Vehicle load real-time monitoring method

Abstract

The invention discloses a vehicle load real-time monitoring method. The method comprises the following steps: S1, acquiring load data of each part of a vehicle in real time; s2, preprocessing the load data transmitted to the data processing unit; s3, calculating the actual load of each monitoring position by using a weighted least square method, and calculating the total load information of the vehicle based on the actual load; s4, storing the encrypted total load information of the vehicle into a block chain network by using a block chain technology, and generating encrypted total load data; s5, transmitting the encrypted total load data stored on the block chain to a remote monitoring platform; s6, performing real-time analysis based on the encrypted total load data stored in the block chain, and when the encrypted total load data exceeds a preset threshold value, automatically triggering an alarm mechanism; and S7, regularly auditing the encrypted total load data stored in the block chain. The method has the advantages of being high in data security, high in real-time performance, high in data consistency and optimized in precision.

Description

Technical Field

[0001] The present invention relates to the technical field of real-time monitoring, and particularly to a method for real-time monitoring of vehicle load. Background Art

[0002] With the continuous development of intelligent transportation and automation technologies, vehicle load monitoring has become a key link in logistics management and traffic safety. Especially in highway and urban transportation systems, real-time monitoring of vehicle load information is of great significance for preventing overloading, ensuring road safety, and improving transportation efficiency. To achieve effective vehicle load monitoring, existing technologies mainly rely on static load sensors and a central processing system to obtain the total vehicle load data and perform certain processing and analysis.

[0003] However, there are some obvious deficiencies in existing vehicle load monitoring technologies, which are mainly manifested in the following aspects:

[0004] 1. Poor data transmission security: Existing methods usually rely on a centralized server to store and transmit vehicle load data, which makes the data vulnerable to hacker attacks or tampering during transmission. Especially in the transportation and logistics processes involving multiple parties, how to ensure the authenticity and integrity of the data has become an urgent problem to be solved.

[0005] 2. Insufficient real-time data processing ability: Most current load monitoring systems rely on single-sensor data collection and analysis methods, and fail to effectively fuse and optimize data from different monitoring locations. Therefore, in some complex application scenarios, the real-time processing ability and prediction accuracy of the system are limited, and potential problems such as uneven load distribution cannot be detected in a timely manner.

[0006] 3. Data consistency and auditing issues: In traditional methods, vehicle load data is often stored in local databases, lacking effective auditing and verification mechanisms. This may result in data tampering, loss, or inconsistency during data storage or transmission, seriously affecting the credibility of the data.

[0007] 4. Lack of multi-dimensional data analysis: Existing load monitoring systems generally only analyze simple sensor data, ignoring multi-dimensional load analysis and optimization. The lack of comprehensive fusion and prediction of data from multiple monitoring points leads to insufficient accuracy and reliability of the monitoring results.

[0008] Therefore, how to provide a method for real-time monitoring of vehicle load is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0009] An object of the present invention is to propose a real-time vehicle load monitoring method based on blockchain technology and weighted Gaussian process regression algorithm. By combining blockchain encrypted storage, weighted Gaussian process regression algorithm to optimize load distribution, hash algorithm to verify data consistency, and smart contract for data auditing, the present invention realizes real-time monitoring of vehicle load information, secure data storage and multi-dimensional analysis, and has the advantages of high data security, strong real-time performance, high data consistency and precision optimization.

[0010] A real-time vehicle load monitoring method according to an embodiment of the present invention includes the following steps:

[0011] S1. Deploy a distributed fiber optic sensor array at multiple load monitoring positions of the vehicle, collect the load data of each part of the vehicle in real time, and transmit the load data to the data processing unit inside the vehicle through the optical fiber;

[0012] S2. Preprocess the load data transmitted to the data processing unit, remove noise and perform preliminary correction to obtain the preprocessed load data;

[0013] S3. Based on the preprocessed load data, use the weighted least squares method to calculate the actual load of each monitoring position, and deduce the total vehicle load information of the vehicle based on the actual load;

[0014] S4. After encrypting the calculated total vehicle load information, use blockchain technology to store the encrypted total vehicle load information in the blockchain network to generate encrypted total load data;

[0015] S5. Build a remote monitoring platform, and transmit the encrypted total load data stored on the blockchain to the remote monitoring platform through a wireless communication network;

[0016] S6. On the remote monitoring platform, perform real-time analysis based on the encrypted total load data stored in the blockchain. When the encrypted total load data exceeds a preset threshold, automatically trigger an alarm mechanism;

[0017] S7. Regularly audit the encrypted total load data stored in the blockchain.

[0018] Optionally, the S2 specifically includes:

[0019] S21. Perform preliminary screening on the load data transmitted to the data processing unit to remove instantaneous interference signals during the acquisition process;

[0020] S22. Denoise the preliminarily screened load data, adopt a multi-point weighted Kalman filtering algorithm, and dynamically adjust the Kalman gain matrix in combination with the spatial position and measurement accuracy of each sensor : ; ;

[0021] Among them, is the load calculation parameter at the th iteration, is the estimated value of the load data at the previous update, is the Kalman gain matrix weighted based on the sensor positions, which adjusts the noise effects of different sensors, is the th load data measurement value, is the observation matrix, is the weight of the th sensor, is the covariance matrix of the th sensor at the previous moment, is the covariance of the measurement noise, represents the number of sensors participating in the filtering process;

[0022] S23. Based on the load data after Kalman filtering, data correction is performed. Based on the relationship between the sensor positions and the vehicle load distribution, using the known load standard, the load data measured by each sensor is adjusted;

[0023] S24. The corrected load data is standardized, and the load data of different sensors is converted into a unified dimension and magnitude, so that the output data of each sensor is comparable, and the deviations of different sensors are eliminated in the data processing;

[0024] S25. A method based on adaptive wavelet packet transform is used to perform time-domain smoothing on the load data after noise removal, filtering, correction and standardization, and generate the preprocessed load data: ;

[0025] Among them, is the smoothed load data, is the adaptive wavelet basis function, whose shape and scale are controlled by the parameter , is dynamically adjusted according to the local characteristics of the load data, is the original load data, is the unit impulse response function, which is used to dynamically adjust the delay effect of wavelet transform, is the translation parameter, represents the convolution operation, represents the time variable.

[0026] Optionally, the specific steps of S3 include:

[0027] S31. Construct a load distribution prediction model between the monitoring locations and the total vehicle load based on the preprocessed load data;

[0028] S32. Set the load data for each monitoring location and represent the load prediction error through the objective function of the weighted least squares method: ;

[0029] where, is the objective function, representing the weighted load prediction error, is the actual load at the th monitoring location, representing the load information collected by the fiber optic sensor, is the predicted load data at the th monitoring location calculated based on the load distribution prediction model parameters, is the weight of the rd sensor, representing the influence of the measurement accuracy, sensor location, and environmental factors at this location, is the load calculation parameter to be optimized, including all variables related to load calculation, is the total number of monitoring locations, is the regularization parameter, used to control the complexity of the model and prevent overfitting, is the square norm of the parameter , used to constrain the complexity of the model;

[0030] S33. Optimize the objective function through the gradient descent method and update the load calculation parameter : ;

[0031] where, is the load calculation parameter at the th iteration, is the load calculation parameter at the th iteration, is the learning rate, is the objective function with respect to the load calculation parameter , that is: ;

[0032] where, represents the partial derivative of the load prediction function at the th monitoring location with respect to the parameter , is the total number of monitoring locations;

[0033] S34. Through iterative optimization calculation, until the objective function converges, obtain the final optimized parameters , and recalculate the actual load at each monitoring position ;

[0034] S35. After obtaining the actual load at each monitoring position recalculated , infer the total vehicle load information based on the weighted average method : ;

[0035] wherein, is the total vehicle load information, is the weight of the th sensor, is the total number of monitoring positions;

[0036] S36. If the load distribution between monitoring positions is uneven, adopt the weighted Gaussian process regression algorithm to optimize the load information inference and generate an optimized load distribution prediction model;

[0037] S37. Based on the total vehicle load information, use the time series analysis method to predict the future load across time periods.

[0038] Optionally, the specific content of S36 includes:

[0039] S361. First, detect the load distribution between monitoring positions, judge whether there is uneven distribution of load data, and determine the range of monitoring positions that need to be optimized;

[0040] S362. According to the detected uneven load distribution, adopt the weighted Gaussian process regression algorithm for optimization: ;

[0041] wherein, is the weighted Gaussian process regression function, is the feature vector of the monitoring position, is the actual load data of each monitoring position, is the hyperparameter of the Gaussian process regression model, is the kernel matrix, is the total number of monitoring positions;

[0042] S363. Based on the output of the weighted Gaussian process regression model, update the load value of each monitoring position in the load distribution prediction model to obtain the optimized load distribution prediction result;

[0043] S364. According to the optimized load distribution prediction result, generate a new load distribution prediction model and use it to infer the load information of other monitoring positions.

[0044] Optionally, S4 specifically includes:

[0045] S41. Encrypt the total vehicle load information using a symmetric encryption algorithm to obtain the encrypted total vehicle load information ;

[0046] S42. Utilize blockchain technology to store the encrypted total vehicle load information in the blockchain network, generating a transaction data containing the encrypted load information;

[0047] S43. After verifying and confirming the validity of the transaction data through the consensus mechanism of the blockchain network, generate and store the final block data;

[0048] S44. Based on the encrypted total vehicle load information, through the data access permission management mechanism in the blockchain network, only authorized users can access the relevant total vehicle load information.

[0049] Optionally, S6 specifically includes:

[0050] S61. Receive the encrypted total load data from the blockchain network through the remote monitoring platform , and decrypt the received encrypted total load data to obtain the decrypted total vehicle load information ;

[0051] S62. Based on the decrypted total vehicle load information , compare it with a preset load threshold to obtain a judgment result of overloaded load. The judgment condition is: ;

[0052] Wherein, represents a judgment function for whether the load is overloaded. If the judgment result is , it means overloading. If the judgment result is , it means no overloading. is the decrypted total vehicle load information, is the preset load threshold;

[0053] S63. If the judgment result is true, that is, the decrypted total vehicle load information of the vehicle exceeds the preset threshold , then automatically trigger the alarm mechanism for overloading alarm processing;

[0054] S64. Send alarm information to relevant personnel through the remote monitoring platform. The alarm information includes the decrypted total vehicle load information of overloading and the actual load at the relevant monitoring location .

[0055] Optionally, the S7 specifically includes:

[0056] S71. Regularly extract the encrypted total load data from the blockchain network , and verify the extracted encrypted total load data;

[0057] S72. Use the hash algorithm to perform a hash operation on the extracted encrypted total load data to generate a data hash value , and compare the data hash value with the previous hash value stored in the blockchain network to determine whether the encrypted total load data has been tampered with;

[0058] S73. If the comparison result indicates that the encrypted total load data has not been tampered with, continue the audit process. If the encrypted total load data has been tampered with, trigger a data anomaly alarm and record the tampering log;

[0059] S74. During the audit process, further regularly verify the encrypted total load data in the blockchain through a smart contract and update the relevant audit report;

[0060] S75. Generate an audit report from the audit results and store it in the blockchain;

[0061] S76. To quantify the data consistency during the audit process, evaluate the degree of data change by calculating the data consistency index : ;

[0062] Wherein, is the data consistency index during the audit process, is the total number of encrypted total load data extracted during the audit period, and are the hash values of the current and the previous data blocks respectively. If the consistency index value is close to zero, it indicates that the data consistency is high and no abnormal changes have occurred.

[0063] The beneficial effects of the present invention are:

[0064] (1) By combining blockchain technology, weighted Gaussian process regression algorithm and smart contract, the present invention monitors and encrypts the vehicle load information in real time, ensuring the security and immutability of the load data, thus effectively solving the security hidden danger problems in the data storage and transmission processes of traditional methods. Utilizing the decentralized feature of the blockchain, the present invention can prevent the data from being illegally tampered with and ensure the integrity and reliability of the data.

[0065] (2) By optimizing the load distribution through the weighted Gaussian process regression algorithm, the present invention can accurately calculate the load information at different monitoring positions, solving the problem of low prediction accuracy of load distribution in traditional methods. Through weighted optimization, this algorithm better reflects the variation law of load data, improving the accuracy and real-time performance of the prediction model.

[0066] (3) By combining the hash algorithm and the audit mechanism of the blockchain network, the present invention realizes the regular audit and verification of the load data stored in the blockchain, preventing data tampering during storage and transmission. By quantifying the data consistency index, the present invention can efficiently evaluate the data change situation, further ensuring the credibility and data consistency of the vehicle load information.

[0067] (4) The present invention provides a real-time monitoring and automatic alarm triggering mechanism to ensure that when the vehicle load exceeds the limit, an alarm can be immediately sent to relevant personnel for timely processing, reducing the need for manual intervention and improving the response speed and processing efficiency. Description of the Drawings

[0068] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0069] Figure 1 is a flowchart of a real-time vehicle load monitoring method proposed by the present invention. Detailed Embodiment

[0070] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only illustrating the basic structure of the present invention in a schematic manner, so they only show the components related to the present invention.

[0071] Refer to Figure 1 , a real-time vehicle load monitoring method, including the following steps:

[0072] S1. Deploy a distributed fiber optic sensor array at multiple load monitoring positions of the vehicle, collect the load data of each part of the vehicle in real time, and transmit the load data to the data processing unit inside the vehicle through the optical fiber;

[0073] In this embodiment, by deploying a distributed fiber optic sensor array at multiple load monitoring positions of the vehicle, collecting the load data of each part in real time and transmitting it to the data processing unit through the optical fiber, high-precision and real-time load data monitoring can be achieved.

[0074] S2. Preprocess the load data transmitted to the data processing unit, remove noise and perform preliminary calibration to obtain the preprocessed load data;

[0075] In this embodiment, S2 specifically includes:

[0076] S21. Initially screen the load data transmitted to the data processing unit to remove instantaneous interference signals during the acquisition process;

[0077] S22. Denoise the initially screened load data. Using the multi-point weighted Kalman filtering algorithm, in combination with the spatial positions and measurement accuracies of each sensor, dynamically adjust the Kalman gain matrix : ; ;

[0078] where is the load calculation parameter at the th iteration, is the estimated value of the load data at the previous update, is the Kalman gain matrix weighted based on the sensor positions, adjusting the noise effects of different sensors, is the th load data measurement value, is the observation matrix, is the th sensor's weight, is the covariance matrix of the th sensor at the previous moment, is the covariance of the measurement noise, represents the number of sensors participating in the filtering process;

[0079] S23. Based on the load data after Kalman filtering, perform data correction. Based on the relationship between the sensor positions and the vehicle load distribution, use the known load standard to adjust the load data measured by each sensor;

[0080] S24. Standardize the corrected load data, converting the load data of different sensors into a unified dimension and magnitude, making the output data of each sensor comparable and eliminating the biases of different sensors in data processing;

[0081] S25. Use the method based on adaptive wavelet packet transform to perform time-domain smoothing on the load data after noise removal, filtering, correction, and standardization, generating the preprocessed load data: ;

[0082] where is the smoothed load data, is the adaptive wavelet basis function, whose shape and scale are controlled by the parameter ; Dynamically adjust according to the local characteristics of the load data is the original load data is the unit impulse response function, which is used to dynamically adjust the time delay effect of wavelet transform is the translation parameter represents the convolution operation represents the time variable

[0083] In this embodiment, by preprocessing the load data transmitted to the data processing unit, removing noise and performing preliminary correction, the accuracy and reliability of the load data can be effectively improved. By eliminating the interference and errors that may occur during the sensor acquisition process, the accuracy of subsequent data analysis and processing is ensured, and misjudgment or misoperation caused by noise or inconsistent data is avoided

[0084] S3. Based on the preprocessed load data, use the weighted least squares method to calculate the actual load at each monitoring position, and deduce the total vehicle load information of the vehicle based on the actual load

[0085] In this embodiment, S3 specifically includes

[0086] S31. According to the preprocessed load data, construct a load distribution prediction model between the monitoring position and the total vehicle load

[0087] S32. Set the load data at each monitoring position, and represent the load prediction error through the objective function of the weighted least squares method ;

[0088] where is the objective function, representing the weighted load prediction error is the actual load at the th monitoring position, representing the load information collected by the fiber optic sensor is the predicted load data at the th monitoring position calculated based on the load distribution prediction model parameters is the weight of the th sensor, representing the influence of the measurement accuracy, sensor position and environmental factors at this position is the load calculation parameter to be optimized, including all variables related to load deduction is the total number of monitoring positions is the regularization parameter, which is used to control the complexity of the model and prevent overfitting is the parameter squared norm, which is used to constrain the complexity of the model

[0089] S33. Optimize the objective function by the gradient descent method and update the load calculation parameters : ;

[0090] Among them, is the load calculation parameter at the th iteration, is the load calculation parameter at the th iteration, is the learning rate, is the objective function with respect to the load calculation parameter gradient, that is: ;

[0091] Among them, represents the partial derivative of the load prediction function at the th monitoring position with respect to the parameter , is the total number of monitoring positions;

[0092] S34. Through iterative optimization calculation until the objective function converges, obtain the final optimized parameters , and recalculate the actual load at each monitoring position ;

[0093] S35. After obtaining the actual load at each recalculated monitoring position , estimate the total vehicle load information based on the weighted average method : ;

[0094] Among them, is the total vehicle load information, is the weight of the th sensor, is the total number of monitoring positions;

[0095] S36. If the load distribution among the monitoring positions is uneven, then use the weighted Gaussian process regression algorithm to optimize the load information estimation and generate an optimized load distribution prediction model;

[0096] S37. Based on the total vehicle load information, use the time series analysis method to predict the future load across time periods.

[0097] The specific content of the said S36 includes:

[0098] S361. First, detect the load distribution among the monitoring positions, judge whether there is an uneven distribution of the load data, and determine the range of monitoring positions that need to be optimized;

[0099] S362. Optimize according to the detected uneven load distribution condition by using the weighted Gaussian process regression algorithm: ;

[0100] Wherein, is the weighted Gaussian process regression function, is the feature vector of the monitoring position, is the actual load data of each monitoring position, is the hyperparameter of the Gaussian process regression model, is the kernel matrix, is the total number of monitoring positions;

[0101] S363. Based on the output of the weighted Gaussian process regression model, update the load value of each monitoring position in the load distribution prediction model to obtain the optimized load distribution prediction result;

[0102] S364. According to the optimized load distribution prediction result, generate a new load distribution prediction model and use it to calculate the load information of other monitoring positions.

[0103] In this embodiment, by calculating the actual load of each monitoring position using the weighted least squares method and calculating the total load information of the vehicle based on the actual load, the load distribution condition of each monitoring position can be accurately reflected. This method effectively considers the different weights of each monitoring position and ensures the accurate calculation and inference of the load information.

[0104] S4. After encrypting the calculated total vehicle load information, use blockchain technology to store the encrypted total vehicle load information in the blockchain network to generate encrypted total load data;

[0105] In this embodiment, S4 specifically includes:

[0106] S41. Encrypt the total vehicle load information by using a symmetric encryption algorithm to obtain the encrypted total vehicle load information ;

[0107] S42. Use blockchain technology to store the encrypted total vehicle load information in the blockchain network to generate a transaction data containing the encrypted load information;

[0108] S43. After verifying and confirming the validity of the transaction data through the consensus mechanism of the blockchain network, generate and store the final block data;

[0109] S44. Based on the encrypted total vehicle load information, through the data access permission management mechanism in the blockchain network, only authorized users can access the relevant total vehicle load information.

[0110] In this embodiment, the calculated total vehicle load information is encrypted, and the encrypted data is stored in the blockchain network using blockchain technology, ensuring the security and integrity of the vehicle load information during transmission and storage.

[0111] S5. Build a remote monitoring platform, and through the wireless communication network, transmit the encrypted total load data stored on the blockchain to the remote monitoring platform;

[0112] In this embodiment, the encrypted total load data stored on the blockchain is transmitted to the remote monitoring platform through the wireless communication network, realizing the convenience of remote real-time monitoring and data access. Using the wireless communication network, it can ensure the timely transmission of data, without physical restrictions, and realize the remote monitoring and real-time acquisition of vehicle load information.

[0113] S6. On the remote monitoring platform, based on the encrypted total load data stored in the blockchain, conduct real-time analysis. When the encrypted total load data exceeds the preset threshold, automatically trigger the alarm mechanism;

[0114] In this embodiment, S6 specifically includes:

[0115] S61. Receive the encrypted total load data from the blockchain network through the remote monitoring platform , and decrypt the received encrypted total load data to obtain the decrypted total vehicle load information ;

[0116] S62. Based on the decrypted total vehicle load information , compare it with the preset load threshold to obtain the judgment result of overloading. The judgment condition is: ;

[0117] Among them, represents the judgment function of whether overloading occurs. If the judgment result is , it means overloading. If the judgment result is , it means no overloading. is the decrypted total vehicle load information, is the preset load threshold;

[0118] S63. If the judgment result is true, that is, the decrypted total vehicle load information of the vehicle exceeds the preset threshold , then automatically trigger the alarm mechanism for overloading alarm processing;

[0119] S64. Send an alarm message to relevant personnel through the remote monitoring platform. The alarm message includes the decrypted total vehicle load information of overloading and the actual load at the relevant monitoring location .

[0120] In this embodiment, by analyzing the encrypted total load data stored in the blockchain in real time on the remote monitoring platform, it is possible to timely detect whether the load data exceeds the preset safety threshold. When the encrypted total load data exceeds the set threshold, the system can automatically trigger the alarm mechanism, thus realizing the real-time monitoring and rapid response to the vehicle load status.

[0121] S7. Regularly audit the encrypted total load data stored in the blockchain

[0122] In this embodiment, S7 specifically includes:

[0123] S71. Regularly extract the encrypted total load data from the blockchain network , and verify the extracted encrypted total load data;

[0124] S72. Use the hash algorithm to perform a hash operation on the extracted encrypted total load data to generate a data hash value , and compare the data hash value with the previous hash value stored in the blockchain network to determine whether the encrypted total load data has been tampered with;

[0125] S73. If the comparison result indicates that the encrypted total load data has not been tampered with, continue the audit process. If the encrypted total load data has been tampered with, trigger a data anomaly alarm and record the tampering log;

[0126] S74. During the audit process, further regularly verify the encrypted total load data in the blockchain through a smart contract and update the relevant audit report;

[0127] S75. Generate an audit report from the audit results and store it in the blockchain;

[0128] S76. To quantify the data consistency during the audit process, evaluate the degree of data change by calculating the data consistency index : ;

[0129] Where is the data consistency index during the audit process, is the total number of encrypted total load data extracted during the audit period, and are the hash values of the current and the previous data blocks respectively. If the consistency index A value close to zero indicates a high degree of data consistency and no abnormal changes.

[0130] In this embodiment, by regularly auditing the encrypted total load data stored in the blockchain, the integrity and security of the data are ensured. During the auditing process, the hash algorithm is used to verify whether the data has been tampered with, and the data verification and report update are automated through smart contracts.

[0131] Example:

[0132] To verify the feasibility of the present invention, in this embodiment, the method is applied to the fleet management system of a logistics transportation company (hereinafter referred to as "logistics company"). The company focuses on the transportation of heavy goods, and all vehicles need to strictly comply with the load requirements during transportation. To avoid overloading, reduce traffic accidents caused by overloading, and prevent legal penalties, the logistics company decides to introduce a real-time vehicle load monitoring system based on distributed fiber optic sensors, blockchain technology, and a remote monitoring platform.

[0133] The logistics company has more than 100 transport trucks, and each truck needs to transport heavy goods, which poses strict requirements for vehicle load monitoring. Since most of the transportation routes pass through mountain roads and remote areas, traditional weighing methods (such as static weighing on the road surface) often cannot monitor the load of the fleet in real time and are easily affected by human factors. To improve the monitoring accuracy and ensure timely detection of overloading, the logistics company decides to adopt the real-time vehicle load monitoring method of the present invention, and collect the load data of each truck through a distributed fiber optic sensor array.

[0134] First, the logistics company deployed a distributed fiber optic sensor array at multiple key load monitoring positions of each transport truck. These sensors can collect the load data of each monitoring position in real time and transmit the data to the data processing unit inside the truck through the fiber optic transmission system. This not only enables detailed load monitoring of various parts of the vehicle but also ensures the real-time and accuracy of the data.

[0135] Next, all the load data transmitted to the data processing unit will be preprocessed, including noise removal and preliminary correction. The preprocessed data will calculate the actual load of each monitoring position through the weighted least squares method. Based on the calculated data, the system will further deduce the total load information of the entire vehicle. These data will be encrypted and stored in the blockchain network using blockchain technology to ensure the security and immutability of the data.

[0136] Then, the encrypted total load data stored in the blockchain is transmitted to the remote monitoring platform via a wireless communication network. The monitoring platform will conduct real-time analysis based on this data. When the encrypted total load data exceeds the preset threshold, the system will automatically trigger an alarm mechanism to alert relevant personnel to take actions.

[0137] Finally, to ensure the long-term validity and credibility of the data, the system will also regularly audit the encrypted total load data in the blockchain. Through smart contracts and hash algorithms, the system can verify whether the data has been tampered with, ensuring the integrity and reliability of the vehicle load data.

[0138] During the implementation period, the logistics company conducted a 3-month test on the system. During the test period, the system monitored the load conditions of 20 transport trucks in real time, especially focusing on monitoring during peak periods and sections where overloading frequently occurred. The following is the specific data report during the test:

[0139] Table 1: Report on the Monitoring Effect of Vehicle Loads of the Logistics Company

[0140] As can be seen from Table 1, after implementing the vehicle load real-time monitoring system of the present invention, the logistics company has significantly improved the detection accuracy of overloading incidents, reduced the false alarm rate, and greatly shortened the response time for overloading incidents. Specifically, before the system was deployed, the logistics company detected approximately 42 overloading incidents per month, while after the system was deployed, this figure dropped to only 3 per month. In addition, the response time for overloading incidents has also been shortened from 120 minutes to only 15 minutes, ensuring the immediate handling of overloading problems and avoiding traffic accidents and legal risks caused by overloading.

[0141] In terms of the system alarm response, through the real-time alarm mechanism, the response time after the system triggers the alarm has been shortened from 45 minutes to 5 minutes, greatly improving the processing efficiency. Due to the application of blockchain technology, the load data of all vehicles has been encrypted and protected and cannot be tampered with. The system ensures the credibility and transparency of the data by regularly auditing this data.

[0142] After implementing the vehicle load real-time monitoring method of the present invention, the logistics company has been significantly improved in terms of monitoring accuracy, response speed for overloading incidents, system stability, etc. The introduction of data auditing and blockchain technology makes the data processing and storage of the system more secure and reliable, and effectively reduces the safety hazards and legal risks brought by overloading.

[0143] The above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, shall be covered by the protection scope of the present invention.

Claims

1. A real-time vehicle load monitoring method, characterized in that, It includes the following steps: S1. Deploy a distributed fiber optic sensor array at multiple load monitoring positions of the vehicle, collect the load data of each part of the vehicle in real time, and transmit the load data to the data processing unit inside the vehicle through the optical fiber; S2. Preprocess the load data transmitted to the data processing unit, remove noise and perform preliminary correction to obtain the preprocessed load data; S3. Based on the preprocessed load data, use the weighted least squares method to calculate the actual load of each monitoring position, and deduce the total vehicle load information of the vehicle based on the actual load; S4. After encrypting the calculated total vehicle load information, use blockchain technology to store the encrypted total vehicle load information in the blockchain network to generate encrypted total load data; S5. Build a remote monitoring platform, and transmit the encrypted total load data stored on the blockchain to the remote monitoring platform through a wireless communication network; S6. On the remote monitoring platform, perform real-time analysis based on the encrypted total load data stored in the blockchain. When the encrypted total load data exceeds the preset threshold, automatically trigger the alarm mechanism; S7. Regularly audit the encrypted total load data stored in the blockchain.

2. The vehicle load real-time monitoring method according to claim 1, characterized in that, The specific content of S2 includes: S21. Perform preliminary screening on the load data transmitted to the data processing unit to remove the instantaneous interference signals during the acquisition process; S22. Denoise the load data after preliminary screening. Using the multi-point weighted Kalman filtering algorithm, combine the spatial position and measurement accuracy of each sensor to dynamically adjust the Kalman gain matrix : ； ； Among them, is the load calculation parameter at the th iteration, is the estimated value of the load data at the previous update, is the Kalman gain matrix weighted based on the sensor positions, which adjusts the noise effects of different sensors, is the th measured value of the load data, is the observation matrix, is the weight of the th sensor, is the covariance matrix of the th sensor at the previous moment, is the covariance of the measurement noise, represents the number of sensors participating in the filtering process; S23. Based on the load data after Kalman filtering processing, perform data correction. Based on the relationship between the sensor position and the vehicle load distribution, use the known load standard to adjust the load data measured by each sensor; S24. Perform standardization processing on the corrected load data, convert the load data of different sensors into a unified dimension and magnitude, make the output data of each sensor comparable, and eliminate the deviation of different sensors in data processing; S25. Use the method based on adaptive wavelet packet transform to perform time-domain smoothing processing on the load data after noise removal, filtering, correction and standardization to generate the preprocessed load data: ； in, is the smoothed load data, is an adaptive wavelet basis function, whose shape and scale are determined by the parameters control, Dynamically adjust according to the local characteristics of the load data, is the original load data, is the unit impulse response function, which is used to dynamically adjust the delay effect of wavelet transform. is the translation parameter, represents the convolution operation, Represents a time variable.

3. A real-time vehicle load monitoring method according to claim 1, characterized in that, The specific content of S3 includes: S31. According to the preprocessed load data, build a load distribution prediction model between the monitoring position and the total vehicle load; S32. Set the load data of each monitoring position, and represent the load prediction error through the objective function of the weighted least squares method; ； Among them, is the objective function, representing the weighted load prediction error, is the actual load at the th monitoring location, representing the load information collected by the fiber optic sensor, is the predicted load data at the th monitoring location calculated based on the load distribution prediction model parameters, is the weight of the th sensor, representing the influence of the measurement accuracy, sensor location and environmental factors at this location, is the load calculation parameter to be optimized, including all variables related to load calculation, is the total number of monitoring locations, is the regularization parameter, used to control the complexity of the model and prevent overfitting, is the square norm of the parameter , used to constrain the complexity of the model; S33. Optimize the objective function by gradient descent method and update the load calculation parameters : ； Among them, is the load calculation parameter at the th iteration, is the load calculation parameter at the th iteration, is the learning rate, is the objective function with respect to the load calculation parameter gradient, that is: ； Among them, represents the partial derivative of the load prediction function for the -th monitoring position with respect to the parameter , and is the total number of monitoring positions; S34. Through iterative optimization calculations until the objective function converges, the final optimized parameters are obtained , and recalculate the actual load at each monitoring location ; S35. After obtaining the actual load at each monitoring location that has been recalculated infer the total vehicle load information based on the weighted average method : ； Among them, is the total vehicle load information, is the weight of the th sensor, is the total number of monitoring positions; S36. If the load distribution between the monitoring positions is uneven, use the weighted Gaussian process regression algorithm to optimize the load information deduction to generate an optimized load distribution prediction model; S37. Based on the total vehicle load information, use the time series analysis method to perform cross-period load prediction on the future load.

4. A real-time vehicle load monitoring method according to claim 1, characterized in that The specific content of S36 includes: S361. First, detect the load distribution between the monitoring positions, judge whether there is an uneven distribution of the load data, and determine the range of monitoring positions that need to be optimized; S362. According to the detected uneven load distribution situation, use the weighted Gaussian process regression algorithm for optimization: ； Among them, is the weighted Gaussian process regression function, is the feature vector of the monitoring location, is the actual load data of each monitoring location, is the hyperparameter of the Gaussian process regression model, is the kernel matrix, is the total number of monitoring locations; S363. Based on the output of the weighted Gaussian process regression model, update the load value of each monitoring position in the load distribution prediction model to obtain the optimized load distribution prediction result; S364. Generate a new load distribution prediction model based on the optimized load distribution prediction results and use it to calculate the load information at other monitoring positions.

5. A real-time vehicle load monitoring method according to claim 1, characterized in that, The specific steps of S4 are as follows: S41. Use a symmetric encryption algorithm to encrypt the total vehicle load information to obtain the encrypted total vehicle load information ; S42. Use blockchain technology to store the encrypted total vehicle load information into the blockchain network, generating a transaction data containing the encrypted load information. S43. After verifying and confirming the validity of the transaction data through the consensus mechanism of the blockchain network, generate and store the final block data. S44. Based on the encrypted total vehicle load information, through the data access permission management mechanism in the blockchain network, only authorized users can access the relevant total vehicle load information.

6. A real-time vehicle load monitoring method according to claim 1, characterized in that The specific steps of S6 are as follows: S61. Receiving encrypted total load data from a blockchain network through a remote monitoring platform , and decrypting the received encrypted total load data to obtain the decrypted vehicle total load information ; S62. Based on the decrypted total vehicle load information , compare it with the preset load threshold to obtain the judgment result of overloaded load. The judgment condition is: ； Among them, is a judgment function for determining whether the load exceeds the limit. If the judgment result is , it means overloading. If the judgment result is , it means no overloading. is the total vehicle load information after decryption, is the preset load threshold; S63. If the judgment result is true, that is, the total vehicle load information after vehicle decryption exceeds the preset threshold , the alarm mechanism will be automatically triggered to handle the overloading alarm; S64. Send an alarm message to relevant personnel through the remote monitoring platform. The alarm message includes the decrypted total vehicle load information of overloading and the actual load at the relevant monitoring location .

7. A real-time vehicle load monitoring method according to claim 1, characterized in that, The specific steps of S7 are as follows: S71. Regularly extract the encrypted total load data from the blockchain network , and verify the extracted encrypted total load data; S72. Use a hash algorithm to perform a hash operation on the extracted encrypted total payload data to generate a data hash value , and compare the data hash value with the previous hash value stored in the blockchain network to determine whether the encrypted total payload data has been tampered with; S73. If the comparison result shows that the encrypted total load data has not been tampered with, continue the audit process; if the encrypted total load data has been tampered with, trigger a data anomaly alarm and record the tampering log. S74. During the audit process, further periodically verify the encrypted total load data in the blockchain through a smart contract and update the relevant audit report. S75. Generate an audit report from the audit results and store it in the blockchain. S76. To quantify the data consistency in the audit process, the data consistency index is calculated to evaluate the degree of data change: ； Among them, is the data consistency index during the audit process, is the total number of encrypted total load data extracted within the audit period, and are the hash values of the current and previous data blocks respectively. If the consistency index value is close to zero, it indicates high data consistency and no abnormal changes have occurred.

Images