A container positioning error self-calibration method based on LoRa+GNSS joint work

Abstract

The present invention relates to the field of intelligent logistics positioning technology, and specifically to a container positioning error self-calibration method based on the combined operation of LoRa and GNSS. The method comprises: deploying a dual-mode sensing unit integrating a LoRa communication module and a door magnetic state sensor on the inside of the container door frame, and fixing a GNSS positioning module at an asymmetric position on the container roof; collecting the GNSS raw multipath error spectrum and the LoRa signal cavity attenuation factor when the door magnetic state triggers the door opening and closing; inputting these two types of data into a metal shielding coefficient calibration model to output a dynamic shielding coefficient matrix; and then, based on this matrix, performing point-by-point error reverse compensation on the GNSS trajectory within 30 seconds before and after the door opening and closing to generate a self-calibrated positioning coordinate sequence. The method fully utilizes the door magnetic triggering events and the changes in the dual-mode signal characteristics to achieve closed-loop optimization of GNSS error modeling and trajectory calibration in shielded environments, significantly improving the positioning accuracy and trajectory continuity of containers in complex environments.

Description

Technical Field

[0001] The present invention relates to the field of intelligent logistics positioning technology, and in particular to a container positioning error self-calibration method based on LoRa+GNSS joint operation. Background Art

[0002] In modern multimodal transport and intelligent logistics systems, containers serve as core transport units, and their full-cycle precise positioning capabilities are crucial for transport safety, asset management, and scheduling optimization. The Global Positioning System (GNSS), as a mainstream spatial positioning method, has been widely used for container tracking. However, because container structures generally utilize large, enclosed metal sheets, their complex cavity characteristics significantly interfere with GNSS signal propagation. This is particularly true in scenarios where doors are frequently opened and closed, and the shielding environment undergoes drastic changes. This can easily lead to enhanced GNSS multipath effects, increased signal loss rates, and positioning drift. Furthermore, the magnetic opening and closing status of container door sensors is a typical time-series event in actual use and is highly observable. Combining this with GNSS signal fluctuations to analyze and construct a dynamic shielding modeling mechanism will provide a key entry point for error compensation.

[0003] Existing technologies primarily rely on external augmentation base stations, inertial navigation assistance systems, or signal filtering algorithms to optimize GNSS positioning accuracy. These methods struggle to adapt to the dynamic shadowing caused by metallic structures within containers. Furthermore, a small number of studies have attempted to utilize IoT communication modules (such as LoRa and NB-IoT) for assisted positioning. However, these efforts often fail to incorporate the actual impact of structural shadowing and lack the ability to collaboratively model GNSS error characteristics. This results in inadequately targeted, robust, and timely error compensation mechanisms. Summary of the Invention

[0004] The present invention provides a container positioning error self-calibration method based on the joint operation of LoRa+GNSS. It provides a self-calibration method that can integrate the characteristics of GNSS and low-frequency communication signals, respond to changes in door magnetic state, and have real-time error correction capabilities, so as to improve the positioning stability and trajectory continuity of containers in shielded environments.

[0005] A container positioning error self-calibration method based on LoRa+GNSS joint operation includes the following steps:

[0006] S1, dual-mode sensing unit deployment: A dual-mode sensing unit integrating a LoRa communication module and a door magnetic status sensor is deployed on the inside of the container door frame, and a GNSS positioning module is fixed in an asymmetric position on the top of the container;

[0007] S2, door status trigger signal acquisition: When the door magnetic status sensor detects the opening and closing of the container door, the original multipath error spectrum of the GNSS positioning module and the LoRa signal cavity attenuation factor of the dual-mode sensing unit are synchronously collected;

[0008] S3, dynamic calibration of metal shielding coefficient: input the original multipath error spectrum and the LoRa signal cavity attenuation factor into the pre-trained metal shielding coefficient calibration model, and output the dynamic shielding coefficient matrix of the current container metal structure to the positioning signal;

[0009] S4, positioning trajectory back-calibration: Based on the dynamic shielding coefficient matrix, the GNSS positioning trajectory within 30 seconds before and after the door opening and closing action is reversely compensated for errors to generate a self-calibration positioning coordinate sequence.

[0010] Optionally, the S1 includes:

[0011] S11, dual-mode sensing unit deployment: Install the dual-mode sensing unit on the inner surface of the container door frame, ensuring that the antenna radiation surface of the LoRa communication module faces the center of the container cavity, and the magnetic sensing end of the door magnetic status sensor is aligned with the edge of the container door leaf;

[0012] S12, GNSS positioning module installation coordinate calculation: Using the container top plane as the reference coordinate system, calculate the installation coordinates that avoid the top angle symmetry axis and are greater than 15 cm away from the edge of the container.

[0013] Optionally, the S1 further includes:

[0014] S13, asymmetric position fixing: according to the installation coordinates, the GNSS positioning module is fixed at an asymmetric position on the top of the box, and the elevation angle of the receiving antenna is adjusted to be greater than 60 degrees.

[0015] Optionally, the S2 includes:

[0016] S21, door opening and closing action detection: when the door magnetic state sensor detects that the container door changes from a closed state to a continuously open state, a door state trigger signal is generated;

[0017] S22, dual-parameter synchronous acquisition: In response to the door state trigger signal, synchronously perform the following operations:

[0018] Collecting the original multipath error spectrum output by the GNSS positioning module;

[0019] Record the LoRa communication module received signal strength indicator value 5 seconds before the door opens and closes and 5 seconds after the door is fully opened;

[0020] S23, cavity attenuation factor calculation: Calculate the LoRa signal cavity attenuation factor according to the received signal strength indicator value.

[0021] Optionally, the LoRa signal cavity attenuation factor is calculated as:

[0022] ;

[0023] in, Indicates the average RSSI value before the door is closed. The RSSI value is the average RSSI value after the door is opened. RSSI represents the received signal strength indicator value.

[0024] Optionally, the S3 includes:

[0025] S31, time-space alignment of error spectrum: extract the time domain sequence of the original multipath error spectrum within the door opening and closing action time window, and align it with the LoRa signal cavity attenuation factor at the same timestamp;

[0026] S32, shielding coefficient calibration: input the aligned original multipath error spectrum and the LoRa signal cavity attenuation factor into a pre-trained metal shielding coefficient calibration model, and the metal shielding coefficient calibration model outputs the spatial distribution vector of the current metal shielding coefficient.

[0027] Optionally, the S3 further includes:

[0028] S33, dynamic matrix generation: mapping the spatial distribution vector into a 4×4 grid coefficient matrix according to the three-dimensional coordinates of the container, superimposing a timestamp mark and outputting a dynamic shielding coefficient matrix.

[0029] Optionally, the S4 includes:

[0030] S41, trajectory data extraction: obtaining the original coordinate sequence of the GNSS positioning trajectory within 30 seconds before and after the door opening and closing action;

[0031] S42, reverse error compensation: input the dynamic shielding coefficient matrix and the original coordinate sequence of the GNSS positioning trajectory into a compensation function, and calculate the compensated coordinate value point by point.

[0032] Optionally, the S4 further includes:

[0033] S43, calibration sequence generation: reorganize all calculated compensation coordinate values ​​in chronological order to generate a self-calibration positioning coordinate sequence with a time stamp.

[0034] Beneficial effects of the present invention:

[0035] This invention deploys a dual-mode sensing unit integrating a LoRa communication module and a door magnetic status sensor inside the container door frame, and installs a GNSS positioning module asymmetrically on the container roof, creating a highly sensitive sensing system suitable for complex metal structures. The LoRa communication module's antenna faces the interior of the cavity, enabling it to fully respond to changes in obscuration. The door magnetic status sensor accurately detects door opening and closing, and the GNSS module is installed away from symmetrical interference zones, significantly reducing multipath interference signal sources. This coordinated deployment ensures the system possesses a rapid triggering mechanism and the ability to detect spatial obscuration differences, laying a high-quality signal source foundation for subsequent error feature extraction and modeling.

[0036] This method, under precise timing control triggered by the door magnetic state, synchronously collects the GNSS raw multipath error spectrum and the LoRa signal cavity attenuation factor. These two data sets are mapped using a pre-trained metal shielding coefficient calibration model to generate a dynamic shielding coefficient matrix for the container's metal structure. This matrix, using door opening and closing behavior as the physical boundary, incorporates information about the propagation interference of different electromagnetic signals in the cavity environment. This matrix retains the high spatial resolution of GNSS while incorporating LoRa's highly sensitive perception of near-field shielding changes. This overcomes the inability of traditional GNSS errors to quantify shielding patterns, thereby enhancing the system's ability to accurately model changes in the metal environment.

[0037] The present invention extracts the original GNSS trajectory within 30 seconds before and after the door opens and closes, uses the dynamic shielding coefficient matrix as the input compensation factor, and utilizes a compensation function to perform inverse correction of the coordinate-level error of the trajectory point at each moment. This inverse compensation uses a dynamic matrix multiplied by a reference error vector to directly map the impact of the shielding structure into a change in coordinate displacement, effectively reducing GNSS positioning drift, jump points, and height anomalies caused by interference from the container's metal cavity. Experimental verification shows that this method can compress the positioning error from more than 5 meters to less than 1 meter in scenarios with strong multipath interference, significantly improving the trajectory availability and continuity of containers in typical scenarios such as storage and transshipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0039] Figure 1 Schematic diagram of a method flow in an embodiment of the present invention;

[0040] Figure 2 This is a schematic diagram of the S3 process of an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It is also noted that, to provide a more detailed description, the following embodiments are best and preferred embodiments, and those skilled in the art may employ alternative methods for implementing certain known technologies. Furthermore, the accompanying drawings are intended only to provide a more detailed description of the embodiments and are not intended to limit the present invention.

[0042] It should be noted that references in the specification to "one embodiment," "an embodiment," "exemplary embodiments," "some embodiments," etc. indicate that the described embodiments may include specific features, structures, or characteristics, but not necessarily every embodiment will include such specific features, structures, or characteristics. Furthermore, when specific features, structures, or characteristics are described in conjunction with an embodiment, it is within the knowledge of persons skilled in the relevant art to implement such features, structures, or characteristics in conjunction with other embodiments (whether or not explicitly described).

[0043] In general, terms can be understood, at least in part, from their use in context. For example, depending at least in part on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending at least in part on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0044] like Figure 1-Figure 2 As shown, a container positioning error self-calibration method based on LoRa+GNSS joint operation includes the following steps:

[0045] S1, dual-mode sensing unit deployment: A dual-mode sensing unit integrating a LoRa communication module and a door magnetic status sensor is deployed on the inside of the container door frame, and a GNSS positioning module is fixed in an asymmetric position on the top of the container;

[0046] S2, door status trigger signal acquisition: When the door magnetic status sensor detects the opening and closing of the container door, it synchronously collects the original multipath error spectrum of the GNSS positioning module and the LoRa signal cavity attenuation factor of the dual-mode sensing unit (calculated by the signal strength mutation rate before and after the door is opened and closed);

[0047] S3, dynamic calibration of metal shielding coefficient: The original multipath error spectrum and the LoRa signal cavity attenuation factor are input into the pre-trained metal shielding coefficient calibration model, and the dynamic shielding coefficient matrix of the current container metal structure to the positioning signal is output;

[0048] S4, positioning trajectory back-calibration: Based on the dynamic shielding coefficient matrix, the GNSS positioning trajectory within 30 seconds before and after the door opening and closing action is reversely compensated for the error to generate a self-calibration positioning coordinate sequence.

[0049] S1 includes:

[0050] To deploy the S11 dual-mode sensor unit, select a flat area on the inside of the container door frame as the mounting surface and securely install the dual-mode sensor unit, which integrates a LoRa communication module and a door magnetic status sensor. The LoRa communication module's antenna radiating surface should face the center of the container cavity to ensure a relatively uniform signal reflection environment within the container. The magnetic sensing end of the door magnetic status sensor should be precisely aligned with the edge of the container door leaf to enhance trigger sensitivity for door opening and closing.

[0051] To ensure the sensor triggering accuracy and the timing consistency of subsequent signal mutation detection, the system performs offline calibration on the trigger threshold of the door magnetic state sensor. The trigger conditions for door opening and closing action recognition are set as follows:

[0052] ;

[0053] in, is the LoRa signal strength when the door is closed, is the LoRa signal strength when the door is open, The threshold for judging signal mutation is 8dBm.

[0054] S12, GNSS positioning module installation coordinate calculation: To reduce the multipath interference and shielding effect of the container metal structure on the GNSS reception signal, the GNSS module needs to be fixed at an asymmetric position on the top of the container. Establish a three-dimensional reference coordinate system based on the container top plane. , where the origin is set to the front left corner of the box (door side), the X axis is along the length of the box, the Y axis is along the width of the box, and the Z axis is the vertical direction.

[0055] Candidate installation coordinates of the GNSS positioning module The following constraints should be met:

[0056] ;

[0057] in, 、 、 are the length, width and height of the container respectively. The installation coordinates of the GNSS module, excluding all or The position of the axis of symmetry.

[0058] For example, if the dimensions of a standard 20-foot container are:

[0059] 605.8cm, 243.8cm, 259.1cm, then the effective installation coordinate range of the GNSS module should be:

[0060] ;

[0061] S13, asymmetric position fixation: According to the installation coordinates determined in S12, select a set of asymmetric positions that meet the conditions on the top of the container, and use industrial-grade adhesive or riveted brackets to firmly install the GNSS positioning module. The elevation angle of the GNSS module's receiving antenna needs to be adjusted to greater than 60 degrees, that is, the angle with the horizontal plane. , in order to minimize the signal shielding from the metal structure at the edge of the box top. The adjusted GNSS antenna direction vector can be expressed as:

[0062] ;

[0063] in, is the elevation angle, and its value range is , It is the azimuth angle, which is set according to the orientation of the container.

[0064] The spatial deployment of the dual-mode sensing unit and the GNSS positioning module is completed in the above manner, which can provide a high-precision basic spatial reference frame for the subsequent door opening and closing trigger signal and error compensation data acquisition.

[0065] S2 includes:

[0066] S21, Door Opening and Closing Movement Detection: The door magnetic state sensor is embedded in the inner side of the container door frame. It monitors the physical opening and closing state of the door in real time through the relative displacement between its magnetic sensing end and the edge of the door leaf. When the sensor detects that the door body switches from the closed state (door magnetic remains engaged) to the continuously open state (magnetic field sensing interruption exceeds the set threshold time, such as 2 seconds), the system immediately generates a door state trigger signal, triggering downstream collection operations. The judgment rule for door state changes is expressed as follows:

[0067] ;

[0068] in, for The door magnetic state at that moment (1 means closed, 0 means open), The door is continuously open, which triggers the threshold time. It is usually set to 2 seconds. It is the door status trigger signal flag.

[0069] S22, dual-parameter synchronous acquisition: In response to the generation of the door status trigger signal, the system enters the dual-parameter synchronous acquisition mode and performs the following operations:

[0070] GNSS raw multipath error spectrum acquisition: Call the intermediate frequency data cache interface of the GNSS module to collect raw multipath error spectrum data before and after the door opening and closing states are switched, and record the pseudorange offset, code phase difference, and carrier phase jump characteristics of the satellite signal of each channel;

[0071] LoRa received signal strength indicator value sampling: 5 seconds before the door is completely closed (recorded as ) and 5 seconds after the door is fully opened (recorded as ) as the central time point, continuously collect the received signal strength indicator (RSSI) of the LoRa communication module near two time points. To reduce the interference of instantaneous fluctuations, set a fixed sampling window (such as ±1 second) at each time point and perform N samplings to obtain the average. The average value is calculated as follows:

[0072] ;

[0073] in, is the number of sampling times, set to 10, Indicates the RSSI value of the kth sampling.

[0074] S23, cavity attenuation factor calculation: The LoRa signal shows a significant signal strength mutation before and after the door is opened and closed due to the structural shielding change of the container metal cavity. This characteristic can be used to construct the LoRa signal cavity attenuation factor as a sensitive indicator of metal shielding changes.

[0075] The cavity attenuation factor is calculated according to the following formula:

[0076] ;

[0077] in, Indicates the average RSSI value before the door is closed. is the average RSSI value after the door is opened. The attenuation factor is the relative change percentage, reflecting the degree of LoRa signal shielding caused by the opening and closing of the container. RSSI represents the received signal strength indicator value.

[0078] Example description:

[0079] like The sampling average is , The sampling average is ,but:

[0080] ;

[0081] The calculated cavity attenuation factor, together with the original GNSS multipath error spectrum, is used as input parameters to the metal shielding coefficient calibration model to infer the shielding characteristics of the current metal structure.

[0082] S3 includes:

[0083] S31, time-space alignment of error spectrum: First, extract the time window of the original multipath error spectrum obtained in S2. Assume that the central timestamp of the door opening and closing action is , select a window with 5 seconds on each side and centered on the time The error spectrum data in the multipath error time series is constructed:

[0084] ;

[0085] in, For the Satellite signals in The pseudorange error value corresponding to the time, is the number of valid satellites participating in the fitting within the window.

[0086] At the same time, the LoRa signal cavity attenuation factor calculated in S2 is calculated according to the same timestamp Since the cavity attenuation factor is a global quantity between the two states of the door opening and closing, the normalized time calibration is performed in the model input as follows: ;

[0087] Finally, a joint input pair with temporal consistency is formed , used to drive the subsequent metal shading modeling process.

[0088] S32, masking coefficient calibration: the “original multipath error spectrum sequence” after time and space alignment and "LoRa signal cavity attenuation factor" The data is input into the pre-trained metal shielding coefficient calibration model. The model uses a deep feature fusion mechanism to mine the shielding response relationship of electromagnetic signals of different frequency bands in the metal cavity and outputs the spatial distribution vector of the metal shielding coefficient corresponding to the local area of ​​the container:

[0089] ;

[0090] in, is a spatial distribution vector with 16 elements, each , represents the relative shielding degree of the sub-area corresponding to the container, function It represents the inference mapping process of the metal shielding coefficient calibration model. The model structure can adopt a multi-channel CNN-Transformer fusion architecture or a temporal fusion network based on an LSTM-MLP hybrid decoder.

[0091] S33, Dynamic Matrix Generation: Transform the spatial distribution vector Remap the 2D projection grid onto the 3D container structure to construct a 4×4 gridding coefficient matrix , its elements and Each component in - corresponds to:

[0092] ;

[0093] To support subsequent time series trajectory backtracking and dynamic compensation, each frame of the mask matrix needs to be superimposed with the corresponding timestamp , forming a dynamic shielding coefficient matrix with the following structure:

[0094] ;

[0095] This matrix is ​​output as the time-tagged spatial masking state and passed to S4 for GNSS trajectory error compensation inversion operation.

[0096] The dynamic shielding coefficient matrix in the present invention In the form of a structure, containing a gridded masking matrix and its generation timestamp , which is referenced in subsequent steps In fact, both Middle matrix part Right now:

[0097] ;

[0098] Indicates definition or assignment.

[0099] S4 includes:

[0100] S41, trajectory data extraction: Extract the original coordinate sequence of the GNSS positioning trajectory within 30 seconds before and after the container door opening and closing action from the cache of the GNSS positioning module, and set the door opening and closing event triggering center time as , then the extraction time interval is:

[0101] ;

[0102] Construct the original 3D trajectory point set:

[0103] ;

[0104] The sampling interval of the trajectory points is determined by the output frequency of the GNSS module, which is usually 1 Hz (one frame per second), so the total number of sampling points is about 61.

[0105] S42, Error Reverse Compensation: To offset the shielding interference caused by the container metal structure to the GNSS signal, a dynamic shielding coefficient matrix is ​​used. Perform reverse error compensation on each frame of original trajectory points. Introduce the preset GNSS reference error vector:

[0106] , is the preset GNSS reference error vector;

[0107] This quantity is an empirically defined systematic offset estimate that represents the direction and magnitude of common GNSS system errors (such as Z-axis deviation and Y-axis drift due to metal reflections) when no shielding compensation is applied. The compensation calculation formula is as follows:

[0108] ;

[0109] in,( , , ) is the original positioning trajectory coordinate, ( , , ) is the compensation positioning coordinate. The compensation calculation is performed independently in each time frame to form a point-by-point correction sequence.

[0110] Example description: Assume that the original coordinates at a certain moment are , the preset reference error vector is: ;

[0111] Dynamic shading coefficient matrix The first three columns are used to compensate for the vector transformation, which is assumed to be simplified as:

[0112] ;

[0113] but:

[0114] ;

[0115] This coordinate is the self-calibration coordinate after inverse compensation of the shading error.

[0116] S43, calibration sequence generation: all the moments The compensation coordinate values ​​are combined into a compensation trajectory sequence:

[0117] ;

[0118] This sequence is the final output self-calibrated positioning coordinate sequence, which has higher spatial continuity and occlusion robustness, and can be used for precise dynamic positioning and trajectory tracing scenarios of containers.

[0119] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0120] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A container positioning error self-calibration method based on LoRa+GNSS joint work, characterized in that: The following steps are involved: S1, dual-mode sensing unit deployment: A dual-mode sensing unit integrating a LoRa communication module and a door magnetic status sensor is deployed inside the container door frame, and a GNSS positioning module is fixed in an asymmetric position on the container roof; S2, door status trigger signal acquisition: When the door magnetic status sensor detects the opening and closing of the container door, the original multipath error spectrum of the GNSS positioning module and the LoRa signal cavity attenuation factor of the dual-mode sensing unit are synchronously collected; S3, dynamic calibration of metal shielding coefficient: input the original multipath error spectrum and the LoRa signal cavity attenuation factor into the pre-trained metal shielding coefficient calibration model, and output the dynamic shielding coefficient matrix of the current container metal structure to the positioning signal; S4, positioning trajectory back-calibration: Based on the dynamic shielding coefficient matrix, the GNSS positioning trajectory within 30 seconds before and after the door opening and closing action is reversely compensated for errors to generate a self-calibration positioning coordinate sequence.

2. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 1 is characterized in that: Said S1 comprises: S11, dual-mode sensing unit deployment: Install the dual-mode sensing unit on the inner surface of the container door frame, ensuring that the antenna radiation surface of the LoRa communication module faces the center of the container cavity, and the magnetic sensing end of the door magnetic status sensor is aligned with the edge of the container door leaf; S12, GNSS positioning module installation coordinate calculation: Using the container top plane as the reference coordinate system, calculate the installation coordinates that avoid the top angle symmetry axis and are greater than 15 cm away from the edge of the container.

3. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 2 is characterized in that: Said S1 further comprises: S13, asymmetric position fixing: according to the installation coordinates, the GNSS positioning module is fixed at an asymmetric position on the top of the box, and the elevation angle of the receiving antenna is adjusted to be greater than 60 degrees.

4. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 3 is characterized in that: The S2 includes: S21, door opening and closing action detection: when the door magnetic state sensor detects that the container door changes from a closed state to a continuously open state, a door state trigger signal is generated; S22, dual-parameter synchronous acquisition: In response to the door state trigger signal, synchronously perform the following operations: Collecting the original multipath error spectrum output by the GNSS positioning module; Record the LoRa communication module received signal strength indicator value 5 seconds before the door opens and closes and 5 seconds after the door is fully opened; S23, cavity attenuation factor calculation: Calculate the LoRa signal cavity attenuation factor according to the received signal strength indicator value.

5. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 4 is characterized in that: The LoRa signal cavity attenuation factor is calculated as: ； in, Indicates the average RSSI value before the door is closed. The RSSI value is the average RSSI value after the door is opened. RSSI represents the received signal strength indicator value.

6. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 5 is characterized in that: The S3 includes: S31, time-space alignment of error spectrum: extract the time domain sequence of the original multipath error spectrum within the door opening and closing action time window, and align it with the LoRa signal cavity attenuation factor at the same timestamp; S32, shielding coefficient calibration: input the aligned original multipath error spectrum and the LoRa signal cavity attenuation factor into a pre-trained metal shielding coefficient calibration model, and the metal shielding coefficient calibration model outputs the spatial distribution vector of the current metal shielding coefficient.

7. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 6 is characterized in that: Said S3 further comprises: S33, dynamic matrix generation: mapping the spatial distribution vector into a 4×4 grid coefficient matrix according to the three-dimensional coordinates of the container, superimposing a timestamp mark and outputting a dynamic shielding coefficient matrix.

8. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 7 is characterized in that: The S4 includes: S41, trajectory data extraction: obtaining the original coordinate sequence of the GNSS positioning trajectory within 30 seconds before and after the door opening and closing action; S42, reverse error compensation: input the dynamic shielding coefficient matrix and the original coordinate sequence of the GNSS positioning trajectory into a compensation function, and calculate the compensated coordinate value point by point.

9. The container positioning error self-calibration method based on LoRa+GNSS joint operation according to claim 8 is characterized in that: Said S4 further comprises: S43, calibration sequence generation: reorganize all calculated compensation coordinate values ​​in chronological order to generate a self-calibration positioning coordinate sequence with a time stamp.

Images