, the environmental feature information collected by the new sensor is matched with the pre-stored prior map, and the fitting value of the rematching result is calculated. This step is to evaluate the accuracy of the positioning information provided by the new sensor.

[0058] Repeat judgment and switching: If the fitting value of the re-matching result is still lower than the preset fitting value threshold, the strategy will continue to switch to the next sensor and repeat the above matching and evaluation process until a sensor with a fitting value not lower than the preset fitting value is found.

[0059] Optimization and final matching: For sensors whose fitting values ​​are not less than the preset fitting values, the strategy uses the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environment characteristics, thereby improving positioning accuracy.

[0060] For example, the sensor (lidar) matching is divided into three categories: front radar, rear radar and full radar (front radar and rear radar). The matching of each radar is judged by the fitting value. When the radar matching is lower than the threshold, other radar matching methods are enabled. This logic can enhance the positioning robustness. First, all radars are matched; second, when the fitting value after all radars are matched is lower than the threshold, the front radar is used for matching; third, when the fitting value after the current radar is matched is lower than the threshold, the rear radar is used for matching. After actual testing, the threshold is 0.35.

[0061] Furthermore, in one embodiment, before S100, the following steps are included:

[0062] S000: Acquire environmental feature information collected by the sensor, and pre-process the collected environmental feature information using a voxel filtering algorithm to reduce the density of point cloud data and retain point cloud data representing environmental features.

[0063] In this embodiment, the environmental feature information processing flow is as follows: Use sensors (such as laser radar) to collect environmental feature information. Preprocessing: Apply voxel filtering algorithm to preprocess the collected environmental feature information. Reduce the density of point cloud data while retaining point cloud data representing environmental features. Feature extraction: Extract key environmental features from the preprocessed point cloud data. To achieve more accurate environmental perception and positioning.

[0064] Furthermore, in one embodiment, before S100, the following steps are included:

[0065] S000: Obtain environmental feature information collected by the sensor, and use a random filtering algorithm to pre-process the collected environmental feature information to reduce the density of point cloud data.

[0066] In this embodiment, the environmental feature information is obtained by using sensors to collect surrounding environment data and extract key features such as distance, angle, etc. The random filtering algorithm is applied to reduce the density of point cloud data and randomly select part of the point cloud data for retention. The filtering ratio or number is determined, and the specified ratio / number of point clouds are randomly selected, and the unselected point cloud data is deleted, so as to significantly reduce the density of point cloud data and improve the efficiency of subsequent processing.

[0067] Furthermore, in one embodiment, before S100, the following steps are included:

[0068] The environmental feature information collected by the sensor is obtained, and the Euclidean distance point cloud filtering method is used to remove the data that is not within the set threshold range in the environmental feature information collected by the sensor.

[0069] In this embodiment, for the point cloud filtering algorithm, this method uses the Euclidean distance point cloud filtering method to remove the data that is too far or too close in each frame of the environmental feature information. First, set the too close point threshold τ and the too far point threshold according to the specific application environment of the algorithm. Then calculate the Euclidean distance of the point [x, y, z] as follows,

[0070]

[0071] If the Euclidean distance of the point does not satisfy Then remove this point.

[0072] Among them, the voxel filtering algorithm can be before or after the step of the Euclidean distance point cloud filtering method, depending on specific needs. The voxel filtering algorithm is mainly used for the preprocessing of point cloud data to optimize the subsequent processing process by reducing the density of the data. It can downsample the point cloud to the same spacing without destroying the geometric structure of the point cloud itself, and remove a certain degree of noise points and outliers. The Euclidean distance point cloud filtering method is a distance-based point cloud filtering method that calculates the Euclidean distance between each point in the point cloud and removes points that are too far or too close to retain point cloud data representing environmental characteristics.

[0073] Furthermore, in one embodiment, before S100, the following steps are included:

[0074] S000: Obtain environmental feature information collected by lidar or visual camera.

[0075] In this embodiment, the laser radar provides high-precision and high-resolution environmental data, and is particularly good at generating three-dimensional maps and images. The visual camera obtains RGBD information of the environment. RGBD is an image or data format that contains color (RGB) and depth (Depth) information. It is suitable for indoor and outdoor environments, but there are blind spots and it is easily affected by light. The laser radar emits laser pulses and receives reflected pulses, calculates the distance, and combines the POS system data to obtain the three-dimensional coordinates of the ground point. The visual camera captures the environmental image and extracts feature information through image processing algorithms.

[0076] In the second aspect, an embodiment of the present application also provides a positioning device based on environmental features, which includes: a map subset matching module, which is used to match the environmental feature information collected by the sensor with a pre-stored prior map, and use the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environmental features; a current posture calculation module, which is used to calculate the transformation matrix between point pairs based on the historical posture true value of the vehicle and the true value of the feature point coordinates provided by the matched map subset, and solve the current vehicle posture.

[0077] Furthermore, in one embodiment, the current posture calculation module is also used to match the environmental feature information collected by the sensor with a pre-stored prior map, and calculate the fitting value of the initial matching result; determine whether the fitting value is lower than a preset fitting value, and if so, switch to the next sensor, and match the environmental feature information collected by the next sensor with the pre-stored prior map, and calculate the fitting value of the re-matching result; repeat the steps of determining the fitting value until a sensor whose fitting value is not lower than the preset fitting value is found; until a sensor whose fitting value is not lower than the preset fitting value is found; for sensors whose fitting value is not lower than the preset fitting value, use the Gauss-Newton algorithm to optimize the matching process to find a map subset that best matches the current environmental features.

[0078] Among them, the functional implementation of each module in the above-mentioned environmental feature-based positioning device corresponds to the various steps in the above-mentioned environmental feature-based positioning method embodiment, and its functions and implementation processes will not be repeated here one by one.

[0079] In a third aspect, an embodiment of the present application provides a positioning device based on environmental features. The positioning device based on environmental features may be a personal computer (PC), a laptop computer, a server, or other device with data processing capabilities.

[0080] Reference Figure 2 , Figure 2 Schematic diagram of the hardware structure of the positioning device based on environmental features involved in the embodiment of the present application. In the embodiment of the present application, the positioning device based on environmental features may include a processor, a memory, a communication interface and a communication bus.

[0081] The communication bus may be of any type and is used to interconnect the processor, the memory, and the communication interface.

[0082] The communication interface includes input / output (I / O) interface, physical interface and logical interface, etc., which are used to realize the interconnection of devices inside the positioning device based on environmental features, and the interface used to realize the interconnection between the positioning device based on environmental features and other devices (such as other computing devices or user devices). The physical interface can be an Ethernet interface, a fiber optic interface, an ATM interface, etc.; the user device can be a display, a keyboard, etc.

[0083] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0084] The processor may be a general-purpose processor, and the general-purpose processor may call the environment feature-based positioning program stored in the memory, and execute the environment feature-based positioning method provided by the embodiment of the present application. For example, the general-purpose processor may be a central processing unit (CPU). The method executed when the environment feature-based positioning program is called may refer to the various embodiments of the environment feature-based positioning method of the present application, which will not be repeated here.

[0085] Those skilled in the art will understand that Figure 2 The hardware structure shown in the figure does not constitute a limitation on the present application, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0086] In a fourth aspect, an embodiment of the present application also provides a readable storage medium.

[0087] The readable storage medium of the present application stores a positioning program based on environmental features, wherein when the positioning program based on environmental features is executed by a processor, the steps of the positioning method based on environmental features as described above are implemented.

[0088] Among them, the method implemented when the environmental feature-based positioning program is executed can refer to the various embodiments of the environmental feature-based positioning method of the present application, and will not be repeated here.

[0089] It should be noted that the serial numbers of the above-mentioned embodiments of the present application are only for description and do not represent the advantages or disadvantages of the embodiments.

[0090] The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices. The terms "first", "second" and "third" are used to distinguish different objects, etc., and do not represent a sequence, nor do they limit "first", "second" and "third" to different types.

[0091] In the description of the embodiments of the present application, "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a specific way.

[0092] In the description of the embodiments of the present application, unless otherwise specified, “ / ” means or, for example, A / B can mean A or B; the “and / or” in the text is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, “multiple” refers to two or more than two.

[0093] In some processes described in the embodiments of the present application, multiple operations or steps that appear in a specific order are included, but it should be understood that these operations or steps may not be executed in the order in which they appear in the embodiments of the present application or in parallel, and the sequence number of the operation is only used to distinguish the different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed in sequence or in parallel, and these operations or steps may be combined.

[0094] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD) as described above, and includes a number of instructions for a terminal device to execute the methods described in each embodiment of the present application.

[0095] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A positioning method based on environmental features, It is characterized in that The environmental feature-based positioning method comprises: Match the environmental feature information collected by the sensor with the pre-stored prior map, and use the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environmental features; According to the historical true values ​​of the vehicle's posture and the true values ​​of the feature point coordinates provided by the matched map subset, the transformation matrix between the point pairs is calculated to solve the current vehicle's posture.

2. The method for positioning based on environmental features as claimed in claim 1, It is characterized in that The environmental feature information collected by the sensor is matched with the pre-stored prior map, and the Gauss-Newton algorithm is used to optimize the matching process to find the map subset that best matches the current environmental features, including: Matching the environmental feature information collected by the sensor with the pre-stored priori map, and calculating the fitting value of the initial matching result; Determine whether the fitting value is lower than the preset fitting value. If so, switch to the next sensor, match the environmental feature information collected by the next sensor with the pre-stored prior map, and calculate the fitting value of the re-matching result; Repeat the step of determining the fitting value until a sensor whose fitting value is not less than the preset fitting value is found; For sensors whose fitting values ​​are not lower than the preset fitting values, the Gauss-Newton algorithm is used to optimize the matching process to find the map subset that best matches the current environment characteristics.

3. The method for positioning based on environmental features as claimed in claim 1, It is characterized in that Before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes: The environmental feature information collected by the sensor is obtained, and the collected environmental feature information is preprocessed using a voxel filtering algorithm to reduce the density of the point cloud data and retain the point cloud data representing the environmental features.

4. The method for positioning based on environmental features as claimed in claim 1, It is characterized in that Before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes: Obtain the environmental feature information collected by the sensor, use the random filtering algorithm to pre-process the collected environmental feature information, and reduce the density of point cloud data.

5. The method for positioning based on environmental features as claimed in claim 1, It is characterized in that Before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes: The environmental feature information collected by the sensor is obtained, and the Euclidean distance point cloud filtering method is used to remove the data that is not within the set threshold range in the environmental feature information collected by the sensor.

6. The method for positioning based on environmental features as claimed in claim 1, It is characterized in that Before matching the environmental feature information collected by the sensor with the pre-stored priori map and optimizing the matching process using the Gauss-Newton algorithm to find the map subset that best matches the current environmental features, the method includes: Obtain environmental feature information collected by lidar or visual camera.

7. A positioning device based on environmental features, It is characterized in that The positioning device based on environmental features comprises: A map subset matching module is used to match the environmental feature information collected by the sensor with the pre-stored prior map, and use the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environmental features; The current posture calculation module is used to calculate the transformation matrix between point pairs based on the historical posture true value of the vehicle and the true value of the feature point coordinates provided by the matched map subset, and solve the current vehicle posture.

8. The positioning device based on environmental features as claimed in claim 7, It is characterized in that The current posture calculation module is also used to match the environmental feature information collected by the sensor with the pre-stored prior map, and calculate the fitting value of the initial matching result; determine whether the fitting value is lower than the preset fitting value, if so, switch to the next sensor, and match the environmental feature information collected by the next sensor with the pre-stored prior map, and calculate the fitting value of the re-matching result; repeat the step of determining the fitting value until a sensor whose fitting value is not lower than the preset fitting value is found; for sensors whose fitting value is not lower than the preset fitting value, use the Gauss-Newton algorithm to optimize the matching process to find the map subset that best matches the current environmental features.

9. A positioning device based on environmental features, It is characterized in that The environmental feature-based positioning device includes a processor, a memory, and an environmental feature-based positioning program stored in the memory and executable by the processor, wherein when the environmental feature-based positioning program is executed by the processor, the steps of the environmental feature-based positioning method as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a positioning program based on environmental features, wherein when the positioning program based on environmental features is executed by a processor, the steps of the positioning method based on environmental features as described in any one of claims 1 to 6 are implemented.

Images