A cargo transportation state monitoring system based on internet of things

Abstract

The application discloses a kind of based on Internet of Things's cargo transport state monitoring system, it is related to logistics management technical field, by electromagnetic interference monitoring module real-time record each frequency band electromagnetic signal intensity in container, combine the multiple-point temperature data and temperature deviation information collected by redundant sensor module, utilize data processing and transmission module to the signal is preprocessed and uploaded to cloud platform, comprehensive analysis module extracts frequency spectrum interference intensity abnormal characteristics and temperature difference fluctuation characteristics, accurately assesses the influence of electromagnetic interference on sensor data accuracy;Subsequently, interference division module divides interference event into instantaneous interference and sustained interference, early warning processing module triggers temperature control equipment adjustment with low-level early warning for instantaneous interference delay, and generates grading early warning signal and guides coping strategy for sustained interference by in-depth analysis, effectively improve the data reliability, equipment stability and goods protection capability of cold chain transport system in complex electromagnetic environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of logistics management, and particularly relates to a cargo transportation state monitoring system based on Internet of Things. BACKGROUND

[0002] Cargo transportation state monitoring based on Internet of Things refers to real-time monitoring and management of various state information of goods in the transportation process, such as location, temperature, humidity, vibration, inclination, illumination and other environmental parameters, through Internet of Things technology. Such a system usually includes sensors, communication modules, cloud platforms and data analysis tools. The sensors collect cargo state information, the communication modules transmit data to the cloud, and the data can provide real-time warnings and reports after being analyzed, helping logistics companies and customers improve transportation efficiency and safety. Specifically, in cold chain logistics, when transporting fresh food or vaccines, temperature is a key factor. After the goods are loaded, the temperature sensors in each container will monitor the temperature data in real time, and the data will be transmitted to the cloud platform through the cellular network. Once the temperature exceeds the preset range, the system will immediately send an alarm to the driver and the management personnel, reminding them to take measures (such as adjusting the refrigeration equipment or re-planning the transportation route). This way ensures the quality of the goods, while improving transportation transparency and customer satisfaction.

[0003] The prior art has the following deficiencies:

[0004] Some cold chain transportation vehicles use high-power motors or other electrical equipment, which may cause electromagnetic interference (EMI) to the communication or readings of the sensors. This can cause the sensors to collect false data or fail to communicate normally. In addition, when electromagnetic interference causes sensor data distortion, the temperature control equipment may receive false information, causing abnormal temperature regulation. For example, when transporting fresh fruits, the temperature control equipment may overcool the fruits, causing them to freeze, affecting quality and value. SUMMARY

[0005] The purpose of the present application is to provide a cargo transportation state monitoring system based on Internet of Things to solve the deficiencies in the background art.

[0006] In order to achieve the above-mentioned purpose, the present application provides the following technical solution: a cargo transportation state monitoring system based on Internet of Things, comprising an electromagnetic interference monitoring module, a redundant sensor module, a data processing and transmission module, a comprehensive analysis module, an interference division module and an early warning processing module.

[0007] The electromagnetic interference monitoring module: a plurality of electromagnetic interference monitors are installed in the container for real-time recording of electromagnetic signal strength data in several frequency bands.

[0008] Redundant sensor module: comprising at least two temperature sensors for simultaneously collecting temperature data in the container, and calculating temperature deviation data between the temperature sensors in several time periods;

[0009] Data processing and transmission module: for preprocessing electromagnetic signal intensity data and temperature deviation data, and uploading the processed data to the cloud platform through the cellular network;

[0010] Comprehensive analysis module: respectively extracting the spectral interference intensity abnormality features in the preprocessed electromagnetic signal intensity data and the temperature difference fluctuation features in the temperature deviation data, and evaluating the influence degree of electromagnetic interference on sensor data accuracy according to the extracted spectral interference intensity abnormality features and temperature difference fluctuation features;

[0011] Interference division module: for classifying interference events into transient interference and continuous interference according to the influence degree of electromagnetic interference on sensor data accuracy;

[0012] Early warning processing module: for transient interference events, delaying the triggering of temperature control equipment adjustment, and sending low-level early warning notification to the manager; for continuous interference events, further analyzing the influence degree of electromagnetic interference on sensor data accuracy in a fixed time period, and classifying and processing the continuous interference events according to the analysis results.

[0013] Preferably, in the comprehensive analysis module, the spectral interference intensity abnormality index is generated after analyzing the extracted spectral interference intensity abnormality features in the electromagnetic signal intensity data, and the method for obtaining the spectral interference intensity abnormality index is as follows:

[0014] Let the spectral signal data be X, with dimension n x m, where n is the number of samples and m is the number of features, is the signal intensity of the i-th sample in the j-th frequency band, first standardize X, the expression is: ; In the formula, is the standardized data, is the mean of the signal intensity of the j-th frequency band, is the standard deviation of the signal intensity of the j-th frequency band, get the standardized matrix Z, calculate the covariance matrix C of the standardized matrix Z, the expression is: ; In the formula, C is an m x m covariance matrix, each element represents the linear correlation between frequency band j and frequency band k, T is the matrix transpose, perform eigenvalue decomposition on the covariance matrix C to get: ; In the formula, is the k-th eigenvalue, and represents the variance contribution rate of the principal component, For the kth eigenvector, the direction corresponding to the principal component, the eigenvalues are arranged in descending order, and the first p largest eigenvalues and the corresponding eigenvectors constitute the principal component subspace, and the standardized data Z is projected into the principal component subspace: ; In the formula, Y is the projected principal component matrix, the dimension is n×p, is a matrix composed of the first p eigenvectors, the dimension is m×p, and the original data is reconstructed using the projected principal component data Y: ; In the formula, is the reconstructed standardized data, the dimension is n×m, and the reconstruction error of each sample is calculated: ; In the formula, is the reconstruction error of the ith sample, and the spectral interference intensity anomaly index is calculated according to the reconstruction error , the expression is: ; In the formula, is the mean of the reconstruction error of all samples, std(E) is the standard deviation of the reconstruction error of all samples, and HKS is the spectral interference intensity anomaly index.

[0015] Preferably, in the comprehensive analysis module, the temperature difference fluctuation characteristics in the extracted temperature deviation data are analyzed to generate a temperature difference fluctuation index, and the method for obtaining the temperature difference fluctuation index is:

[0016] Let the temperature difference data be T(t), defined as a one-dimensional time series, with a length of N and a sampling interval of Δt, ; The time series T(t) is converted into a frequency domain representation F(f) using Fourier transform, and the expression is: ; In the formula, is a complex value of frequency , j is an imaginary unit, is the kth frequency component, and the amplitude spectrum is extracted from , and the expression is: ; In the formula, is the real part and the imaginary part of ; The amplitude represents the contribution strength of the frequency component to the temperature difference fluctuation, the main frequency characteristics are extracted, the total spectral energy of the temperature difference sequence is calculated , and the expression is: ; reflects the overall strength of the entire temperature difference fluctuation, the energy of the high-frequency component is extracted , and the expression is: ; In the formula, is the threshold frequency of the high-frequency component, and the proportion of the high-frequency energy in the total energy is calculated , and the expression is: ; In the formula, Quantify the proportion of severe fluctuations in temperature difference data, calculate the temperature difference fluctuation index CHQ as the normalized form of high-frequency energy proportion, the expression is: ; In the formula, GHQ is the temperature difference fluctuation index, 、 , the mean and standard deviation of high-frequency energy proportion, for normalization processing.

[0017] Preferably, in the comprehensive analysis module, the spectral interference intensity anomaly index and the temperature difference fluctuation index are converted into a comprehensive feature vector, the comprehensive feature vector is taken as the input of a machine learning model, the machine learning model takes the influence degree value label of electromagnetic interference on sensor data accuracy as the prediction target, and the training target is to minimize the sum of prediction errors of all influence degree value labels of electromagnetic interference on sensor data accuracy, the machine learning model is trained until the sum of prediction errors converges, and the influence degree value of electromagnetic interference on sensor data accuracy is determined according to the model output result, wherein the machine learning model is a polynomial regression model.

[0018] Preferably, in the interference division module, the interference event is classified into transient interference and sustained interference according to the influence degree of electromagnetic interference on sensor data accuracy, specifically:

[0019] The influence degree value of electromagnetic interference on sensor data accuracy obtained is compared with the influence degree value reference threshold preset according to historical data, if the influence degree value of electromagnetic interference on sensor data accuracy is greater than or equal to the influence degree value reference threshold, it means that the influence degree of electromagnetic interference on sensor data accuracy is high, at this time a high interference degree signal is generated, and the interference event is classified as sustained interference; if the influence degree value of electromagnetic interference on sensor data accuracy is less than the influence degree value reference threshold, it means that the influence degree of electromagnetic interference on sensor data accuracy is low, at this time a low interference degree signal is generated, and the interference event is classified as transient interference.

[0020] Preferably, in the early warning processing module, for sustained interference events, the influence degree of electromagnetic interference on sensor data accuracy in a fixed time period is further analyzed, and the sustained interference events are classified and processed according to the analysis result, specifically:

[0021] For sustained interference events, that is, the influence degree value of electromagnetic interference on sensor data accuracy generated in a fixed time period is greater than or equal to the influence degree value reference threshold, the influence degree value of electromagnetic interference on sensor data accuracy greater than or equal to the influence degree value reference threshold generated in the subsequent fixed time period is collected, and a corresponding data set is established, the mean and standard deviation of the data set are calculated, and the sustained interference events are classified and processed according to the analysis result.

[0022] Preferably, if the mean value of the influence degree value in the data set is greater than or equal to the reference threshold value of the mean value of the influence degree value, and the standard deviation of the influence degree value is less than the reference threshold value of the standard deviation of the influence degree value, the interference degree is high and stable, a first-level warning signal is generated at this time, the interference source needs to be checked in priority, and redundant data or protection mode is enabled in the system;

[0023] If the mean value of the influence degree value is greater than or equal to the reference threshold value of the mean value of the influence degree value, and the standard deviation of the influence degree value is greater than or equal to the reference threshold value of the standard deviation of the influence degree value, the interference degree is high and the fluctuation is large, a second-level warning signal is generated at this time, the interference fluctuation trend is monitored, the system adjustment decision is delayed, and the sensor data processing algorithm is optimized;

[0024] If the mean value of the influence degree value is less than the reference threshold value of the mean value of the influence degree value, and the standard deviation of the influence degree value is greater than or equal to the reference threshold value of the standard deviation of the influence degree value, the interference degree is low but the fluctuation is large, a third-level warning signal is generated at this time, and further monitoring is required without immediate intervention;

[0025] If the mean value of the influence degree value is less than the reference threshold value of the mean value of the influence degree value, and the standard deviation of the influence degree value is less than the reference threshold value of the standard deviation of the influence degree value, the interference degree is low and the fluctuation is small, no warning is needed at this time, the system remains normal operation, and no additional adjustment measures are taken.

[0026] In the above technical solution, the present application provides technical effects and advantages:

[0027] 1. The present application effectively solves the problem of sensor data distortion caused by electromagnetic interference in the cold chain transportation process by integrating electromagnetic interference monitoring, redundant sensor data acquisition, cloud data processing and transmission, comprehensive analysis and early warning mechanism. The influence of interference is quantitatively evaluated using the spectral interference intensity anomaly index and the temperature difference fluctuation index, and the analysis accuracy is further improved through a machine learning model to provide a scientific basis for interference event classification and early warning. The present application can accurately identify transient and continuous interference events in complex electromagnetic environments, and based on the hierarchical warning strategy of interference influence degree, it realizes the delayed triggering and intelligent response of device adjustment, effectively ensuring the quality and transportation safety of cold chain goods.

[0028] 2. The present application significantly improves the stability and reliability of the cold chain transportation state monitoring system through innovative data analysis and processing scheme. Especially in the case of continuous interference, the mean value and standard deviation analysis method is used to realize the hierarchical processing of interference events, which helps management personnel to take accurate adjustment measures according to different warning levels. In addition, the system can also reduce the misoperation and energy loss of temperature control equipment, while enhancing the real-time monitoring capability of the transportation environment of goods, and improving the efficiency and intelligent level of cold chain logistics management. BRIEF DESCRIPTION OF DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0030] Figure 1 The system module diagram of the present application. DETAILED DESCRIPTION

[0031] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0032] Embodiment, please refer to Figure 1 As shown in the figure, the present embodiment describes a cargo transportation state monitoring system based on Internet of Things, which comprises an electromagnetic interference monitoring module, a redundant sensor module, a data processing and transmission module, a comprehensive analysis module, an interference division module and an early warning processing module.

[0033] The electromagnetic interference monitoring module: a plurality of electromagnetic interference monitors are installed in the cargo box for real-time recording of electromagnetic signal strength data in several frequency bands.

[0034] The redundant sensor module: comprising at least two temperature sensors for simultaneously collecting temperature data in the cargo box and calculating temperature deviation data between the temperature sensors in several time periods.

[0035] The data processing and transmission module: for preprocessing electromagnetic signal strength data and temperature deviation data, and uploading the processed data to the cloud platform through the cellular network.

[0036] The comprehensive analysis module: respectively extracting the spectral interference intensity abnormal features in the preprocessed electromagnetic signal strength data and the temperature difference fluctuation features in the temperature deviation data, and evaluating the influence degree of electromagnetic interference on the accuracy of sensor data according to the extracted spectral interference intensity abnormal features and temperature difference fluctuation features.

[0037] The interference division module: for classifying interference events into transient interference and continuous interference according to the influence degree of electromagnetic interference on the accuracy of sensor data.

[0038] Early warning processing module: for transient interference events, delay triggering temperature control device adjustment and send low-level early warning notice to management personnel; for continuous interference events, further analyze the influence degree of electromagnetic interference on sensor data accuracy within a fixed time period, and process the continuous interference events according to the analysis results.

[0039] In the electromagnetic interference monitoring module, the electromagnetic interference monitor should have: frequency range: support 50Hz to 1GHz frequency band monitoring, covering common electromagnetic interference sources: low frequency band (50Hz-1kHz): related to the interference of industrial equipment such as motors and frequency converters. Medium frequency band (1kHz-30MHz): related to the interference of communication equipment or switching power supply. High frequency band (30MHz-1GHz): related to wireless communication (such as Wi-Fi, 4G / 5G) and microwave interference. Signal acquisition accuracy: dynamic range: -120dBm to 0dBm, ensuring the ability to capture signals from weak interference to strong interference. Resolution bandwidth (RBW): support at least 10kHz RBW, used to identify detailed interference characteristics. Real-time and data transmission: sampling frequency: ≥1 / s, ensuring real-time monitoring of interference changes. Data transmission: built-in cellular communication module or connected with vehicle gateway, supporting real-time uploading of interference data to cloud platform. Operating temperature range: -20°C to 60°C, suitable for cold chain transportation environment. Protection level: IP67, dustproof and waterproof, ensuring normal operation of the device in humid or condensation conditions.

[0040] Multiple monitors are evenly arranged at key positions within the container, such as near sensors and refrigeration equipment, to capture electromagnetic signals in different areas. Increase the density of monitors at known interference source locations, such as near motors or frequency converters. Maintain appropriate spacing (such as ≥1 meter) between monitors to avoid mutual influence of signals between monitors. Small containers (≤10m³): install 2-3 monitors, covering the front, middle and rear of the container. Medium and large containers (10-30m³): install 4-6 monitors, symmetrically arranged along the two sides of the container, focusing on covering sensor and refrigeration equipment positions. In the case of complex sensor layout or dense interference source distribution, increase the number of monitors or use mobile arrangement.

[0041] Each monitor collects electromagnetic signals at fixed time intervals (e.g., 1 second), records signal strength (unit: dBm) and corresponding frequency. The data is time-stamped and sent to the data processing module in real time. Signal preprocessing includes: noise reduction: remove background noise through a band-pass filter, only keep the effective signal of the interference frequency band. Peak detection: extract the interference peak signal in a specific frequency band, which is used to locate the strongest interference source. Feature extraction: record the mean, peak and spectral distribution characteristics of the interference signal. Data is transmitted to the cloud or local storage device through the vehicle gateway. The system integrates and analyzes multi-point monitoring data to generate a thermal distribution map of electromagnetic interference in the cargo box.

[0042] The monitor is calibrated regularly (e.g., every 3 months) to ensure measurement accuracy, especially after long-term use in cold chain environments. When the monitor detects an interference signal strength exceeding the set threshold (e.g., -50 dBm), an alarm mechanism is triggered immediately and the driver or management platform is notified. For example, in cold chain transportation, four electromagnetic interference monitors are installed in the cargo box: two monitors are located at the front and rear of the cargo box, covering the refrigeration equipment; two monitors are located in the middle of the cargo box, near the temperature sensor. Real-time capture of high-frequency interference signals when the refrigeration equipment is running, and monitoring of medium and low-frequency interference near the sensor. When a medium-frequency interference signal of -60 dBm is detected, the system generates an interference source positioning map through the cloud platform, prompting the driver to adjust the operating parameters of the refrigeration equipment to reduce interference.

[0043] In the redundant sensor module, at least two temperature sensors are installed in each cargo box. For larger cargo boxes or temperature distribution uneven scenarios, the number of sensors can be increased (e.g., 4-6). Sensors should be evenly distributed in the cargo box to cover areas with large temperature differences. Increase the sensor density in areas that are easily affected by interference or the environment (e.g., near the refrigeration equipment or dense cargo areas). Measurement range: support -40°C to +85°C to meet the needs of cold chain environments. Accuracy requirements: high-precision sensors with ±0.1°C accuracy for monitoring temperature-sensitive goods (e.g., vaccines). Response time: fast response capability (<2 seconds) to ensure temperature changes are captured. Communication interface: support I²C or UART protocol for easy integration with data processing modules.

[0044] Each sensor collects temperature data at fixed intervals (e.g., 1 second), time-stamps and stores it in local storage or uploads it to the cloud. The deviation calculation formula is: ; T1, T2 are temperature data of two sensors, AT is the temperature deviation between them. Calculate the mean and standard deviation of the deviation data for several time periods (e.g. 1 minute, 5 minutes) to extract temperature difference fluctuation characteristics. Set thresholds (e.g. 0.2°C) according to the type of the container and the performance of the temperature control equipment. If AT > threshold, mark it as data anomaly. Trigger alarm if it exceeds the threshold for several times, indicating interference or equipment failure. When the deviation is within the threshold range, consider the data reliable and directly enter the system analysis. If the deviation exceeds the range, compare and verify it with historical data and spatial models. Weight the data of the two sensors to reduce the influence of individual data distortion: when the deviation of one sensor exceeds a certain fixed threshold (e.g. 1°C), directly exclude the sensor data.

[0045] Minor deviation: deviation within threshold range, marked as minor anomaly, no immediate intervention required.

[0046] Moderate deviation: deviation exceeds threshold for a short time but has no obvious trend, system enables data correction and notifies the management background for attention. Severe deviation: deviation continues to exceed the threshold, system triggers an alarm, prompting to check the sensor or interference source. Automatically exclude abnormal data and enable backup sensors (if more than 2 sensors).

[0047] Calibrate the sensors regularly (e.g. every 6 months) to ensure data accuracy, especially in extreme cold chain environments. The module has a health status monitoring function (e.g. fault self-diagnosis) to report sensor performance in real time. For example, when transporting vaccines in a cold chain, 4 sensors are installed in the container, one at each corner. Calculate the pairwise deviation of the 4 sensors every minute to generate an average deviation matrix. If sensor 3 deviates from other sensors by more than 0.5°C continuously, the system automatically excludes sensor 3 data and enables other sensor data for temperature control equipment adjustment.

[0048] In the data processing and transmission module, receive the frequency spectrum intensity data through the communication protocol of the monitor (e.g. I²C, UART or SPI). Receive real-time temperature values of two or more temperature sensors and calculate their deviation. Convert the collected data into a unified format, such as JSON or CSV format, for subsequent processing and transmission, including fields: timestamp, data type (electromagnetic signal intensity or temperature deviation), data value, device ID or location identifier; Apply a bandpass filter to electromagnetic signal intensity data to filter background noise and irrelevant frequency band signals. Use smoothing algorithms (such as moving average or Kalman filter) for temperature deviation data to reduce data fluctuations.

[0049] Identify abnormal peaks (e.g., interference intensity significantly higher than average) and flag as suspicious interference events. Remove abnormal data with deviation values exceeding a set threshold range (e.g., 0.5°C or higher). Package processed data into batch data packets at fixed time intervals (e.g., every minute) to optimize transmission efficiency. Support 4G / 5G communication protocols to ensure high-speed and stable transmission performance. Design an automatic switching mechanism to switch to Wi-Fi or other backup networks when cellular network signal is weak.

[0050] Encrypt data during transmission (e.g., TLS / SSL protocol) to prevent tampering or theft. Use a verification algorithm (e.g., CRC or MD5) to verify the integrity of transmitted data to ensure error-free upload. Upload electromagnetic signal and temperature deviation data at a frequency of seconds to ensure real-time monitoring on the cloud. For historical data or low-priority data, upload at a minute or longer time interval to reduce communication burden.

[0051] Use standardized interfaces (e.g., RESTful API) to interface with the cloud platform to ensure smooth data upload and reception. Upload logs and status monitoring: record the timestamp, data packet size, and upload status (success or failure) of each data upload. Automatically retry or temporarily store locally when uploading fails and re-upload after network recovery.

[0052] In the event of network interruption, temporarily store the unuploaded data in local storage (e.g., SD card or built-in flash memory) to prevent data loss. After network recovery, upload the unsent data in chronological order in batches. When the network fails or the cloud is unavailable, enable edge computing devices to analyze data locally and generate temporary alerts or adjust strategies. Regularly check the running status of the cellular module, detect communication signal strength and module health status.

[0053] Comprehensive analysis module: extract the spectral interference intensity anomaly features from the pre-processed electromagnetic signal intensity data and the temperature difference fluctuation features from the temperature deviation data, and evaluate the influence of electromagnetic interference on the accuracy of sensor data based on the extracted spectral interference intensity anomaly features and temperature difference fluctuation features.

[0054] Extracting the spectral interference intensity anomaly features in the pre-processed electromagnetic signal intensity data, specifically: intensity mean feature: calculate the average value of electromagnetic signal intensity in the monitoring time period, used to measure the overall interference level. Higher mean value indicates the presence of persistent interference sources in the environment. Peak intensity feature: identify the maximum value of signal intensity in the frequency spectrum and the corresponding frequency band position, which can reveal the main interference source frequency range, such as device operation or external communication signals. Spectral distribution feature: analyze the distribution of interference signals in different frequency bands, record the interference concentrated frequency band and signal width, help to locate specific interference equipment or source. Volatility feature: identify the stability of interference signals by monitoring the amplitude of signal intensity changes over time. Sharp fluctuations usually indicate intermittent interference, while smooth changes may be related to continuously running devices. Interference event density feature: count the number of times the interference intensity exceeds the set threshold, used to quantify the frequency of interference events.

[0055] Extracting the temperature deviation fluctuation features in the pre-processed temperature deviation data, specifically: average deviation feature: calculate the average value of temperature deviation between multiple sensors, used to evaluate overall consistency. Larger average deviation may indicate data distortion or interference influence. Maximum deviation feature: extract the maximum temperature deviation in the time period, used to identify severe abnormal events such as single sensor failure or electromagnetic interference peak. Fluctuation amplitude feature: analyze the up and down fluctuation range of temperature deviation data, larger fluctuation amplitude may indicate that the system is greatly affected by external interference. Deviation trend feature: observe the temperature deviation trend over time, continuous rising or falling deviation may indicate system failure or long-term influence of interference source. Abnormal deviation event feature: count the number of abnormal events where the deviation exceeds the set safety range, help to identify potential problem areas in the temperature control system.

[0056] After analyzing the extracted spectral interference intensity anomaly features in the electromagnetic signal intensity data, generate the spectral interference intensity anomaly index, the method for obtaining the spectral interference intensity anomaly index is:

[0057] Let the spectral signal data be X, with dimensions n x m, where: n is the number of samples (the number of time points sampled), m is the number of features (signal intensity of different frequency bands in the spectral distribution), is the signal intensity of the i-th sample in the j-th frequency band. In order to eliminate the dimensional differences of different frequency band signal intensities, first standardize X, the expression is: ; In the formula, is the standardized data, is the mean value of the j-th frequency band signal intensity, is the standard deviation of the j-th frequency band signal intensity, and the standardized matrix Z (dimension n x m) is obtained. Calculate the covariance matrix C of the standardized matrix Z, the expression is: ; In the formula, C is an m x m covariance matrix, each element represents the linear correlation between frequency band j and frequency band k, T is the matrix transpose, the covariance matrix C is subjected to eigenvalue decomposition to obtain: ; in the formula, is the kth eigenvalue, and the variance contribution rate of the principal component is represented, is the kth eigenvector, corresponding to the direction of the principal component. The eigenvalues are arranged in descending order, and the first p largest eigenvalues and the corresponding eigenvectors are taken to form a principal component subspace, and the standardized data Z is projected into the principal component space: ; in the formula, Y is the projected principal component matrix, the dimension is n×p, is a matrix composed of the first p eigenvectors, the dimension is m×p, and the principal component matrix Y captures the main change mode of the original data. The original data is reconstructed using the projected principal component data Y: ; in the formula, is the reconstructed standardized data, the dimension is n×m, and the reconstruction error of each sample is calculated: ; in the formula, is the reconstruction error of the ith sample, and the spectral interference intensity anomaly index is calculated according to the reconstruction error , and the expression is: ; in the formula, is the mean of the reconstruction error of all samples, std(E) is the standard deviation of the reconstruction error of all samples, and HKS is the spectral interference intensity anomaly index.

[0058] When the spectral interference intensity anomaly index is large, it usually indicates that the spectral signal of a certain sample deviates greatly from the normal distribution predicted by the principal component model, which means that there is strong electromagnetic interference. A larger anomaly index indicates that the interference signal is strong and volatile, or is distributed in a key frequency band (such as the sensor communication frequency band), which has a significant impact on the sensor data communication link or sampling circuit. In this case, the sensor data may have a large deviation or even fail, thereby directly affecting the accuracy and reliability of the monitoring system.

[0059] When the spectral interference intensity anomaly index is small, it means that the spectral characteristics of the sample are close to the normal range predicted by the principal component model, and the strength and distribution of the interference signal have little effect on the sensor data. A smaller anomaly index indicates that the level of electromagnetic interference in the environment is low and stable, and the sensor can operate normally, and the accuracy and consistency of data acquisition are basically not affected by interference. In this case, the system can rely on sensor data for accurate state monitoring and decision-making.

[0060] After analyzing the temperature difference fluctuation characteristics in the extracted temperature deviation data, a temperature difference fluctuation index is generated, and the method for obtaining the temperature difference fluctuation index is:

[0061] Let the temperature difference data be T(t), defined as a one-dimensional time series with length N and sampling interval Δt, ; convert the time series T(t) to the frequency domain representation F(f) using Fourier transform, expressed as: ; where, is the complex value of the frequency , containing amplitude and phase information, and j is the imaginary unit, is the kth frequency component with unit Hz, extracted from Extract the amplitude spectrum , expressed as: ; where, is the real and imaginary parts of ; the amplitude represents the contribution of the frequency component to the temperature difference fluctuation. Extract the main frequency feature and calculate the total spectral energy of the temperature difference sequence , expressed as: ; reflects the overall intensity of the entire temperature difference fluctuation. Extract the energy of the high-frequency component (such as the frequency component of ) from the total energy , expressed as: ; where, is the threshold frequency of the high-frequency component, used to distinguish between smooth fluctuations and sharp fluctuations. Calculate the proportion of high-frequency energy in the total energy , expressed as: ; where, quantifies the proportion of sharp fluctuations in the temperature difference data. Calculate the temperature difference fluctuation index CHQ as the normalized form of the high-frequency energy proportion, expressed as: ; where, GHQ is the temperature difference fluctuation index, , is the mean and standard deviation of the high-frequency energy proportion, used for normalization processing.

[0062] When the temperature difference fluctuation index is larger, it indicates that the temperature difference fluctuation between temperature sensors is significantly enhanced, especially the proportion of high-frequency components increases, showing short-term sharp fluctuations. This may be due to strong electromagnetic interference causing instability in sensor data collection, such as transient signal distortion or reading jumps, making the sensor unable to accurately reflect the actual temperature conditions inside the cargo box. When the fluctuation index is large, the accuracy of the sensor data may decrease significantly, which may lead to misjudgment of the temperature control equipment, thereby affecting the safety of the goods.

[0063] When the temperature difference fluctuation index is smaller, it indicates that the temperature difference fluctuation between the temperature sensors is more stable, the proportion of the high-frequency component of the spectrum is low, and the influence of electromagnetic interference on the sensors is limited. The low fluctuation index reflects that the sensors work in good environmental conditions, and the collected data has high stability and consistency. At this time, the sensor data has high precision, which can provide reliable reference for the temperature control system and ensure the normal operation of the cargo state monitoring and control.

[0064] The spectrum interference intensity anomaly index and the temperature difference fluctuation index are converted into a comprehensive feature vector, the comprehensive feature vector is taken as the input of a machine learning model, the machine learning model takes the influence degree value label of electromagnetic interference on sensor data precision as the prediction target, and the training target is to minimize the sum of prediction errors of all influence degree value labels of electromagnetic interference on sensor data precision, the machine learning model is trained until the sum of prediction errors converges, and the influence degree value of electromagnetic interference on sensor data precision is determined according to the model output result, wherein the machine learning model is a polynomial regression model.

[0065] The method for obtaining the influence degree value of electromagnetic interference on sensor data precision is as follows: from the comprehensive feature vector training data of the trained machine learning model, the corresponding function expression is obtained: ; in the formula, is the output function of the model, HKS is the spectrum interference intensity anomaly index, GHQ is the temperature difference fluctuation index, is the influence degree value of electromagnetic interference on sensor data precision.

[0066] The interference division module is used for classifying the interference events into transient interference and continuous interference according to the influence degree of electromagnetic interference on sensor data precision.

[0067] The obtained influence degree value of electromagnetic interference on sensor data precision is compared with the influence degree value reference threshold preset according to historical data, if the influence degree value of electromagnetic interference on sensor data precision is greater than or equal to the influence degree value reference threshold, it indicates that the influence degree of electromagnetic interference on sensor data precision is high, at this time, a high interference degree signal is generated, and the interference event is classified as continuous interference; if the influence degree value of electromagnetic interference on sensor data precision is less than the influence degree value reference threshold, it indicates that the influence degree of electromagnetic interference on sensor data precision is low, at this time, a low interference degree signal is generated, and the interference event is classified as transient interference.

[0068] Early warning processing module: for transient interference events, delay triggering temperature control device adjustment, and send low-level early warning notice to the management personnel; for continuous interference events, further analyze the influence degree of electromagnetic interference on sensor data accuracy within a fixed time period, and process the continuous interference events according to the analysis results.

[0069] For transient interference events, delay triggering temperature control device adjustment, and send low-level early warning notice to the management personnel, specifically: real-time acquisition of sensor temperature data and electromagnetic interference signal intensity. By setting a transient interference threshold (such as electromagnetic intensity change amplitude or sensor reading jump amplitude), identify interference events. Transient interference characteristics: short duration (such as <10 seconds). The temperature difference fluctuation index or the spectral interference anomaly index increases transiently, but there is no sustained trend.

[0070] After the occurrence of a transient interference event, a buffer time (such as 5-10 seconds) is set, during which the system does not trigger the adjustment operation of the temperature control device. If the interference disappears (the temperature difference fluctuation index returns to the normal range) within the buffer time, the device does not need to be adjusted. If the temperature data returns to normal after the interference ends, the system ignores the event and continues to run. If the interference continues or multiple transient interferences occur within the buffer time, trigger a higher level of processing mechanism (such as entering the moderate interference processing flow). Even if the interference has not completely ended, only low-amplitude adjustments (such as small-range adjustment of refrigeration power) of the temperature control device are allowed after the buffer time to avoid frequent switching causing device overload or large temperature fluctuations of the goods.

[0071] When a transient interference event occurs, record the interference data, including: time stamp interference intensity (such as spectral intensity anomaly index), sensor fluctuation amplitude (such as temperature difference fluctuation index), generate a low-priority early warning, prompt the management personnel that there is a transient interference, but it does not pose a significant risk to the device and goods. Send a low-level early warning notice to the management system or mobile terminal, display the following information: event type: transient interference event. Current state: temperature control device has not triggered adjustment, system monitoring is normal. Recommended measures: no urgent intervention is required, only subsequent attention to interference trends is recommended. Early warning classification: if multiple transient interference events occur in succession (such as 5 times within 1 hour), automatically upgrade to a medium-priority early warning, and recommend that the management personnel investigate the source of the interference.

[0072] For continuous interference events, further analyze the influence degree of electromagnetic interference on sensor data accuracy within a fixed time period, and process the continuous interference events according to the analysis results, specifically:

[0073] For a continuous interference event, that is, the impact degree value of electromagnetic interference generated in a fixed time period on the sensor data accuracy is greater than or equal to the impact degree value reference threshold, the impact degree values of electromagnetic interference greater than or equal to the impact degree value reference threshold generated in subsequent fixed time periods on the sensor data accuracy are collected, and a corresponding data set is established. The mean and standard deviation of the data set are calculated, analyzed, and classified according to the analysis results.

[0074] If the mean of the impact degree values in the data set is greater than or equal to the reference threshold of the mean of the impact degree values, and the standard deviation of the impact degree values is less than the reference threshold of the standard deviation of the impact degree values, the interference degree is high and stable. It indicates that the continuous interference is strong, but the fluctuation is small, and the interference environment is relatively stable. At this time, a first-level warning signal is generated, and the interference source needs to be prioritized to be investigated to try to reduce or shield the interference. Redundant data or protection mode is enabled in the system to ensure the reliability of the sensor data.

[0075] If the mean of the impact degree values is greater than or equal to the reference threshold of the mean of the impact degree values, and the standard deviation of the impact degree values is greater than or equal to the reference threshold of the standard deviation of the impact degree values, the interference degree is high and fluctuates greatly. It indicates that the continuous interference is strong and there is significant fluctuation, which may be caused by multiple source interference superposition or device instability. At this time, a second-level warning signal is generated, and the interference fluctuation trend is monitored, and the system adjustment decision is appropriately delayed. The sensor data processing algorithm (such as weighted average or filtering) is optimized to reduce the influence of fluctuation.

[0076] If the mean of the impact degree values is less than the reference threshold of the mean of the impact degree values, and the standard deviation of the impact degree values is greater than or equal to the reference threshold of the standard deviation of the impact degree values, the interference degree is low but the fluctuation is large. It indicates that the overall impact of the interference is small, but there is instability, which may be an occasional interference or a short-term device problem. At this time, a third-level warning signal is generated, and continuous monitoring is required without immediate intervention. The fluctuation event is recorded for subsequent environment optimization or device fault diagnosis.

[0077] If the mean of the impact degree values is less than the reference threshold of the mean of the impact degree values, and the standard deviation of the impact degree values is less than the reference threshold of the standard deviation of the impact degree values, the interference degree is low and the fluctuation is small. It indicates that the interference intensity is small and stable, and the impact on the system can be ignored. At this time, no warning is needed, the system remains normal operation, and no additional adjustment measures are taken. Only the data is recorded as a long-term monitoring reference.

[0078] It should be noted that the importance of the first-level warning signal is greater than that of the second-level warning signal, and the importance of the second-level warning signal is greater than that of the third-level warning signal. Relevant personnel can take corresponding processing measures according to different warning signal levels.

[0079] In the embodiment, the intelligent management of the electromagnetic interference and temperature monitoring in the container is realized by multi-module cooperation: the electromagnetic interference monitoring module installs multiple monitors to record the electromagnetic signal strength in each frequency band in real time; the redundant sensor module uses at least two temperature sensors to collect temperature data and calculate temperature deviation; the data processing and transmission module pre-processes the electromagnetic signal strength data and temperature deviation data, and uploads them to the cloud platform through the cellular network; the comprehensive analysis module extracts the abnormal characteristics of the frequency spectrum interference strength and the temperature difference fluctuation characteristics, and evaluates the influence of electromagnetic interference on the sensor data accuracy; the interference division module classifies the interference events into instantaneous interference or continuous interference; the early warning processing module implements hierarchical processing according to the interference type, delays the trigger of the temperature control equipment adjustment for instantaneous interference, and sends a low-level warning notification, and further analyzes the continuous interference for hierarchical processing, and provides the corresponding adjustment strategy and warning information for the management personnel.

[0080] The above formulas are dimensionless values calculated, and the formulas are obtained by software simulation of a large amount of data to obtain a formula of the most recent real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0081] The above embodiments can be realized wholly or partially by software, hardware, firmware or any other combination. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are wholly or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transferred from one computer readable storage medium to another, for example, the computer instructions can be transferred from one website, computer, server or data center to another by wired (such as infrared, wireless, microwave, etc.) mode. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center and the like containing one or more available medium collections. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD) or a semiconductor medium. The semiconductor medium can be a solid state disk.

[0082] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A cargo transportation status monitoring system based on the Internet of Things, characterized by: It includes electromagnetic interference monitoring module, redundant sensor module, data processing and transmission module, comprehensive analysis module, interference classification module and early warning processing module; Electromagnetic interference monitoring module: Multiple electromagnetic interference monitors are installed in the cargo box to record electromagnetic signal strength data in several frequency bands in real time; Redundant sensor module: includes at least two temperature sensors, used to simultaneously collect temperature data inside the cargo box and calculate temperature deviation data between the temperature sensors over several time periods; Data processing and transmission module: used to pre-process electromagnetic signal strength data and temperature deviation data, and upload the processed data to the cloud platform through the cellular network; Comprehensive analysis module: This module extracts the abnormal spectrum interference intensity features from the pre-processed electromagnetic signal strength data and the temperature difference fluctuation features from the temperature deviation data. Based on these extracted abnormal spectrum interference intensity features and temperature difference fluctuation features, it evaluates the impact of electromagnetic interference on the sensor data accuracy. Specifically, the method includes: converting the spectrum interference intensity anomaly index and the temperature difference fluctuation index into a comprehensive feature vector, using the comprehensive feature vector as the input of the machine learning model, using the machine learning model to predict the impact degree value label of electromagnetic interference on sensor data accuracy for each set of comprehensive feature vectors as the prediction target, minimizing the sum of the prediction errors of the impact degree value labels of all electromagnetic interference on sensor data accuracy as the training target, training the machine learning model until the sum of the prediction errors reaches convergence, stopping the model training, and determining the impact degree value of electromagnetic interference on sensor data accuracy based on the model output results, wherein the machine learning model is a polynomial regression model; Interference classification module: used to classify interference events into transient interference and continuous interference according to the degree of impact of electromagnetic interference on sensor data accuracy; Early warning processing module: For instantaneous interference events, the module delays the triggering of temperature control equipment adjustments and sends a low-level early warning notification to management personnel. For continuous interference events, the module further analyzes the impact of electromagnetic interference on sensor data accuracy within a fixed time period and classifies continuous interference events based on the analysis results.

2. The cargo transportation status monitoring system based on the Internet of Things according to claim 1, characterized in that: In the comprehensive analysis module, the spectrum interference intensity anomaly characteristics in the extracted electromagnetic signal strength data are analyzed to generate a spectrum interference intensity anomaly index. The spectrum interference intensity anomaly index is obtained as follows: Assume that the spectrum signal data is X, with a dimension of n×m, where n is the number of samples and m is the number of features. is the signal strength of the i-th sample in the j-th frequency band. First, X is standardized and the expression is: Where, is the standardized data, is the mean value of the signal strength of the jth frequency band, is the standard deviation of the signal strength in the jth frequency band, and the standardized matrix Z is obtained. The covariance matrix C of the standardized matrix Z is calculated, and the expression is: ; Where C is the m×m covariance matrix, each element represents the linear correlation between frequency band j and frequency band k, T is the matrix transpose, and the eigenvalue decomposition of the covariance matrix C is obtained: Where, is the kth eigenvalue, which represents the variance contribution rate of the principal component. is the kth eigenvector, corresponding to the direction of the principal component, arrange the eigenvalues ​​in descending order, take the first p largest eigenvalues ​​and the corresponding eigenvectors to form the principal component subspace, and project the standardized data Z into the principal component space: ; Where Y is the principal component matrix after projection, with dimension n×p, is a matrix composed of the first p eigenvectors with a dimension of m×p. The original data is reconstructed using the projected principal component data Y: Where, The dimension of the reconstructed normalized data is n×m, and the reconstruction error of each sample is calculated: Where, is the reconstruction error of the i-th sample, according to the reconstruction error Calculate the spectrum interference intensity anomaly index, the expression is: Where, is the mean of the reconstruction errors of all samples, std(E) is the standard deviation of the reconstruction errors of all samples, and HKS is the spectrum interference intensity anomaly index.

3. The cargo transportation status monitoring system based on the Internet of Things according to claim 2, characterized in that: In the comprehensive analysis module, the temperature difference fluctuation characteristics in the extracted temperature deviation data are analyzed to generate a temperature difference fluctuation index. The method for obtaining the temperature difference fluctuation index is as follows: Assume that the temperature difference data is T(t), which is defined as a one-dimensional time series with a length of N and a sampling interval of Δt. ; Use Fourier transform to convert the time series T(t) into frequency domain representation F(f), the expression is: Where, Frequency The complex value of , j is the imaginary unit, is the kth frequency component, from Extracting the amplitude spectrum , the expression is: Where, for The real and imaginary parts of Indicates the contribution intensity of the frequency component to the temperature difference fluctuation, extracts the main frequency characteristics, and calculates the total spectrum energy of the temperature difference sequence , the expression is: ; Reflects the overall intensity of the entire temperature fluctuation and extracts the energy of the high-frequency components , the expression is: Where, is the threshold frequency of the high-frequency component, and calculates the proportion of high-frequency energy in the total energy , the expression is: Where, The proportion of violent fluctuations in the temperature difference data was quantified, and the temperature difference fluctuation index CHQ was calculated as the normalized form of the high-frequency energy proportion, and the expression is: ; Where GHQ is the temperature fluctuation index, 、 is the mean and standard deviation of the high-frequency energy ratio, which is used for normalization.

4. The cargo transportation status monitoring system based on the Internet of Things according to claim 1, characterized in that: The interference classification module is used to classify interference events into transient interference and continuous interference based on the degree of impact of electromagnetic interference on sensor data accuracy. Specifically: The obtained electromagnetic interference impact value on the sensor data accuracy is compared with the impact value reference threshold preset based on historical data. If the electromagnetic interference impact value on the sensor data accuracy is greater than or equal to the impact value reference threshold, it means that the electromagnetic interference has a high impact on the sensor data accuracy. In this case, a high interference degree signal is generated, and the interference event is classified as continuous interference. If the electromagnetic interference impact value on the sensor data accuracy is less than the impact value reference threshold, it means that the electromagnetic interference has a low impact on the sensor data accuracy. In this case, a low interference degree signal is generated, and the interference event is classified as instantaneous interference.

5. The cargo transportation status monitoring system based on the Internet of Things according to claim 1 is characterized in that: In the early warning processing module, for continuous interference events, the impact of electromagnetic interference on sensor data accuracy within a fixed time period is further analyzed, and continuous interference events are graded based on the analysis results. Specifically: For continuous interference events, that is, the impact degree value of the electromagnetic interference generated within a fixed time period on the sensor data accuracy is greater than or equal to the impact degree value reference threshold, the impact degree values ​​of the electromagnetic interference generated within a subsequent fixed time period that are greater than or equal to the impact degree value reference threshold on the sensor data accuracy are collected, and a corresponding data set is established. The mean and standard deviation of the data set are calculated and analyzed, and the continuous interference events are graded according to the analysis results.

6. The cargo transportation status monitoring system based on the Internet of Things according to claim 5, characterized in that: If the mean of the impact values ​​in the data set is greater than or equal to the reference threshold of the mean of the impact values, and the standard deviation of the impact values ​​is less than the reference threshold of the standard deviation of the impact values, the interference level is high and stable. At this time, a level 1 warning signal is generated, and the interference source needs to be checked first, and redundant data or protection mode should be enabled in the system. If the mean of the impact values ​​is greater than or equal to the reference threshold of the mean of the impact values, and the standard deviation of the impact values ​​is greater than or equal to the reference threshold of the standard deviation of the impact values, the interference level is high and the fluctuation is large. At this time, a secondary warning signal is generated to monitor the trend of interference fluctuations, delay system adjustment decisions, and optimize the sensor data processing algorithm; If the mean of the impact values ​​is less than the reference threshold of the mean of the impact values, and the standard deviation of the impact values ​​is greater than or equal to the reference threshold of the standard deviation of the impact values, the interference level is low but the fluctuation is large. In this case, a level 3 warning signal is generated and continued monitoring is required, but no immediate intervention is required. If the mean of the impact degree values ​​is less than the reference threshold of the mean of the impact degree values, and the standard deviation of the impact degree values ​​is less than the reference threshold of the standard deviation of the impact degree values, the interference level is low and the fluctuation is small. At this time, no early warning is required, the system maintains normal operation, and no additional adjustment measures are taken.

Images