Goods Dynamic Tracking System and Method Using Satellite Precise Positioning

Abstract

The present invention discloses a goods dynamic tracking system and method using satellite precise positioning, specifically relating to the technical field of goods dynamic tracking, and includes the following steps: By collecting real-time goods positioning data and performing accuracy analysis, it is judged whether there are early signs of insufficient positioning accuracy; if there are early signs, dynamic tracking features are extracted and the monitoring situation is evaluated to form a feature data group, which is input into a pre-trained machine learning model and classified as low, medium, and high-quality tracking according to the signal environment and path stability; for high-quality and low-quality tracking, the positioning method is optimized to maintain accuracy and improve system efficiency, ensuring the accuracy and stability of goods dynamic tracking; the present invention can ensure the quality of positioning signals and path stability by flexibly adjusting the positioning method, thereby improving the accuracy of goods tracking, while reducing the use of redundant technologies, thus avoiding unnecessary errors and computational burdens, and improving the system response speed and real-time performance.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic tracking of goods. More specifically, the present invention relates to a dynamic tracking system and method for goods using satellite precise positioning. Background Art

[0002] With the increasing development of modern logistics and goods transportation, an accurate dynamic tracking system plays a crucial role in improving the efficiency of goods transportation, reducing transportation costs, and enhancing customer satisfaction. Especially in complex environments (such as urban canyons, tunnels, underground parking lots, etc.), how to ensure the positioning accuracy and stability of goods has become a challenge for modern goods tracking technology.

[0003] Traditional methods for dynamic tracking of goods mainly rely on single positioning technologies, such as Global Navigation Satellite System (GNSS), Wireless Local Area Network (Wi-Fi), Ultra-Wideband (UWB), etc. These technologies have their respective advantages, but single technologies are restricted under certain environmental conditions (such as signal occlusion, interference, etc.). To solve this problem, the industry has gradually introduced multi-modal fusion positioning technologies to improve the positioning accuracy and stability by combining the advantages of multiple positioning technologies.

[0004] However, although multi-modal fusion positioning technologies can improve positioning accuracy, they also bring problems such as increased computational burden and aggravated energy consumption. Especially in high-quality tracking scenarios, when the positioning accuracy is already sufficient, redundant positioning technologies will lead to unnecessary waste of computing resources. Therefore, how to intelligently optimize the positioning method under different tracking quality conditions has become an urgent problem to be solved in the current goods dynamic tracking technology. For this reason, the present invention provides a dynamic tracking system and method for goods using satellite precise positioning in order to solve the above problems. Summary of the Invention

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for dynamic tracking of goods using satellite precise positioning, comprising the following steps:

[0007] During the goods transportation process, real-time collect the positioning data of the goods to obtain a positioning data set;

[0008] Based on the positioning data, perform positioning accuracy analysis, and determine whether there are early signs of insufficient positioning accuracy during the goods transportation process through the positioning accuracy analysis result;

[0009] When there are early signs of insufficient positioning accuracy, dynamic tracking feature extraction is performed, and then the monitoring of cargo dynamic tracking is evaluated to provide a feature data set for subsequent current dynamic tracking quality classification. The feature data set is imported into a pre-trained machine learning model, and the monitoring of cargo dynamic tracking is classified into low-quality tracking, medium-quality tracking, or high-quality tracking according to the degree to which the positioning signal is affected by the environment and the stability of the positioning data path;

[0010] For the monitoring of high-quality tracking and low-quality tracking, the current positioning method is optimized to keep the cargo positioning accuracy at a standard state, and the optimized positioning method is used to dynamically track the cargo, store the tracking data, and achieve traceability and remote monitoring.

[0011] In a preferred embodiment, performing positioning accuracy analysis based on positioning data refers to:

[0012] In the current time window, the average deviation between the current positioning data and the reference position inferred based on the historical trajectory is calculated to measure the overall positioning accuracy and obtain the average offset error value; the standard deviation of the rate of change of the cargo movement speed at adjacent time points in the current time window is calculated to obtain the speed jump value; the standard deviation of the time interval between two consecutive positioning data in the current time window is calculated to obtain the sampling jump value, and then the average offset error value, speed jump value and sampling jump value are summarized together as the actual positioning vector.

[0013] In a preferred embodiment, judging whether there are early signs of insufficient positioning accuracy during cargo transportation through the positioning accuracy analysis result refers to:

[0014] The actual positioning vector is compared with the preset standard vector, the Euclidean distance between the actual positioning vector and the preset standard vector is calculated, and the Euclidean distance is compared with the preset warning threshold. If the Euclidean distance is greater than the preset warning threshold, it is determined that there are early signs of insufficient positioning accuracy. If the Euclidean distance is less than or equal to the preset warning threshold, it is determined that there are no early signs of insufficient positioning accuracy.

[0015] In a preferred embodiment, the cargo dynamic tracking feature extraction includes positioning environment information and positioning stability information, and then a spatial signal environment index is generated based on the positioning environment information, and the spatial signal environment index is used to evaluate the degree of influence of the environment in which the cargo is located during transportation on the positioning signal; a signal path integrity index is generated based on the positioning stability information, and the signal path integrity index is used to evaluate the stability of the positioning data path of the cargo during transportation; the spatial signal environment index and the signal path integrity index are used together to reflect the monitoring status of cargo dynamic tracking.

[0016] In a preferred embodiment, the acquisition logic of the spatial signal environment index is as follows:

[0017] During the transportation of goods, collect the following four types of environmental information based on the currently adopted positioning method: received signal strength coefficient, available signal quality coefficient, signal interference coefficient, and signal drift coefficient. To ensure the unified processing of data from different sensors, normalize the received signal strength coefficient, available signal quality coefficient, signal interference coefficient, and signal drift coefficient respectively:

[0018] Adopt the minimum-maximum normalization processing method for the received signal strength coefficient:

[0019] ; is the normalized result of the received signal strength coefficient corresponding to the positioning method type i, is the received signal strength coefficient corresponding to the positioning method type i, , are the minimum and maximum values of the standard received signal strength coefficient corresponding to the positioning method type i;

[0020] Adopt the proportional normalization processing method for the available signal quality coefficient:

[0021] ; is the normalized result of the available signal quality coefficient corresponding to the positioning method type i, is the available signal quality coefficient corresponding to the positioning method type i, is the standard available signal quality coefficient value corresponding to the positioning method type i;

[0022] Adopt the exponential decay normalization processing method for the signal interference coefficient:

[0023] ; is the normalized result of the signal interference coefficient corresponding to the positioning method type i, is the signal interference coefficient corresponding to the positioning method type i;

[0024] Adopt the exponential decay normalization processing method for the signal drift coefficient:

[0025] ; is the normalized result of the signal drift coefficient corresponding to the positioning method type i, is the signal interference coefficient corresponding to the positioning method type i;

[0026] The calculation formula for the spatial signal environment value is:

[0027] ; is the spatial signal environment value, is the total number of types of positioning methods in the current time window, , , and are all preset non-zero proportionality coefficients, is the preset exponential weight, and its value range is [0.5, 2]. Finally, the average value of all spatial signal environment values in the current time window is obtained to get the spatial signal environment index .

[0028] In a preferred embodiment, the acquisition logic of the signal path integrity index is as follows:

[0029] In the current time window, calculate the difference between the data reception time and the expected reception time corresponding to time t to obtain the signal delay value corresponding to time t; calculate the time ratio of signal loss within the sub-time period to which time t belongs to obtain the signal loss degree corresponding to time t; calculate the ratio of the deviation between two adjacent positioning points at time t and time t - 1 to the maximum allowable deviation to obtain the data consistency corresponding to time t;

[0030] The calculation formula of the signal path integrity index is:

[0031] ; i represents the number of the type of positioning method, is the total number of types of positioning methods in the current time window, represents the standard deviation of all signal delay values corresponding to the positioning method type i in the current time window, represents the standard deviation of all signal loss degrees corresponding to the positioning method type i in the current time window, represents the standard deviation of all data consistencies corresponding to the positioning method type i in the current time window, , are all preset influence coefficients, represents the signal path integrity index.

[0032] In a preferred embodiment, the machine learning model is a convolutional neural network model. The feature data groups, namely the spatial signal environment index and the signal path integrity index, are imported into the pre-trained convolutional neural network model together. According to the degree of influence of the positioning signal by the environment and the stability of the positioning data path, the convolutional neural network model classifies the monitoring situation of the dynamic tracking of the goods into low-quality tracking, medium-quality tracking, or high-quality tracking.

[0033] In a preferred embodiment, for the monitoring scenarios of high-quality tracking and low-quality tracking, fuzzy inference is used to optimize the current positioning method. The spatial signal environment index, the signal path integrity index, and the positioning method in the current time window are used as the input variables of fuzzy logic, and the optimized positioning method is used as the output variable of fuzzy logic. The input variables are fuzzified, converting the values of the input variables into fuzzy sets, and the output variable is fuzzified, converting the output variable into a fuzzy set. Fuzzy rules are formulated to describe the adaptability of each positioning method type under different combinations of data types. The fuzzified input variables are inferred through the fuzzy rules to obtain the optimized positioning method and apply it.

[0034] In a preferred embodiment, a satellite-precision positioning-based cargo dynamic tracking system includes:

[0035] A data acquisition module that, during the transportation of goods, collects the positioning data of the goods in real time to obtain a positioning data set;

[0036] A positioning accuracy analysis module that performs positioning accuracy analysis based on the positioning data and determines whether there are early signs of insufficient positioning accuracy during the transportation of goods through the positioning accuracy analysis results;

[0037] A dynamic tracking feature extraction module that, when there are early signs of insufficient positioning accuracy, extracts dynamic tracking features and then evaluates the monitoring situation of the cargo dynamic tracking to provide a feature data set for the subsequent classification of the current dynamic tracking quality;

[0038] A dynamic tracking quality classification module that imports the feature data set into a pre-trained machine learning model and classifies the monitoring situation of the cargo dynamic tracking as low-quality tracking, medium-quality tracking, or high-quality tracking according to the degree of influence of the positioning signal by the environment and the stability of the positioning data path;

[0039] A positioning method optimization module that, in the monitoring scenarios of high-quality tracking and low-quality tracking, optimizes the current positioning method to keep the cargo positioning accuracy in a standard state and uses the optimized positioning method for the dynamic tracking of the goods.

[0040] The technical effects and advantages of the present invention:

[0041] The present invention combines positioning technologies such as GNSS, Wi-Fi, and UWB to provide precise positioning services by leveraging their respective advantages in different environments. For high-quality tracking, sufficient accuracy can be achieved using a single or a few of these technologies, avoiding the use of redundant technologies and thus unnecessary errors and computational burdens. In complex environments (such as tunnels, underground parking lots, urban canyon effects, etc.), the system can ensure the quality of positioning signals and path stability by flexibly adjusting the positioning method, thereby improving the accuracy of cargo tracking.

[0042] The present invention can optimize computational resources and reduce energy consumption by reducing redundant positioning methods. When the system determines that the current positioning accuracy has met the requirements, it uses intelligent optimization methods to streamline the positioning method. For example, in the case of high-quality tracking, an efficient combination such as GNSS+INS is used instead of UWB and Wi-Fi, thus reducing unnecessary computational burdens and power consumption. Through this optimization, the system can not only maintain high-precision positioning but also significantly reduce energy consumption and extend the service life of the device. Intelligent adjustment of the positioning method: Through fuzzy inference, the system can dynamically adjust the positioning method so that each positioning technology is used when necessary, avoiding the system from simultaneously activating multiple positioning methods when unnecessary, thereby optimizing the computational burden and resource consumption.

[0043] The present invention improves the system response speed and real-time performance and reduces the computational burden. By reducing redundant technologies and optimizing the positioning method, the system can significantly reduce the computational burden and enhance the system's response speed. For high-quality tracking, the optimization of the positioning method can be achieved through real-time data adjustment, enabling the system to make optimal decisions in a shorter time and provide fast and efficient dynamic tracking services.

[0044] By combining the spatial signal environment index and the signal path integrity index, the present invention can monitor the changes in the environment where the goods are located in real time and automatically adjust the positioning strategy according to these changes. When the environmental signal changes, the system can quickly switch to the most suitable combination of positioning technologies, thus adapting to the fluctuations in environmental conditions and ensuring that the cargo tracking is always in the best state. Since the system can dynamically adjust the positioning technology according to environmental changes, it has stronger robustness against various interference sources (such as signal occlusion, electromagnetic interference, etc.) and can maintain stable and reliable positioning performance in complex and uncertain environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the accompanying drawings;

[0046] Figure 1 It is the schematic diagram of the method for dynamically tracking goods using satellite precise positioning in the present invention.

[0047] Figure 2This is the schematic diagram of the cargo dynamic tracking system using satellite precise positioning in the present invention. Detailed implementation manners

[0048] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0049] Refer to Figure 1 - Figure 2 The following embodiments are obtained:

[0050] Embodiment 1:

[0051] In the process of modern cargo transportation, especially in complex environments, positioning technologies are widely used in cargo tracking and management. Traditional cargo tracking systems mainly rely on single positioning technologies, such as GNSS (Global Navigation Satellite System) or Wi-Fi, etc. However, these single technologies often show problems of insufficient accuracy when facing signal interference, occlusion (such as tunnels, basements, urban canyon effect, etc.) and environmental changes. In addition, with the diversification of transportation environments and cargo types, the existing tracking technologies have the following problems: Fluctuation of positioning accuracy: In complex environments, positioning signals may be affected by multipath effects or signal occlusion, resulting in fluctuations in positioning accuracy and affecting the tracking of goods. Signal loss: Signal loss or instability leads to missing positioning data, reducing the reliability of the tracking system. Poor path stability: Even if multiple positioning methods are used, without an effective path optimization mechanism, large path deviations may still occur.

[0052] To solve the above problems and improve the accuracy and stability of the cargo dynamic tracking system, the present invention proposes a dynamic tracking optimization system based on multi-modal positioning. This system automatically adjusts the positioning method by comprehensively considering the characteristics of different positioning methods, environmental signal effects, and path stability, and optimizes the positioning method in different quality tracking situations. Specifically, the reasons for proposing the present invention include:

[0053] To address positioning problems in complex environments: Different positioning methods have different performances in different environments. For example, GNSS signals may be lost in tunnels or dense urban environments, while Wi-Fi and UWB perform better in these environments. Through the optimization of multi-modal positioning methods, positioning errors can be minimized to the greatest extent.

[0054] Improve the flexibility and reliability of the system: Dynamically adjust the positioning method based on environmental impact and path stability, enabling the system to adjust its working mode according to real-time positioning data, avoiding the problem that a single positioning method cannot handle complex situations.

[0055] Automated tracking optimization: Traditional cargo tracking systems often rely on manual intervention to adjust equipment or strategies. In contrast, this invention uses machine learning and fuzzy inference technologies to automatically optimize the positioning method, reduce manual intervention, and improve the efficiency and accuracy of the system.

[0056] This invention aims to improve the performance of the cargo dynamic tracking system in the following aspects:

[0057] Achieve the integration of multi-modal positioning methods: In scenarios such as tunnels, underground parking lots, urban high-rise dense areas (urban canyon effect), and inside warehouses, GNSS signals are easily blocked or affected by multipath interference, resulting in a decrease in positioning accuracy or even signal loss. By combining multi-modal fusion positioning technologies such as inertial navigation (INS), Wi-Fi positioning, Bluetooth beacons, and UWB (ultra-wideband), the tracking ability in complex environments is enhanced. Integrate multiple positioning technologies such as GNSS, INS, Wi-Fi, and UWB, leveraging their respective advantages to improve positioning accuracy and stability.

[0058] Provide dynamic tracking quality classification: Optimize the positioning method through machine learning models and fuzzy inference, enabling the system to automatically judge the tracking quality according to environmental changes and classify it as low-quality, medium-quality, or high-quality tracking.

[0059] Optimize the positioning method: Whether it is high-quality or low-quality tracking, the system can optimize the positioning method in real-time based on environmental signal quality and path stability to ensure that the cargo positioning accuracy remains at the standard level.

[0060] Achieve traceability and remote monitoring: The optimized positioning method can track the cargo in real-time and store the data in the cloud or on the blockchain to ensure the traceability and security of the data.

[0061] Optimization of low-quality tracking: When the monitoring result of the cargo dynamic tracking system is low-quality, it indicates that there are large errors in the positioning signal or the path is unstable, which may affect subsequent transportation management and cargo tracking. In this case, it is necessary to optimize the positioning method. For example, if the current single GNSS positioning is used and the signal is unstable or lost, the system needs to switch to a multi-modal positioning method (such as GNSS+INS+UWB) to reduce errors and improve stability.

[0062] In the case of high-quality tracking, the positioning accuracy usually meets the requirements. At this time, it is not necessary to enable multiple positioning methods simultaneously. Instead, the positioning methods should be streamlined to avoid unnecessary calculations and resource consumption. For example, if GNSS+INS already provides sufficiently accurate positioning, there is no need to enable UWB or Wi-Fi, as these technologies may bring additional computational burdens and energy consumption.

[0063] Improve energy efficiency: Multiple positioning methods (such as UWB, Wi-Fi, GNSS, etc.) may require high power and computational resources. If high-quality tracking has been completed through a single or a few technologies, it is very necessary to reduce redundant technologies to reduce energy consumption and extend the battery life of the device.

[0064] Reduce computational burden: Although multi-modal fusion positioning can improve accuracy, it will increase the computational burden, especially in real-time tracking systems. Optimizing the selection of positioning methods during high-quality tracking can reduce the computational burden while maintaining accuracy and improve the system response speed.

[0065] Based on this, the present invention proposes the following: A method for dynamically tracking goods using satellite precise positioning, including the following steps:

[0066] During the transportation of goods, the positioning data of the goods is collected in real time to obtain a positioning data set; Collecting the positioning data of the goods in real time is the basis for dynamic tracking. By collecting positioning information including data such as timestamps, position coordinates, and speeds, the position changes of the goods can be understood in real time to ensure accurate tracking of the transportation process.

[0067] Based on the positioning data, a positioning accuracy analysis is carried out, and whether there are early signs of insufficient positioning accuracy during the transportation of the goods is judged through the results of the positioning accuracy analysis; By analyzing the positioning data, the current positioning accuracy is evaluated to ensure the accuracy of the positioning information. The accuracy analysis helps to judge whether the positioning is interfered by environmental factors (such as signal loss, occlusion, etc.), and whether the system can provide reliable positioning results. If the accuracy analysis shows that there are large deviations or instabilities in the positioning data, by judging whether there are signs of insufficient accuracy, potential problems of the system can be warned in advance. Discovering positioning problems early helps to take measures to avoid tracking failures or accuracy degradation.

[0068] When there are early signs of insufficient positioning accuracy, dynamic tracking features are extracted, and then the monitoring situation of cargo dynamic tracking is evaluated to provide a feature data set for subsequent classification of the current dynamic tracking quality. The feature data set is imported into a pre-trained machine learning model. According to the degree to which the positioning signal is affected by the environment and the stability of the positioning data path, the monitoring situation of cargo dynamic tracking is classified as low-quality tracking, medium-quality tracking, or high-quality tracking; when it is determined that the positioning accuracy is insufficient, by extracting dynamic tracking features, the current tracking status is further analyzed. The extracted features provide data support for subsequent quality evaluation and optimization. By evaluating the extracted features, the quality of cargo dynamic tracking is understood. This evaluation helps to judge the current tracking status (low quality, poor stability, or high quality, strong stability) and provides a basis for subsequent optimization decisions. Based on the feature data set (such as positioning accuracy, path stability), the cargo dynamic tracking situation is classified as low quality, medium quality, or high quality. The classification results help the system identify and solve positioning problems and ensure the accuracy of cargo tracking. Through data classification by a machine learning model, the tracking quality can be automatically evaluated. The pre-trained model can judge the quality of cargo tracking based on historical data and patterns and give reasonable classification results.

[0069] For the monitoring situations of high-quality tracking and low-quality tracking, the current positioning method is optimized to keep the cargo positioning accuracy in a standard state, and the optimized positioning method is used for the dynamic tracking of the cargo. The tracking data is stored to achieve traceability and remote monitoring. According to the classification results, the low-quality or high-quality tracking situations are optimized to ensure the positioning accuracy. If low-quality tracking is found, the system will automatically switch to a more stable positioning method. If it is high-quality tracking, unnecessary positioning methods can be reduced to save resources. With the support of the optimized positioning method, the cargo tracking continues to ensure the tracking accuracy and stability. The optimized positioning method can avoid excessive resource consumption while maintaining high accuracy and improve the system efficiency. The tracking data of the cargo is stored in the cloud or blockchain to ensure the security, integrity, and traceability of the data. The stored tracking data can not only ensure real-time monitoring but also provide historical data support for subsequent cargo transportation analysis, problem tracing, and optimization.

[0070] Performing positioning accuracy analysis based on positioning data means that within the current time window, the average deviation between the current positioning data and the reference position inferred from the historical trajectory is calculated to measure the overall positioning accuracy and obtain the average offset error value; the standard deviation of the change rate of the cargo movement speed at adjacent time points within the current time window is calculated to obtain the speed jump value; the standard deviation of the time interval between two consecutive positioning data within the current time window is calculated to obtain the sampling jump value, and then the average offset error value, the speed jump value, and the sampling jump value are summarized together as the actual positioning vector.

[0071] By calculating the average deviation between the current positioning data and the reference position inferred from the historical trajectory, the overall accuracy of the positioning system can be evaluated. If the deviation is large, it indicates that the error of the current positioning system is high, which may affect the accurate tracking of goods. This value reflects the long-term stability of the system and helps to evaluate the effectiveness of the current positioning method.

[0072] Calculate the speed jump value: The standard deviation of the change rate of the goods movement speed at adjacent time points can reveal the degree of acceleration or deceleration of the goods during transportation. Excessive speed fluctuations may indicate signal problems, environmental factor interference, or sensor failures. The speed jump value helps to determine whether there are abnormal fluctuations in the positioning data, which in turn affects the stability of the positioning accuracy.

[0073] Calculate the sampling jump value: The standard deviation of the time interval between two consecutive positioning data can reflect the discontinuity in the sampling process. Excessive changes in the time interval may indicate discontinuous data collection, resulting in positioning errors. This value helps to determine whether the sampling process is stable and whether there are cases of data loss or too long sampling intervals.

[0074] Aggregating the above three indicators (average offset error, speed jump value, sampling jump value) into an actual positioning vector can more comprehensively reflect the current state of the goods positioning system. The actual positioning vector provides a multi-dimensional comprehensive evaluation, helping the system to comprehensively judge the quality of the positioning accuracy.

[0075] Judging whether there are early signs of insufficient positioning accuracy during the goods transportation process based on the positioning accuracy analysis results refers to: comparing the actual positioning vector with a preset standard vector, calculating the Euclidean distance between the actual positioning vector and the preset standard vector, comparing the Euclidean distance with a preset warning threshold. If the Euclidean distance is greater than the preset warning threshold, it is determined that there are early signs of insufficient positioning accuracy; if the Euclidean distance is less than or equal to the preset warning threshold, it is determined that there are no early signs of insufficient positioning accuracy.

[0076] By calculating the Euclidean distance between the actual positioning vector and the preset standard vector, the system can quantify the difference between the current positioning accuracy and the standard. The Euclidean distance provides a mathematical measurement method, making the judgment of positioning accuracy more objective and intuitive. The preset warning threshold defines the maximum deviation range tolerated by the system. When the Euclidean distance exceeds the warning threshold, the system will consider the positioning accuracy insufficient, and at this time, warnings and subsequent measures (such as switching the positioning method, enhancing the positioning signal, etc.) are required. If the Euclidean distance is less than or equal to the warning threshold, it indicates that the positioning accuracy is good and the system does not need to intervene. By comparing the Euclidean distance between the actual positioning vector and the standard vector in real time, the system can timely identify whether there are early signs of insufficient positioning accuracy, which is conducive to taking corresponding optimization measures in advance, such as adjusting the positioning method, to ensure the accuracy and reliability of cargo positioning.

[0077] The extraction of cargo dynamic tracking features includes positioning environment information and positioning stability information. Then, a spatial signal environment index is generated based on the positioning environment information, and the spatial signal environment index is used to evaluate the influence degree of the environment where the cargo is located during transportation on the positioning signal. A signal path integrity index is generated based on the positioning stability information, and the signal path integrity index is used to evaluate the stability of the positioning data path during the transportation of the cargo. The spatial signal environment index and the signal path integrity index are jointly used to reflect the monitoring situation of cargo dynamic tracking.

[0078] The extraction of cargo dynamic tracking features is a core step in the entire cargo tracking system. Its significance lies in evaluating the positioning signal quality and path stability of the cargo during transportation through comprehensive analysis of the positioning environment information and the positioning stability information. This process provides key information for subsequent tracking quality evaluation and positioning method optimization.

[0079] The positioning environment information refers to the influence of the environment where the cargo is located during transportation on the positioning signal, such as signal strength, available signal quality, interference sources, etc. These information help the system evaluate how the external environment affects the positioning accuracy. For example, in complex environments such as tunnels and urban canyons, the signal may be blocked, affecting the positioning accuracy.

[0080] The positioning stability information refers to the data consistency during the cargo dynamic tracking process, including the stability of the positioning signal, path stability, etc. The stability information evaluates whether the path is continuous and whether the signal is stable, which helps to identify unstable positioning data and thus improve the tracking quality.

[0081] The Spatial Signal Environment Index (SSEI) is a key metric generated based on positioning environment information, used to quantify the impact of the environment on the quality of positioning signals. Evaluating the degree of signal affected by the environment: By integrating multi-dimensional environmental information such as signal strength, available signal quality, interference situation, and signal drift, SSEI provides an intuitive quantification result for measuring the stability and quality of signals under environmental conditions. Decision support: The higher the value of SSEI, the smaller the impact of the environment where the goods are located on the positioning signal, and the higher the positioning accuracy. Conversely, it indicates that environmental factors have a greater impact on the positioning signal, and it may be necessary to adjust the positioning method or optimize the sensor configuration. Optimizing the positioning scheme: According to SSEI, the system can dynamically adjust the positioning method, using multi-modal positioning (such as GNSS+INS+UWB) in adverse environments, while in good environments, the system can reduce the use of redundant positioning technologies to save computing resources.

[0082] The Signal Path Integrity Index (SPII) is an important metric generated based on positioning stability information, used to evaluate the stability of the positioning data path. Evaluating the stability of the data path: SPII reflects the continuity and consistency of data during the positioning process. A higher SPII value indicates better path stability of the positioning data, without data loss or significant drift, and the positioning system can maintain a high level of accuracy. Detecting abnormal fluctuations: If the SPII value is low, it means there are large fluctuations, data loss, or inconsistencies in the path, and the system needs to take timely measures (such as switching the positioning method, compensating for missing data, etc.) to maintain the tracking quality. Ensuring the tracking quality: The analysis result of SPII directly affects the system's judgment of the positioning signal quality, and thus affects the subsequent selection of positioning methods and optimization strategies. A stable path can ensure the efficiency and accuracy of goods tracking.

[0083] The Spatial Signal Environment Index (SSEI) and the Signal Path Integrity Index (SPII) jointly reflect the monitoring situation of goods dynamic tracking: Comprehensively evaluating the tracking quality: SSEI mainly evaluates the impact of environmental factors on the positioning signal, while SPII evaluates the stability and continuity of the positioning signal. Combining the two can comprehensively evaluate the quality of goods tracking. SSEI provides a judgment on the environment, and SPII provides a judgment on data stability. The combination of the two can help the system better perform quality classification and optimization decisions. By evaluating SSEI and SPII, the system can more accurately classify the quality of the current tracking (low quality, medium quality, high quality) and take corresponding optimization measures based on these metrics. For example, if both SSEI and SPII indicate poor signal quality, the system will preferentially enable high-precision positioning methods, such as combining multiple positioning technologies (GNSS+INS+UWB) to improve tracking accuracy.

[0084] The acquisition logic of the Spatial Signal Environment Index is as follows:

[0085] During the process of goods transportation, the following four types of environmental information are collected based on the currently adopted positioning method: received signal strength coefficient, available signal quality coefficient, signal interference coefficient, and signal drift coefficient. To ensure the unified processing of data from different sensors, the received signal strength coefficient, available signal quality coefficient, signal interference coefficient, and signal drift coefficient are respectively normalized:

[0086] For different positioning methods (such as GNSS, Wi-Fi, UWB, etc.), the measurement ranges, units, and accuracies of the data collected in different environments may vary greatly. Through normalization, various data can be converted into a unified scale, facilitating comprehensive evaluation and comparison. The measurements of different signals (for example, the units of signal strength may be different) may affect the final results. Normalization can remove these unit differences, enabling various environmental information to be calculated under the same standard. After normalization, various coefficients will contribute proportionally to the final spatial signal environment index, making the influence of each factor on the final index under the same standard.

[0087] The minimum-maximum normalization processing method is adopted for the received signal strength coefficient:

[0088] ; is the normalization result of the received signal strength coefficient corresponding to the positioning method type i, is the received signal strength coefficient corresponding to the positioning method type i, 、 are the minimum and maximum values of the standard received signal strength coefficient corresponding to the positioning method type i;

[0089] The proportional normalization processing method is adopted for the available signal quality coefficient:

[0090] ; is the normalization result of the available signal quality coefficient corresponding to the positioning method type i, is the available signal quality coefficient corresponding to the positioning method type i, is the standard available signal quality coefficient value corresponding to the positioning method type i;

[0091] The exponential decay normalization processing method is adopted for the signal interference coefficient:

[0092] ; is the normalization result of the signal interference coefficient corresponding to the positioning method type i, is the signal interference coefficient corresponding to the positioning method type i;

[0093] The exponential decay normalization processing method is adopted for the signal drift coefficient:

[0094] ; is the normalized result of the signal drift coefficient corresponding to the positioning method type i, is the signal interference coefficient corresponding to the positioning method type i;

[0095] The calculation formula for the spatial signal environment value is:

[0096] ; is the spatial signal environment value, is the total number of types of positioning methods in the current time window, , , and are all preset non-zero proportionality coefficients, is the preset exponential weight, with a value range of [0.5, 2]. Finally, the average value of all spatial signal environment values in the current time window is obtained to get the spatial signal environment index . Different weight coefficients are assigned to different environmental information (received signal strength, available signal quality, interference, and drift), and the contribution degree of each factor to the environment can be adjusted according to the actual situation. For example, signal interference may have a greater impact on positioning accuracy in some environments, while received signal strength may have a greater impact in areas with weak signals. Introducing an exponential weight (such as a value range from 0.5 to 2) further enhances the non-linearity of the model, making the impact of environmental factors on signal quality more in line with the sensitivity in actual applications. For example, the changes in signal strength and interference may not be linear, and using an exponential function can better capture such changes.

[0097] The received signal strength coefficient is used to measure the strength of the signal received by each positioning method. In GNSS, Wi-Fi, and UWB, the calculation methods of signal strength are slightly different. The following is an example of the calculation method for a positioning method: GNSS (Global Navigation Satellite System): In the GNSS system, the received satellite signal strength is usually represented by the signal-to-noise ratio (SNR) or received signal strength indicator (RSSI). By measuring the ratio of the signal strength from the satellite to the receiver and the noise signal (signal-to-noise ratio), the signal strength can be obtained. Calculation of the received signal strength coefficient: Signal strength coefficient = (Received signal strength) / (Standard signal strength of this positioning method).

[0098] Wi-Fi (Wireless Local Area Network): In a Wi-Fi system, the received signal strength is usually evaluated by the Received Signal Strength Indicator (RSSI), and the RSSI value is usually proportional to the signal strength received by the device. Calculation of the received signal strength coefficient: Signal strength coefficient = (Signal strength received by Wi-Fi) / (Standard signal strength of this Wi-Fi positioning method)

[0099] UWB (Ultra-Wideband): In a UWB positioning system, the measurement of signal strength is usually based on the intensity of the received electromagnetic wave. Calculation of the received signal strength coefficient: Signal strength coefficient = (Signal strength of received UWB signal) / (Standard UWB signal strength).

[0100] The signal interference coefficient is used to evaluate the degree of external interference suffered by the signal during transmission. Under different positioning methods, the sources of interference may be different, such as multipath effects, electromagnetic interference, etc.

[0101] In a GNSS system, signal interference is usually related to multipath effects (signal reflection) and electromagnetic interference. Calculation of the interference coefficient: Signal interference coefficient = (Interference intensity of the received signal) / (Standard interference intensity). The Wi-Fi signal interference coefficient is usually affected by the surrounding environment, such as electromagnetic wave interference and the presence of physical obstacles such as walls. Calculation of the interference coefficient: Signal interference coefficient = (Interference of the Wi-Fi received signal) / (Standard Wi-Fi interference). UWB signal interference usually comes from interference from neighboring wireless devices or signal reflection. Calculation of the interference coefficient: Signal interference coefficient = (Interference intensity of the UWB signal) / (Standard UWB interference).

[0102] The signal drift coefficient measures the degree of signal drift during transmission and is usually used in calculations in an Inertial Navigation System (INS), affecting positioning accuracy. In a GNSS system, drift may occur due to changes in satellite positions or device errors. Calculation of the drift coefficient: Signal drift coefficient = (Change in signal position) / (Standard drift value). The drift coefficient of a Wi-Fi system may be related to factors such as device errors and moving speed. Calculation of the drift coefficient: Signal drift coefficient = (Change in Wi-Fi positioning data) / (Standard Wi-Fi drift value). The drift coefficient of a UWB positioning system is related to sensor errors and signal propagation characteristics. Calculation of the drift coefficient: Signal drift coefficient = (Drift value of the UWB signal) / (Standard UWB drift value)

[0103] The available signal quality factor reflects the quality of the signal available in the current environment, which is usually determined by factors such as signal stability, signal interference, and signal reliability. In the GNSS system, the signal quality is usually related to the number of satellites received, signal stability, and the geometric distribution of satellites (such as the DOP value). Calculation of the signal quality factor: Available signal quality factor = (Number of satellites received) / (Maximum standard number of satellites). The signal quality of the Wi-Fi system is usually evaluated by signal stability and link quality. The link quality is calculated based on the signal strength of the Wi-Fi access point, interference between access points, and link stability. Calculation of the signal quality factor: Signal quality factor = (Wi-Fi connection stability) / (Standard Wi-Fi stability). The quality of the UWB positioning signal is usually related to signal integrity and reliability. The quality of the UWB signal is affected by the transmission distance between devices, propagation obstacles, and interference sources. Calculation of the signal quality factor: Signal quality factor = (Effective reception time of the UWB signal) / (Standard UWB signal reception time).

[0104] It should be noted that since it is difficult to calculate instantaneous values for some parameter calculations, in the actual calculation process, the current time window can be divided into multiple sub-time windows of the same time length, calculate the parameter values of the corresponding sub-time windows, and replace the instantaneous values corresponding to the time points to make all calculation formulas valid. For example: When calculating the signal drift coefficient, it is usually necessary to estimate based on the continuity of the device (such as the integration error of inertial sensors) or the long-term offset of the signal. If the instantaneous drift coefficient is directly calculated, it is easily affected by noise and may not reflect the actual long-term trend. Therefore, using a time window for smoothing processing and calculating the drift coefficient within each sub-time window can effectively reduce these errors. Specific steps: Divide the time window: Divide the current time window into multiple equal-length sub-time windows, for example, each sub-time window is 1 second or 5 seconds. Calculate the drift coefficient of each sub-time window: Calculate the signal drift within each sub-time window. The signal drift can be integrated based on the data output by the inertial sensor (INS) or calculated according to the offset of the positioning data. For example, the drift coefficient within a certain sub-time window can be obtained by averaging or weighting the signal deviation within that time period. Replace the instantaneous value: Replace the instantaneous drift value at that moment with the average drift coefficient calculated for each sub-time window. In this way, the drift coefficient at each moment can more accurately reflect the long-term drift trend of the device, rather than relying solely on instantaneous data. And in fact, the drift coefficients of all sub-time windows can also be smoothed (such as by the weighted average method) to finally obtain a stable drift coefficient value. This process effectively reduces the errors that may be brought by instantaneous drift values and improves the positioning accuracy.

[0105] The acquisition logic of the signal path integrity index is as follows:

[0106] In the current time window, calculate the difference between the data reception time and the expected reception time corresponding to time t respectively to obtain the signal delay value corresponding to time t; calculate the time ratio of signal loss within the sub-time period to which time t belongs to obtain the signal loss degree corresponding to time t; calculate the ratio of the deviation between two adjacent positioning points at time t and time t-1 to the maximum allowable deviation to obtain the data consistency corresponding to time t.

[0107] The calculation formula of the signal path integrity index is:

[0108] ; i represents the serial number of the positioning method type, is the total number of types of positioning methods in the current time window, represents the standard deviation of all signal delay values corresponding to the positioning method type i in the current time window, represents the standard deviation of all signal loss degrees corresponding to the positioning method type i in the current time window, represents the standard deviation of all data consistencies corresponding to the positioning method type i in the current time window, 、 are both preset influence coefficients, represents the signal path integrity index. Signal path integrity analysis: Signal delay: Measures the time delay from signal transmission to reception, reflecting the time difference in signal transmission. When the delay is too large, it may lead to a decrease in the timeliness of positioning data, thus affecting the accuracy of the entire positioning system. Signal loss: In practical applications, signals may be lost or interrupted due to various factors (such as occlusion or interference). The lost signals will affect the stability and continuity of positioning. Data consistency: Reflects the ratio of the deviation between adjacent positioning data points to the maximum allowable deviation. If the data deviation is too large, it indicates that there is a large instability in the positioning path.

[0109] 、 are both preset influence coefficients. The design of the weights reflects the relative importance of different factors on the positioning accuracy and signal path stability. For example, signal loss and delay may have a greater impact on the positioning accuracy, while data consistency affects the continuity and stability of the path. The standardized calculations of signal delay, signal loss, and data consistency respectively convert these original indicators into a unified standard range, thus avoiding calculation errors caused by different dimensions between different indicators and effectively quantifying the stability of the signal path.

[0110] The exponential part in the formula is the key to balancing the signal influencing factors. The design of the exponential function enhances the non-linear response of the signal path stability to the signal quality. It enables certain factors (such as large signal delays or signal losses) to have a more significant impact on the final exponent, avoiding the over-amplification of minor interferences. By summing up and normalizing the weighted terms, the reasonable weight of each factor is ensured, and finally a single value reflecting the signal path stability is obtained. In this formula, the impacts of various parts of the signal (such as delay, loss, data consistency) are exponentiated after weighted summation, reflecting the non-linear impact. That is, the greater the factors such as signal delay and signal loss, the lower the final SPII value, indicating a higher degree of path instability. Through the comprehensive evaluation of different factors, the SPII index can accurately reflect the signal path stability and reliability in the process of cargo positioning.

[0111] The machine learning model is a convolutional neural network model. The feature data groups, namely the spatial signal environment index and the signal path integrity index, are imported into the pre-trained convolutional neural network model together. According to the degree of influence of the positioning signal by the environment and the stability of the positioning data path, the convolutional neural network model classifies the monitoring situation of cargo dynamic tracking into low-quality tracking, medium-quality tracking or high-quality tracking.

[0112] The structural principle of the convolutional neural network: Input layer: The input data are the spatial signal environment index and the signal path integrity index, and these two feature data groups are the inputs of the convolutional neural network. The spatial signal environment index reflects the impact of the environment where the cargo is located on the positioning signal, and the signal path integrity index evaluates the stability of the positioning data. These two indices can provide sufficient information to distinguish different tracking qualities. Convolutional layer: The convolutional layer is responsible for extracting useful features from the input data. Through the sliding window operation, the convolutional neural network can capture local features in the data, such as signal quality fluctuations and path stability changes. These local features help the convolutional neural network identify and classify different types of cargo tracking qualities, especially for those complex signal patterns that are difficult to simply describe by traditional methods. Pooling layer: The pooling layer is usually located after the convolutional layer. Its role is to reduce the dimensionality of the features output by the convolutional layer, reduce the data volume and computational amount, while retaining important feature information. This helps to improve the generalization ability of the model and avoid the overfitting problem. Fully connected layer: After the convolutional layer and the pooling layer extract sufficient features, the network passes these features to the fully connected layer. The fully connected layer combines the various feature information through multiple neurons and finally generates an output for classification. Output layer: The task of the output layer is to classify the tracking quality of the cargo into low-quality tracking, medium-quality tracking or high-quality tracking according to the input feature data. By learning the feature combinations of different signal environments and path stabilities, the network can predict the current state based on historical data and give the classification of the tracking quality.

[0113] For the monitoring scenarios of high-quality tracking and low-quality tracking, fuzzy inference is used to optimize the current positioning method. The spatial signal environment index, the signal path integrity index, and the positioning method under the current time window are taken as the input variables of fuzzy logic, and the optimized positioning method is taken as the output variable of fuzzy logic. The input variables are fuzzified, converting the values of the input variables into fuzzy sets, and the output variable is fuzzified, converting the output variable into a fuzzy set. Fuzzy rules are formulated to describe the adaptation degree of each positioning method type under different combinations of data types. The fuzzified input variables are inferred through the fuzzy rules to obtain the optimized positioning method and apply it.

[0114] Fuzzification of input variables: The spatial signal environment index and the signal path integrity index are important indicators affecting positioning quality. Since these indices are usually continuous variables, we need to convert them into fuzzy sets, such as "low", "medium", and "high", so that the system can reason according to the fuzzy rule base.

[0115] Definition of fuzzy rules: Fuzzy rules are formulated based on system experience or data analysis to describe decisions under different input situations. For example: If the spatial signal environment index is "high" and the signal path integrity index is "high", then select the "standard positioning method". If the spatial signal environment index is "low" and the signal path integrity index is "low", then select "optimize the current positioning method". If the spatial signal environment index is "medium" and the signal path integrity index is "medium", then select "maintain the current positioning method".

[0116] Derive the output based on fuzzy rules and finally obtain the decision through defuzzification. In the process of fuzzy inference, existing techniques such as trigonometric functions and max-min operations are often used to complete specific operations. For example, trigonometric functions are mainly used in fuzzy inference to handle the non-linear relationship between input variables and output variables. In some cases, the relationship between input variables may not be linear, and trigonometric functions provide a way to capture these changes. In fuzzy inference, the max-min operation is an important technique for synthesizing the inference results of multiple fuzzy rules. The inference process in a fuzzy logic system usually involves multiple rules, and the output of each rule is a fuzzy set. To synthesize these multiple fuzzy outputs into the final decision result, the max-min operation is usually used. Input variables (such as the spatial signal environment index and the signal path integrity index) are converted into fuzzy sets, and the max-min operation can help determine the most appropriate membership degree values during fuzzification. The final output fuzzy set (such as the optimized positioning method) needs to be converted into a specific operation through the defuzzification process. The max-min method is usually used for the aggregation of multiple fuzzy rule outputs to ensure the most appropriate output result.

[0117] Suppose the system is tracking the location of a batch of goods, and currently uses multi-modal positioning (such as GNSS+INS+UWB). At this time, the system evaluates and finds that the spatial signal environment index is "high" and the signal path integrity index is "high". According to the optimization principle, the use of redundant technologies can be reduced through the following steps: Current positioning method: GNSS+INS+UWB: GNSS provides high-precision position data; INS (Inertial Navigation System) provides supplementation, especially in the case of unstable signals, to improve positioning accuracy; UWB (Ultra-Wideband) is used for more accurate short-distance positioning, especially indoors or in occluded areas, but the accuracy improvement it brings is no longer obvious at this time. Current computational burden: Using three technologies to work simultaneously will increase the computational burden, especially when processing real-time data, and the system needs to synchronously process data from multiple sensors. Optimized positioning method: GNSS+INS, UWB can be turned off to reduce unnecessary calculations and resource consumption. This not only improves the system's response speed but also reduces battery consumption and extends the device's working time.

[0118] Embodiment 2: A dynamic tracking system for goods using satellite precise positioning, comprising:

[0119] A data acquisition module that, during the transportation of goods, collects the positioning data of the goods in real time to obtain a positioning data set;

[0120] A positioning accuracy analysis module that performs positioning accuracy analysis based on the positioning data and determines whether there are early signs of insufficient positioning accuracy during the transportation of goods through the positioning accuracy analysis results;

[0121] A dynamic tracking feature extraction module that, when there are early signs of insufficient positioning accuracy, extracts dynamic tracking features, and then evaluates the monitoring situation of the goods' dynamic tracking to provide a feature data set for the subsequent classification of the current dynamic tracking quality;

[0122] A dynamic tracking quality classification module that imports the feature data set into a pre-trained machine learning model and divides the monitoring situation of the goods' dynamic tracking into low-quality tracking, medium-quality tracking, or high-quality tracking according to the degree of influence of the positioning signal by the environment and the stability of the positioning data path;

[0123] A positioning method optimization module that, in the monitoring situations of high-quality tracking and low-quality tracking, optimizes the current positioning method to keep the goods' positioning accuracy in a standard state and uses the optimized positioning method for the dynamic tracking of the goods.

[0124] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula that is closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0125] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not imply the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0126] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present application.

[0127] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0128] As described above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A cargo dynamic tracking method using satellite precise positioning, characterized in that: The following steps are involved: During the transportation of goods, the positioning data of the goods is collected in real time to obtain a positioning data set; Perform positioning accuracy analysis based on positioning data, and use the positioning accuracy analysis results to determine whether there are early signs of insufficient positioning accuracy during the cargo transportation process; When there are early signs of insufficient positioning accuracy, dynamic tracking feature extraction is performed, and then the monitoring of cargo dynamic tracking is evaluated to provide a feature data set for subsequent current dynamic tracking quality classification. The feature data set is imported into a pre-trained machine learning model, and the monitoring of cargo dynamic tracking is classified into low-quality tracking, medium-quality tracking, or high-quality tracking according to the degree to which the positioning signal is affected by the environment and the stability of the positioning data path; For the monitoring of high-quality tracking and low-quality tracking, the current positioning method is optimized to keep the cargo positioning accuracy at the standard state, and the optimized positioning method is used for dynamic tracking of cargo; The cargo dynamic tracking feature extraction includes positioning environment information and positioning stability information. Then, a spatial signal environment index is generated based on the positioning environment information. The spatial signal environment index is used to evaluate the degree of influence of the environment in which the cargo is located during transportation on the positioning signal. A signal path integrity index is generated based on the positioning stability information. The signal path integrity index is used to evaluate the stability of the positioning data path of the cargo during transportation. The spatial signal environment index and the signal path integrity index are used together to reflect the monitoring status of cargo dynamic tracking. The logic for obtaining the spatial signal environment index is: During the cargo transportation process, the following four types of environmental information are collected based on the current positioning method: received signal strength coefficient, available signal quality coefficient, signal interference coefficient and signal drift coefficient. In order to ensure that the data of different sensors can be processed uniformly, the received signal strength coefficient, available signal quality coefficient, signal interference coefficient and signal drift coefficient are normalized respectively: The minimum-maximum normalization method is used for the received signal strength coefficient: ; is the normalized result of the received signal strength coefficient corresponding to the positioning mode type i, is the received signal strength coefficient corresponding to positioning mode type i, , is the minimum and maximum values ​​of the standard received signal strength coefficient corresponding to the positioning mode type i; The available signal quality coefficients are normalized using a proportional approach: ; is the normalized result of the available signal quality coefficient corresponding to the positioning mode type i, is the available signal quality coefficient corresponding to positioning mode type i, is the standard available signal quality coefficient value corresponding to positioning mode type i; The signal interference coefficient is processed by exponential decay normalization: ; is the normalized result of the signal interference coefficient corresponding to the positioning mode type i, is the signal interference coefficient corresponding to the positioning mode type i; The signal drift coefficient is normalized using exponential decay: ; is the normalized result of the signal drift coefficient corresponding to the positioning mode type i, is the signal interference coefficient corresponding to the positioning mode type i; The calculation formula of spatial signal environment value is: ; is the spatial signal environment value, is the total number of positioning methods in the current time window, , , and are all preset non-zero scale factors, is the preset exponential weight, with a value range of [0.5, 2]. Finally, all spatial signal environment values ​​in the current time window are averaged to obtain the spatial signal environment index. ; The logic for obtaining the signal path integrity index is: In the current time window, the difference between the data receiving time corresponding to time t and the expected receiving time is calculated to obtain the signal delay value corresponding to time t; the time proportion of signal loss in the sub-time period to which time t belongs is calculated to obtain the signal loss degree corresponding to time t; Calculate the ratio of the deviation between two adjacent positioning points at time t and time t-1 to the maximum allowable deviation to obtain the data consistency corresponding to time t; The signal path integrity index is calculated as: ； i represents the number of the positioning method type. is the total number of positioning methods in the current time window, It represents the standard deviation of all signal delay values ​​corresponding to positioning mode type i in the current time window. It represents the standard deviation of all signal loss corresponding to positioning mode type i in the current time window. It represents the standard deviation of the consistency of all data corresponding to the positioning method type i in the current time window. , These are the preset influence coefficients. Represents the signal path integrity index.

2. The cargo dynamic tracking method using satellite precise positioning according to claim 1 is characterized in that: Positioning accuracy analysis based on positioning data refers to: In the current time window, the average deviation between the current positioning data and the reference position inferred based on the historical trajectory is calculated to measure the overall positioning accuracy and obtain the average offset error value; the standard deviation of the rate of change of the cargo movement speed at adjacent time points in the current time window is calculated to obtain the speed jump value; the standard deviation of the time interval between two consecutive positioning data in the current time window is calculated to obtain the sampling jump value, and then the average offset error value, speed jump value and sampling jump value are summarized together as the actual positioning vector.

3. The cargo dynamic tracking method using satellite precise positioning according to claim 2 is characterized in that: The early signs of insufficient positioning accuracy during cargo transportation can be determined through the positioning accuracy analysis results: The actual positioning vector is compared with the preset standard vector, the Euclidean distance between the actual positioning vector and the preset standard vector is calculated, and the Euclidean distance is compared with the preset warning threshold. If the Euclidean distance is greater than the preset warning threshold, it is determined that there are early signs of insufficient positioning accuracy. If the Euclidean distance is less than or equal to the preset warning threshold, it is determined that there are no early signs of insufficient positioning accuracy.

4. The cargo dynamic tracking method using satellite precise positioning according to claim 3 is characterized in that: The machine learning model is a convolutional neural network model. The feature data set, namely the spatial signal environment index and the signal path integrity index, is imported into the pre-trained convolutional neural network model. According to the degree to which the positioning signal is affected by the environment and the stability of the positioning data path, the convolutional neural network model divides the monitoring status of cargo dynamic tracking into low-quality tracking, medium-quality tracking or high-quality tracking.

5. The cargo dynamic tracking method using satellite precise positioning according to claim 4 is characterized in that: For the monitoring of high-quality tracking and low-quality tracking, fuzzy reasoning is used to optimize the current positioning method. The spatial signal environment index, signal path integrity index, and positioning method under the current time window are used as input variables of fuzzy logic. The optimized positioning method is used as the output variable of fuzzy logic. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the adaptability of each positioning method under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the optimized positioning method and apply it.

6. A cargo dynamic tracking system using satellite precise positioning, used to implement a cargo dynamic tracking method using satellite precise positioning as claimed in any one of claims 1 to 5, characterized in that: include: The data collection module collects the positioning data of the goods in real time during the transportation of the goods to obtain the positioning data set; The positioning accuracy analysis module performs positioning accuracy analysis based on positioning data and determines whether there are early signs of insufficient positioning accuracy during cargo transportation through the positioning accuracy analysis results; The dynamic tracking feature extraction module performs dynamic tracking feature extraction when there are early signs of insufficient positioning accuracy, and then evaluates the monitoring status of cargo dynamic tracking to provide a feature data set for subsequent current dynamic tracking quality classification; The dynamic tracking quality classification module imports the feature data group into the pre-trained machine learning model, and classifies the monitoring status of cargo dynamic tracking into low-quality tracking, medium-quality tracking or high-quality tracking according to the degree to which the positioning signal is affected by the environment and the stability of the positioning data path; The positioning method optimization module optimizes the current positioning method under the monitoring conditions of high-quality tracking and low-quality tracking, so that the cargo positioning accuracy remains at the standard state, and uses the optimized positioning method to dynamically track the cargo.

Images