Intelligent monitoring system and method for coal road transportation

Abstract

The invention relates to the technical field of multi-dimensional physical quantity measurement, and particularly discloses a coal road transportation intelligent monitoring system, which comprises the following modules: a multi-dimensional physical quantity acquisition module used for integrating physical quantity synchronous acquisition and data integration of various acquisition devices; the time sequence trend modeling module is used for building an environment physical quantity measurement data time sequence trend model, a carrier operation physical quantity measurement data time sequence trend model and a cargo state measurement data time sequence trend model; the multi-dimensional correlation analysis module is composed of a measurement data feature vector extraction sub-module and a physical quantity trend correlation matrix construction sub-module, strong correlation combination is screened through a matrix, and a cause chain of progressive anomaly is locked. And the sequential variation trend of the environment, carrier and cargo data in the whole transportation link is tracked, and the progressive anomaly is identified.

Description

A smart monitoring system and method for coal road transportation Technical Field

[0001] This application relates to the field of multidimensional physical quantity measurement technology, and specifically discloses an intelligent monitoring system and method for coal road transportation. Background Technology

[0002] In China's multi-industry production, transportation, and storage systems, accurate monitoring and early warning of anomalies are crucial for ensuring operational safety, reducing resource waste, and mitigating environmental risks. Whether it's the transportation of hazardous chemicals, the cold chain storage of fresh produce, or the management of industrial production processes, continuous tracking of physical quantities related to the environment, carriers or equipment, and goods or materials is essential. However, changes in physical quantities across various scenarios are easily affected by multiple factors, such as airflow disturbances, wind fluctuations, and road roughness during transportation; temperature and humidity changes and equipment vibrations during production; and fluctuations in environmental parameters during storage. These factors can easily lead to safety hazards, resource depletion, or environmental pollution due to the accumulation of abnormal physical quantities. The essence of these problems is that abnormal changes in multi-dimensional physical quantities across multiple scenarios are not captured and analyzed in a timely manner, causing risks to gradually escalate from their initial stages.

[0003] To alleviate this common problem across industries, existing technologies have implemented specialized monitoring measures for specific scenarios, such as dust suppression monitoring in some transportation scenarios and single-parameter monitoring in production scenarios. These measures aim to reduce specific risks through targeted technical means. However, significant limitations still exist in practice: on the one hand, the quality standards and implementation effects of specialized monitoring measures vary, and the monitoring accuracy in some scenarios is difficult to achieve the expected results, failing to fundamentally avoid risks; on the other hand, existing monitoring technologies generally lack unified physical quantity measurement standards and data processing systems. The measurement data dimensions of different types of physical quantities are inconsistent, and errors are not corrected, resulting in a lack of comparability and correlation of data, which seriously limits the comprehensiveness and accuracy of monitoring.

[0004] Existing related technical solutions, such as the railway coal transportation dust suppression information monitoring method and system (application number 2019111422673) and the railway coal transportation spillage monitoring system (application number 2019111435635), all suffer from the same core technical deficiency: they are limited to the measurement of specific variables in a single industry and scenario, focusing only on a certain type of physical quantity, such as a single image change or a specific environmental parameter, and lack the ability to perform generalized synchronous measurement and correlation analysis of multi-dimensional physical quantities of the environment, carrier, and cargo. The key problem this leads to is that it cannot capture the temporal trend of physical quantity changes and the inherent correlation between multi-dimensional physical quantities. The monitoring results can only reflect a single instantaneous state and output a single alarm signal, but it cannot identify gradual physical quantity anomalies where a single change does not exceed the standard but the cumulative trend is abnormal. Furthermore, it is difficult to trace the cause of the anomaly and cannot provide a basis for early prediction and precise handling for operational safety in multiple scenarios.

[0005] In addition, existing technologies are mostly designed for specific scenarios and lack general adaptability. Different industries and scenarios need to develop corresponding monitoring systems separately, resulting in low technology reuse rate, high development cost, and difficulty in meeting the common needs of physical quantity monitoring across scenarios.

[0006] To address this, the present invention provides an intelligent monitoring platform and implementation method for coal road transportation. This platform is not specific to any particular industry or type of variable. By integrating scalable, multi-dimensional physical quantity measurement equipment, it establishes a universal and standardized measurement and data processing workflow. It tracks the temporal change trends of three core physical quantities—environment, carrier, and cargo—across various scenarios and the entire supply chain, identifying progressive physical quantity anomalies. This ensures that the monitoring results output is no longer a simple alarm, but a comprehensive early warning report containing the causes of anomalies and trend predictions. This fundamentally solves the shortcomings of existing technologies in the universal measurement of multi-dimensional physical quantities and cross-scenario correlation analysis, adapting to the common monitoring needs of multiple industries such as chemical, fresh food, manufacturing, transportation, and storage. Summary of the Invention

[0007] The purpose of this invention is not to focus on a single state measurement at a certain moment, but to track the temporal change trends of multiple types of data throughout the entire transportation chain and identify progressive anomalies.

[0008] To achieve the above objectives, this invention provides the following basic solution: an intelligent monitoring system for coal road transportation, comprising the following modules: an equipment module, including an expandable physical quantity acquisition device group, which includes, but is not limited to, environmental physical quantity acquisition devices, carrier operation physical quantity acquisition devices, and cargo status acquisition devices; a multi-dimensional physical quantity acquisition module, used to integrate the synchronous acquisition and data integration of physical quantities from various acquisition devices, and supporting the expansion and addition of physical quantity acquisition; a measurement data preprocessing module, which cleans invalid physical quantity data based on general rules and standardizes and transforms invalid physical quantity data; and a time-series trend modeling module, which builds a generalized time-series trend model adaptable to different physical quantities. The system extracts time-series feature vectors, and generalized time-series trend models include, but are not limited to, time-series trend models for environmental physical quantity measurement data, time-series trend models for vehicle operation physical quantity measurement data, and time-series trend models for cargo status measurement data. A multi-dimensional correlation analysis module, consisting of a measurement data feature vector extraction submodule and a physical quantity trend correlation matrix construction submodule, constructs an m×n symmetric trend correlation matrix. Strongly correlated combinations are selected through matrix element filtering to pinpoint the causal chain of progressive anomalies. A tiered early warning and handling module, based on the progressive anomaly identification results, generates early warning information containing anomaly trend details, physical cause predictions, and handling suggestions, which is synchronized to the intelligent monitoring system and the vehicle driving terminal via a communication module.

[0009] Furthermore, it also includes a time-series measurement data storage module used in conjunction with the time-series trend modeling module: enabling the tagged archiving of historical measured physical quantity data, associating monitored objects with measured physical quantity data, and retaining at least six months of measured physical quantity data for model optimization in the time-series trend modeling module.

[0010] Furthermore, it also includes an intelligent management system module: integrating an image analysis and processing unit, an early warning display unit, and a disposal decision-making unit; the image analysis and processing unit connects to the time-series trend modeling module and the multi-dimensional correlation analysis module to complete measurement data calculation and logical judgment; the early warning display unit visualizes the trend change curves and correlation matrices of physical quantities in the form of charts and heat maps; the disposal decision-making unit stores historical early warning data and disposal results to provide decision support for transportation management.

[0011] Furthermore, the tiered early warning and response module classifies early warnings into three levels based on the severity of the progressive abnormal trends in physical quantities and the associated risk levels: Mild warning: A single time-series measurement data trend approaches the cumulative threshold of the physical quantity trend, while the other time-series measurement data have low strong correlation coefficients, prompting subsequent stations to increase the monitoring frequency of the corresponding physical quantity; Moderate warning: Two types of time-series measurement data have high strong correlation coefficients, and their respective physical quantity measurement data are approaching the cumulative threshold, prompting the station to stop at the nearest station ahead and conduct a simple inspection of the monitoring objects corresponding to the associated physical quantities; Severe warning: Two or more types of time-series measurement data have high strong correlation coefficients, and the cumulative change in physical quantities exceeds the threshold, immediately issuing an emergency warning, requiring the carrier to stop at the nearest station, and conducting a comprehensive inspection and handling of the environment, carrier, and cargo status corresponding to the relevant physical quantities.

[0012] Furthermore, the environmental physical quantity acquisition equipment includes a temperature and humidity sensor for measuring ambient temperature and humidity, a wind speed sensor for measuring wind speed, and a visibility detector for measuring visibility; the carrier operation physical quantity acquisition equipment includes an acceleration sensor for measuring carrier acceleration, a tilt sensor for measuring carrier tilt angle, and a braking status monitoring module for monitoring the carrier's braking physical state; the cargo status acquisition equipment is a high-definition image acquisition camera used to acquire images of the cargo's appearance and physical state; all environmental physical quantity acquisition equipment needs to be installed on each loading unit of the carrier, the cargo status acquisition equipment needs to clearly see the cargo, and the carrier needs to be labeled with a unique number, and the loading unit needs to be labeled with a unique number. The high-definition image acquisition camera simultaneously acquires digital data and label data, accurately binding the measured physical quantity data with the monitored object.

[0013] Furthermore, the m×n symmetric trend correlation matrix is ​​at least a 3×3 physical quantity trend correlation matrix. The multi-dimensional correlation analysis module uses the time-series feature vectors of environmental physical quantities, carrier operation physical quantities, and cargo status measurement data as analysis dimensions. It quantifies the trend correlation between pairs of dimensions through the Pearson correlation coefficient, and finally constructs a 3×3 physical quantity trend correlation matrix. All calculations are based on the output results of the time-series trend models of standardized measurement data and environmental physical quantity measurement data, the time-series trend models of carrier operation physical quantity measurement data, and the time-series trend models of cargo status measurement data.

[0014] Furthermore, the processing steps of the measurement data preprocessing module are as follows: Step A1: Full outlier removal is adopted, which includes: removing invalid data from the physical quantity data collected by the multi-dimensional physical quantity acquisition module; removing values ​​that exceed the reasonable measurement range of physical quantities due to sensor failure; removing instantaneous fluctuation measurement data caused by mechanical characteristics at the moment of carrier start-up and shutdown; and excluding invalid cargo state similarity measurement values ​​caused by light obstruction and lens dirt during image acquisition; Step A2: Establish a data standardization conversion model to convert the effective environmental physical quantity measurement data, effective carrier operation physical quantity measurement data, and effective cargo state measurement data into standardized measurement data of the same dimension to ensure the comparability of different types of physical quantity measurement data; Step A3: Based on the actual measurement timestamp of the carrier passing through each station, the environmental physical quantity measurement data, carrier operation physical quantity measurement data, and cargo state measurement data of the same train and the same loading unit are time-axis calibrated to ensure that the three types of effective measurement data of each monitoring node strictly correspond to the transportation physical state at the same time point.

[0015] Furthermore, it also includes monitoring stations and benchmark stations, which are arranged along the transportation route. The monitoring stations are spaced at equal intervals, and the intervals are adaptively adjusted. The benchmark stations are located at the loading points. After the cargo is loaded at the loading points, benchmark physical quantities are measured for each loading unit. The benchmark measurement data includes: a benchmark appearance image of the cargo, initial environmental physical parameters, and initial operating physical parameters of the carrier. All benchmark measurement data are bound to numbers, labels, and measurement timestamps and stored in the system benchmark database.

[0016] This application also discloses an intelligent monitoring method for coal road transportation, applied to an intelligent monitoring system for coal road transportation, comprising the following steps: Step S01: Taking the entire chain of coal transportation—loading point, stations along the route, and destination—as the axis, a benchmark acquisition station is set up at the loading point, and several monitoring stations are evenly distributed along the route according to the transportation mileage. The monitoring stations integrate environmental physical quantity acquisition equipment, carrier operation physical quantity acquisition equipment, and cargo status acquisition equipment. Through the above equipment, continuous physical quantity measurement of the entire chain is realized, and the time-series measurement data of the entire chain is obtained; Step S02: A four-level storage structure of vehicle number, loading unit, measurement data type, and measurement timestamp is constructed to generate a unique time-series measurement data chain for each loading unit. The benchmark measurement data at the loading point and the measurement data at each station along the route are concatenated in the order of transportation time to form a complete time-series measurement sequence from the benchmark measurement data to the destination measurement data; Step S03: The time-series measurement data of the entire chain is processed. Step S04: Optimize the time-series measurement data to convert it into standardized measurement data with a unified dimension; Step S05: Establish a time-series trend model for each type of standardized measurement data based on its physical characteristics, extract core feature parameters reflecting the changing patterns of physical quantities from the time-series trend model, and form a time-series feature vector for each type of measurement data; Step S06: Calculate the trend correlation of the time-series feature vector for each type of measurement data using a multi-dimensional correlation trend algorithm, and construct a physical quantity trend correlation matrix using the time-series feature vector for each type of measurement data as a three-dimensional dimension to reveal the inherent correlation of physical quantity measurement results in different dimensions; Step S07: Based on road transport safety standards, set dual thresholds for the entire-link time-series measurement data. The dual thresholds include a single change threshold for physical quantities and a cumulative trend threshold for physical quantities. Combined with the strong correlation coefficient of the time-series measurement data judged above, if it exceeds the range set by the dual thresholds, it is judged as a gradual anomaly in physical quantities.

[0017] Furthermore, in step S02, an independent data folder is created according to numbers, containing the measurement data of all loading units of the train; the data of each loading unit is classified and stored according to environmental physical quantity measurement data, carrier operation physical quantity measurement data, and cargo status measurement data, ultimately forming a complete time-series measurement sequence from baseline measurement data to endpoint measurement data.

[0018] The principle and effect of this solution are as follows: 1. Compared with existing technologies, existing coal transportation monitoring technologies are mostly limited to the measurement of a single physical quantity or the determination of instantaneous state. They cannot capture gradual risks that do not exceed the standard in a single instance but have an abnormal cumulative trend, such as slow wear on the surface of the cargo or gradual shift of the carrier's tilt angle. This system focuses on three core physical quantities: environment, carrier, and cargo. It achieves continuous measurement across the entire chain through multi-variable synchronous acquisition equipment, constructs a unified time-series monitoring system, and expands the measurement dimension from a single point instantaneous time to a multi-variable full-chain time-series. It accurately identifies gradual anomalies that are difficult to detect with traditional technologies, effectively avoids missed judgments and misjudgments, and fills the technical gap in multi-variable collaborative monitoring of gradual risks in coal transportation.

[0019] 2. Compared with existing technologies, even if existing technologies integrate multiple types of data, they are only used by simple superposition and do not form an effective correlation between multiple variables, making it difficult to pinpoint the root cause of the risk after the warning. This system is built on the core framework of multivariate measurement, extracts the time-series feature vectors of three types of physical quantities: environment, carrier, and goods, and realizes visualization of multivariate causal chains.

[0020] 3. Compared with existing technologies, this system significantly improves trend fitting accuracy and data reuse value: It emphasizes the different characteristics of environmental, carrier, and cargo data to ensure high trend fitting accuracy; at the same time, it constructs a unified standardization and storage architecture, enabling data to be reused even in difficult-to-reuse scenarios. This method not only ensures the feasibility of multi-dimensional data correlation analysis but also realizes the tagging, archiving, and reuse of historical data, providing data support for model iteration and optimization. Compared with the generalized processing methods of existing technologies, the fitting accuracy and data reuse rate are significantly improved.

[0021] 4. Compared with existing technologies, this system achieves a synergistic improvement in environmental and economic benefits, demonstrating significant industrial application value: By accurately identifying the risk of gradual spillage, this system can take preventative measures to avoid continuous coal spillage during transportation. Simultaneously, it reduces coal dust pollution of the atmosphere and soil along the transportation route, lowering environmental remediation costs. Furthermore, by accurately locating the causes of anomalies, it reduces unnecessary vehicle stops and inspections, improving transportation efficiency and reducing logistics costs. Its standardized data architecture and scalable algorithm logic can also adapt to different coal transportation routes, vehicle types, and climate conditions, possessing broad industrial promotion value and providing core technological support for the green and intelligent upgrading of coal road transportation.

[0022] 5. Compared with existing technologies, this system achieves synergistic improvement in benefits across multiple fields and has broad industrial application value: It is not limited to specific variables or a single industry. By accurately identifying progressive risks in various fields, such as hazardous material leakage risks in chemical transportation, quality degradation risks in fresh food transportation, and early signs of equipment failure in industrial production, it allows for proactive measures to prevent risks from escalating. At the environmental level, it can reduce pollutant emissions and resource waste during production and transportation in various industries, lowering environmental governance costs. At the economic level, by accurately locating the causes of anomalies, it reduces unnecessary shutdowns for inspections and handling in various fields, improving operational efficiency and reducing overall costs. Its standardized data architecture and scalable algorithm logic can be adapted to transportation, production, and storage scenarios in different industries, possessing strong cross-domain compatibility. It provides core technical support for the green and intelligent upgrading of multiple industries, including coal, chemicals, fresh food, and manufacturing, demonstrating significant industrial application value. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 shows a schematic diagram of the composition of an intelligent monitoring system for coal road transportation proposed in an embodiment of this application; Figure 2 shows a flowchart of an intelligent monitoring method for coal road transportation proposed in an embodiment of this application. Detailed Implementation

[0025] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0026] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, or it may be an electrical, mechanical, or other form of connection.

[0027] The embodiments are shown in Figures 1 and 2: This invention discloses an intelligent monitoring system for coal road transportation, including the following modules: Equipment module: including an expandable physical quantity acquisition device group, which includes, but is not limited to, environmental physical quantity acquisition devices, carrier operation physical quantity acquisition devices, and cargo status acquisition devices; Multi-dimensional physical quantity acquisition module: used to integrate the synchronous acquisition and data integration of physical quantities from various acquisition devices, and supports the expansion and addition of physical quantity acquisition.

[0028] Specifically: the environmental physical quantity acquisition equipment includes a temperature and humidity sensor for measuring ambient temperature and humidity, a wind speed sensor for measuring wind speed, and a visibility detector for measuring visibility. The carrier operation physical quantity acquisition equipment includes an acceleration sensor for measuring carrier acceleration, a tilt sensor for measuring carrier tilt angle, and a braking status monitoring module for monitoring the carrier braking physical state. The cargo status acquisition equipment is a high-definition image acquisition camera used to acquire images of the cargo's appearance and physical state. All environmental physical quantity acquisition equipment needs to be installed on each loading unit of the coal transport carrier. The cargo status acquisition equipment needs to clearly see the coal, and the carrier needs to be labeled with a unique number, and the loading unit needs to be labeled with a unique number. The high-definition image acquisition camera simultaneously acquires digital data and label data, accurately binding the measured physical quantity data with the monitored object.

[0029] It also includes monitoring stations and benchmark stations, which are arranged along the transportation route. The monitoring stations are spaced at equal intervals and the intervals are adaptively adjusted. The benchmark stations are located at the loading points. After the cargo is loaded at the loading points, benchmark physical quantities are measured for each loading unit. The benchmark measurement data includes: benchmark appearance image of the cargo, initial environmental physical parameters, and initial operating physical parameters of the carrier. All benchmark measurement data are bound to numbers, labels, and measurement timestamps and stored in the system benchmark database.

[0030] For ease of understanding, the full-link timing sequence generated in this embodiment is as follows: full-link timing sequence of reference data - data from site 1 - data from site 2 - end point data. All of these data cover the data obtained by the above-mentioned devices.

[0031] Clearly, not all of these data are valid; errors and related factors cause data distortion. Therefore, a measurement data preprocessing module was set up to process the time-series data across the entire chain and convert it into standardized data of a unified dimension.

[0032] Measurement data preprocessing module: Cleans invalid physical quantity data of measurements based on general rules, and standardizes and transforms invalid physical quantity data of measurements.

[0033] The processing steps of the measurement data preprocessing module are as follows: Step A1: Use full outlier removal. Full outlier removal includes: removing invalid data from the physical quantity data collected by the multi-dimensional physical quantity acquisition module; removing values ​​that exceed the reasonable measurement range of physical quantities due to sensor failure; removing instantaneous fluctuation measurement data caused by mechanical characteristics at the moment of carrier start-up and shutdown; and excluding invalid cargo state similarity measurement values ​​caused by light obstruction and lens dirt during image acquisition.

[0034] In step A1, for ease of understanding, this embodiment lists relevant invalid data, such as abnormal data with wind speed > 20m / s and no corresponding image evidence, and sudden acceleration values ​​at startup. These data are invalid and need to be removed. After removal, the remaining data are valid data, and proceed to step A2.

[0035] Step A2: Establish a data standardization conversion model to convert effective environmental physical quantity measurement data, effective carrier operation physical quantity measurement data, and effective cargo status measurement data into standardized measurement data of a unified dimension, ensuring the comparability of different types of physical quantity measurement data.

[0036] Specifically, the following steps are taken: Step A21: The standardization transformation of effective environmental physical quantity measurement data, effective carrier operation physical quantity measurement data, and effective cargo status measurement data is based on the linear normalization method, which refers to the Min-Max normalization method.

[0037] Step A22: The raw data of effective environmental physical quantity measurement data, effective carrier operation physical quantity measurement data, and effective cargo status measurement data are uniformly mapped to the [0,1] interval.

[0038] Step A23: Construct a normalized data standardization transformation model, with the following formula: ;in, The standardized value, ranging from [0,1], where X is the valid data value after cleaning in step A1. This is the reasonable maximum value for this type of data in the context of coal road transportation. Data exceeding this reasonable maximum value is considered high-risk and triggers a shutdown threshold. This is the reasonable minimum value for this type of data in the context of coal road transportation. This reasonable minimum value is determined based on a combination of industry standards and historical monitoring data.

[0039] For example, environmental physical quantity data includes three core indicators: wind speed, relative humidity, and visibility, which are combined with safety threshold settings for coal transportation scenarios. / The specific conversion is as follows: Taking wind speed as an example: It is a windless state. (Strong wind threshold for coal transportation; operation must be stopped if the wind speed exceeds 15 m / s). Let X be the original wind speed, assumed to be 7.5 m / s. According to the formula: =0.5; Let X be the original fraction, assuming it is 12 m / s. According to the formula, =0.8, at this point the wind speed data has been standardized, and the original data has been uniformly mapped to the [0,1] interval.

[0040] The conversion of relative humidity and visibility is the same as that of wind speed, and will not be repeated in this embodiment. The physical quantity data of the carrier includes three categories: braking frequency, driving tilt angle, and acceleration fluctuation value. The threshold is set based on the safety operation standard of coal transport carrier. The conversion of braking frequency, driving tilt angle, and acceleration fluctuation value is the same as that of wind speed, and will not be repeated in this embodiment. The cargo data is mainly the image deviation angle, which is uniformly mapped to the [0,1] interval through the above conversion formula.

[0041] Step A3: Based on the actual measurement timestamps of the carrier passing through each station, perform time axis calibration on the environmental physical quantity measurement data, carrier operation physical quantity measurement data, and cargo status measurement data for the same train and the same loading unit, so that the three types of valid measurement data of each monitoring node strictly correspond to the transportation physical state at the same time point.

[0042] The time-series trend modeling module is used to build time-series trend models for environmental physical quantity measurement data, carrier operation physical quantity measurement data, and cargo status measurement data. It extracts core feature parameters from the time-series trend models to form time-series feature vectors for each type of data. Regarding the effective environmental data time-series trend model: environmental parameters include wind speed, humidity, and visibility. In the entire coal road transportation chain, these parameters exhibit approximately linear changes due to regional climate and terrain influences, such as the increasing wind speed along the transportation route from plains to mountains. The linear regression algorithm can accurately fit this monotonic trend and directly quantify the direction and rate of change through the slope.

[0043] Regarding the time-series trend model for environmental physical quantity measurement data: The model uses linear regression as its core algorithm, and the model is as follows: ,in, Trend slope >0 indicates that the parameter continues to increase. <0 indicates that the parameter continues to decrease. Site order : Represents the fitted value of the environmental parameter for the i-th site. Trendline intercept, corresponding to The fitted value of =0; where, ;in, ;in, ;in, ; : The standardized environmental parameter value of the i-th site, with a value range of [0,1]; The mean of the site order. : The mean value of standardized environmental parameters.

[0044] This embodiment still uses wind speed as an example, and humidity and visibility are the same as wind speed.

[0045] Assume the standardized environmental wind speeds at four stations: 0, 1, 2, and 3. The values ​​are 0.2, 0.4, 0.5, and 0.7 respectively. The mean of the site order is known. and Given, we obtain , A value greater than or equal to 0 indicates that the standardized wind speed increases by an average of 0.15 for each station visited, meaning that the wind speed shows a continuous increasing trend.

[0046] Regarding the time-series trend model of physical quantity measurement data of the vehicle operation: The vehicle operation parameters include braking frequency, driving tilt angle and acceleration fluctuation value. They are affected by road conditions such as sudden congestion and changes in road slope. They have trends but are accompanied by random fluctuations. For example, the braking frequency gradually increases due to road congestion, but individual stations may fluctuate due to short-term smooth traffic. The quadratic exponential smoothing method eliminates random interference through double smoothing.

[0047] Specifically: The core model is based on quadratic exponential smoothing, and the model is as follows: ;in, Based on the data from the i-th site, predict the carrier parameter values ​​for the m-th subsequent site, where m ≥ 1. : The smoothed value of the i-th station, : The rate of change of the trend at the i-th station, where 'a' is a smoothing coefficient, ranging from 0.1 to 0.3. For standardized braking frequency.

[0048] Then, firstly according to ; This is the first smoothing value, used to eliminate random fluctuations; then according to... ; This is the second smoothing value, used to capture the trend, and then based on... and get and The value of .

[0049] Standardization Taking braking frequency as an example, suppose there are four stations, and the braking frequency of the four stations is... All of this can be obtained because the initial data for the loading points along the transportation chain is clear and can be set. and that is All of the four stations and Given that 'a' is a smoothing coefficient with a value of 0.2, 0.2 is suitable for the moderate fluctuation characteristics of the carrier data, and thus we can obtain... The value, assuming The rate of increase is showing an increasing trend, and it can be predicted that the braking frequency of subsequent stations will continue to accelerate, requiring early warning.

[0050] Regarding the time-series trend model of cargo status measurement data: The changes in cargo data, including image similarity values ​​and cover integrity coefficients, are mostly non-linear decays. For example, in the early stage, slight shaking causes the similarity to decrease slowly, and in the later stage, the decay increases due to damage to the cover. Quadratic polynomials can accurately fit this non-linear trend of slow first and then fast or fast first and then slow, which is more in line with the actual transportation scenario than linear models.

[0051] Therefore, the time series trend model for effective cargo data is set as a quadratic polynomial as the core model: Where a, b, and c are polynomial coefficients, a > 0 indicates a decrease in decay, and a < 0 indicates an increase in decay. : The fitted values ​​of the cargo parameters at station t. Indicates the order of the stations.

[0052] Specifically: Then set the core parameters: , Instantaneous decay rate Once the core parameters of the cargo at station t and the instantaneous decay rate are known, the instantaneous decay rate of each station can be determined. For example, the instantaneous decay rate of station 2 can be compared with that of station 1, and the instantaneous decay rate of station 3 can be compared with that of station 1, thus obtaining a conclusion that the decay acceleration is obvious or not obvious.

[0053] Multi-dimensional correlation analysis module: It consists of a measurement data feature vector extraction submodule and a physical quantity trend correlation matrix construction submodule. It constructs an m×n symmetric trend correlation matrix, and filters strong correlation combinations through matrix elements to lock the cause chain of progressive anomalies.

[0054] Specifically: the m×n symmetric trend correlation matrix is ​​at least a 3×3 physical quantity trend correlation matrix. The multi-dimensional correlation analysis module uses the time-series feature vectors of environmental physical quantities, carrier operation physical quantities, and cargo status measurement data as analysis dimensions. It quantifies the trend correlation between pairs of dimensions through the Pearson correlation coefficient, and finally constructs a 3×3 physical quantity trend correlation matrix. All calculations are based on the output results of the time-series trend models of standardized measurement data and environmental physical quantity measurement data, the time-series trend models of carrier operation physical quantity measurement data, and the time-series trend models of cargo status measurement data.

[0055] The specific steps are as follows: B051: Extract the time series feature vectors of the time series trend models of the physical quantity measurement data of the carrier operation, the physical quantity measurement data of the carrier operation, and the cargo status measurement data. According to the above, the time series feature vectors need to be ≥3; Environment: .

[0056] Goods: .

[0057] Carrier: .

[0058] Each row corresponds to a trend feature of a site, and there are a total of three trend features, corresponding to all sites in the entire chain.

[0059] B052: Calculate the trend correlation r(X,Y) between two types of time series feature vectors using the Pearson correlation coefficient; ; where vector The eigenvalue of the i-th row and k-th column, such as E. These are the feature values ​​in the i-th row and k-th column of V or G; k=1, 2, 3: corresponding to the three core dimensions of the feature vector. or The mean of the k-th column features of vector X or Y. The final trend correlation coefficient between pairwise correlations, | The larger the value of |, the stronger the correlation. >0 indicates a positive correlation, meaning that when X increases, Y increases simultaneously; r<0 indicates a negative correlation, meaning that when X increases, Y decreases simultaneously.

[0060] B053: Construct a symmetrical 3×3 trend correlation matrix, filter strongly correlated combinations through matrix elements, and lock the causal chain of progressive anomalies.

[0061] Specifically: Establish correlation strength grading standards: 0.8 ≤ |r| ≤ 1.0 indicates strong correlation, with highly synchronized trends in both data types, such as a strong positive correlation between rising wind speed and cargo depletion. 0.5 ≤ |r| < 0.8 indicates moderate correlation, with moderately synchronized trends in both data types, such as a moderate correlation between increased braking frequency and cargo depletion. 0.3 ≤ |r| < 0.5 indicates weak correlation, with weakly synchronized trends in both data types, such as a weak correlation between rising humidity and braking frequency. |r| < 0.3 indicates no correlation, with no significant synchronicity between the trends in both data types.

[0062] Tiered early warning and response module: Based on the progressive anomaly identification results of physical quantities, it generates early warning information containing details of abnormal trends, prediction of physical causes, and response suggestions, and synchronizes it to the intelligent monitoring system and the vehicle driving terminal through the communication module.

[0063] The graded early warning and response module divides early warnings into three levels based on the severity of the progressive abnormal trend of physical quantities and the associated risk level.

[0064] Mild warning: The trend of a single time series measurement data is close to the cumulative threshold of the physical quantity trend, while the other time series measurement data have low strong correlation coefficients, suggesting that subsequent sites should increase the monitoring frequency of the corresponding physical quantities.

[0065] Moderate warning: The strong correlation coefficient between the two types of time series measurement data is high and their respective physical quantity measurement data are close to the cumulative threshold, indicating that the device should stop at the nearest station ahead for simple inspection of the monitoring objects corresponding to the associated physical quantities.

[0066] Severe warning: When the strong correlation coefficient of two or more types of time-series measurement data is high and the cumulative change of physical quantities exceeds the threshold, an emergency warning is immediately issued, requiring the carrier to dock at the nearest station and to conduct a comprehensive inspection and handling of the environment, carrier, and cargo status corresponding to the relevant physical quantities.

[0067] It also includes a time series measurement data storage module used in conjunction with the time series trend modeling module: enabling the tagged archiving of historical measured physical quantity data, associating monitored objects with measured physical quantity data, and retaining at least six months of measured physical quantity data for model optimization in the time series trend modeling module.

[0068] It also includes an intelligent management system module: integrating an image analysis and processing unit, an early warning display unit, and a disposal decision-making unit; the image analysis and processing unit connects to the time-series trend modeling module and the multi-dimensional correlation analysis module to complete measurement data calculation and logical judgment; the early warning display unit visualizes the trend change curves and correlation matrices of physical quantities in the form of charts and heat maps; the disposal decision-making unit stores historical early warning data and disposal results to provide decision support for transportation management.

[0069] This application also discloses an intelligent monitoring method for coal road transportation, applied to an intelligent monitoring system for coal road transportation, including the following steps: Step S01: Taking the entire chain of coal transportation, from the loading point, along the route, to the destination, as the axis, a benchmark acquisition station is set up at the loading point, and several monitoring stations are evenly distributed along the route according to the transportation mileage. The monitoring stations integrate environmental physical quantity acquisition equipment, carrier operation physical quantity acquisition equipment, and cargo status acquisition equipment. Through the above equipment, continuous physical quantity measurement of the entire chain is realized, and the time-series measurement data of the entire chain is obtained.

[0070] Step S02: Construct a four-level storage structure consisting of train number, loading unit, measurement data type, and measurement timestamp. Generate a unique time-series measurement data chain for each loading unit. Concatenate the baseline measurement data at the loading point with the measurement data at each station along the route in chronological order of transportation time to form a complete time-series measurement sequence from the baseline measurement data to the destination measurement data. In step S02, create independent data folders by number, containing the measurement data of all loading units for that train number. The data of each loading unit is classified and stored according to environmental physical quantity measurement data, carrier operation physical quantity measurement data, and cargo status measurement data, ultimately forming a complete time-series measurement sequence from the baseline measurement data to the destination measurement data.

[0071] Step S03: Optimize the end-to-end time-series measurement data and convert it into standardized measurement data of a unified dimension.

[0072] Step S04: Establish time series trend models for the physical characteristics of standardized measurement data, extract core feature parameters reflecting the laws of change of physical quantities from the time series trend models, and form time series feature vectors for each type of measurement data.

[0073] Step S05: Calculate the trend correlation of the time series feature vectors of each type of measurement data using a multi-dimensional correlation trend algorithm. Construct a physical quantity trend correlation matrix using the time series feature vectors of each type of measurement data as a three-dimensional dimension to reveal the intrinsic correlation of physical quantity measurement results in different dimensions.

[0074] Step S06: Based on road transport safety standards, set dual thresholds for the time-series measurement data across the entire chain. The dual thresholds include a threshold for a single change in physical quantities and a threshold for the cumulative trend of physical quantities. If the strong correlation coefficient of the time-series measurement data determined above exceeds the range set by the dual thresholds, it is judged as a gradual anomaly in physical quantities.

[0075] The above embodiments are only illustrated using coal transportation as an example. This system and method are not specific to any particular variable or a single industry. By accurately identifying progressive risks in various fields, such as the potential for hazardous material leakage in chemical transportation, the risk of quality degradation in fresh food transportation, and the precursors of equipment failure in industrial production, measures can be taken in advance to prevent the risks from escalating. In terms of environmental protection, it can reduce pollutant emissions and resource waste in the production and transportation processes of various industries, thereby reducing environmental governance costs. In terms of economics, by accurately locating the causes of anomalies, it can reduce unnecessary shutdowns for inspection and handling in various fields, improve operational efficiency, and reduce overall costs. Its standardized data architecture and scalable algorithm logic can be adapted to transportation, production, and storage scenarios in different industries, possessing strong cross-domain compatibility. It provides core technical support for the green and intelligent upgrading of multiple industries such as coal, chemicals, fresh food, and manufacturing, and has significant industrial promotion value.

[0076] For example, in the case of potential leaks of hazardous materials during chemical transportation, the three major models can be adjusted based on industry standards to make the system applicable to the presence of potential leaks of hazardous materials during chemical transportation.

[0077] For example, in the risk of quality degradation during fresh food transportation, the three major models can be adjusted based on industry standards, thereby making this system applicable to whether there is a risk of quality degradation during fresh food transportation.

[0078] The purpose of this invention is not to focus on a single state measurement at a certain moment, but to track the temporal change trends of multiple types of data throughout the entire transportation chain and identify progressive anomalies.

[0079] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A smart monitoring system for coal road transportation, characterized in that, It includes the following modules: Equipment module: including an expandable physical quantity acquisition device group, which includes, but is not limited to, environmental physical quantity acquisition devices, carrier operation physical quantity acquisition devices, and cargo status acquisition devices; Multi-dimensional physical quantity acquisition module: used to integrate the synchronous acquisition and data integration of physical quantities from various acquisition devices, and supports the expansion and addition of physical quantity acquisition; Measurement data preprocessing module: cleans invalid physical quantity data based on general rules and standardizes and transforms invalid physical quantity data. The time series trend modeling module builds a generalized time series trend model, which can be adapted to extract time series feature vectors based on different physical quantity characteristics. The generalized time series trend model includes, but is not limited to, time series trend models for environmental physical quantity measurement data, time series trend models for carrier operation physical quantity measurement data, and time series trend models for cargo status measurement data. Multi-dimensional correlation analysis module: It consists of a measurement data feature vector extraction submodule and a physical quantity trend correlation matrix construction submodule. It constructs an m×n symmetric trend correlation matrix, filters strong correlation combinations through matrix elements, and locks the cause chain of progressive anomalies. Tiered early warning and response module: Based on the progressive anomaly identification results of physical quantities, it generates early warning information containing details of abnormal trends, prediction of physical causes, and response suggestions, and synchronizes it to the intelligent monitoring system and the vehicle driving terminal through the communication module.

2. The intelligent monitoring system for coal road transportation according to claim 1, characterized in that, It also includes a time series measurement data storage module used in conjunction with the time series trend modeling module: enabling the tagged archiving of historical measured physical quantity data, associating monitored objects with measured physical quantity data, and retaining at least six months of measured physical quantity data for model optimization in the time series trend modeling module.

3. The intelligent monitoring system for coal road transportation according to claim 2, characterized in that, It also includes an intelligent management system module: integrating an image analysis and processing unit, an early warning display unit, and a disposal decision-making unit; the image analysis and processing unit connects to the time-series trend modeling module and the multi-dimensional correlation analysis module to complete measurement data calculation and logical judgment; the early warning display unit visualizes the trend change curves and correlation matrices of physical quantities in the form of charts and heat maps; the disposal decision-making unit stores historical early warning data and disposal results to provide decision support for transportation management.

4. The intelligent monitoring system for coal road transportation according to claim 3, characterized in that, The graded early warning and handling module divides the early warning into three levels based on the severity of the progressive abnormal trend of physical quantities and the associated risk level: mild warning: the trend of a single time series measurement data is close to the cumulative threshold of the physical quantity trend, and the other time series measurement data have a low strong correlation coefficient, which suggests that subsequent sites should increase the monitoring frequency of the corresponding physical quantities. Moderate warning: When the strong correlation coefficient between two types of time-series measurement data is high and the physical quantity measurement data of each is close to the cumulative threshold, it is suggested to stop at the nearest station ahead and conduct a simple inspection of the monitoring objects corresponding to the related physical quantities; Severe warning: When the strong correlation coefficient between two or more types of time-series measurement data is high and the cumulative change of physical quantities exceeds the threshold, an emergency warning is immediately issued, requiring the carrier to stop at the nearest station and conduct a comprehensive inspection and handling of the environment, carrier, and cargo status corresponding to the relevant physical quantities.

5. The intelligent monitoring system for coal road transportation according to claim 4, characterized in that, The environmental physical quantity acquisition equipment includes a temperature and humidity sensor for measuring ambient temperature and humidity, a wind speed sensor for measuring wind speed, and a visibility detector for measuring visibility. The carrier operation physical quantity acquisition equipment includes an acceleration sensor for measuring carrier acceleration, a tilt sensor for measuring carrier tilt angle, and a braking status monitoring module for monitoring the carrier's braking physical state. The cargo status acquisition equipment is a high-definition image acquisition camera used to acquire images of the cargo's appearance and physical state. All environmental physical quantity acquisition equipment needs to be installed on each loading unit of the carrier. The cargo status acquisition equipment needs to clearly see the cargo, and the carrier needs to be labeled with a unique number, and the loading unit needs to be labeled with a unique number. The high-definition image acquisition camera simultaneously acquires digital data and label data, accurately binding the measured physical quantity data with the monitored object.

6. The intelligent monitoring system for coal road transportation according to claim 5, characterized in that, The m×n symmetric trend correlation matrix is ​​at least a 3×3 physical quantity trend correlation matrix. The multi-dimensional correlation analysis module uses the time-series feature vectors of environmental physical quantities, carrier operation physical quantities, and cargo status measurement data as analysis dimensions. It quantifies the trend correlation between pairs of dimensions through the Pearson correlation coefficient and finally constructs a 3×3 physical quantity trend correlation matrix. All calculations are based on the output results of the time-series trend models of standardized measurement data and environmental physical quantity measurement data, the time-series trend models of carrier operation physical quantity measurement data, and the time-series trend models of cargo status measurement data.

7. The intelligent monitoring system for coal road transportation according to claim 6, characterized in that, The processing steps of the measurement data preprocessing module are as follows: Step A1: Full outlier removal is adopted, which includes: removing invalid data from the physical quantity data collected by the multi-dimensional physical quantity acquisition module; removing values ​​that exceed the reasonable measurement range of physical quantities due to sensor failure; removing instantaneous fluctuation measurement data caused by mechanical characteristics at the moment of carrier start-up and shutdown; and excluding invalid cargo state similarity measurement values ​​caused by light obstruction and lens dirt during image acquisition; Step A2: Establish a data standardization conversion model to convert the effective environmental physical quantity measurement data, effective carrier operation physical quantity measurement data, and effective cargo state measurement data into standardized measurement data of the same dimension to ensure the comparability of different types of physical quantity measurement data; Step A3: Based on the actual measurement timestamp of the carrier passing through each station, perform time axis calibration on the environmental physical quantity measurement data, carrier operation physical quantity measurement data, and cargo state measurement data of the same train and the same loading unit, so that the three types of effective measurement data of each monitoring node strictly correspond to the transportation physical state at the same time point.

8. The intelligent monitoring system for coal road transportation according to claim 7, characterized in that, It also includes monitoring stations and benchmark stations, which are arranged along the transportation route. The monitoring stations are spaced at equal intervals and the intervals are adaptively adjusted. The benchmark stations are located at the loading points. After the cargo is loaded at the loading points, benchmark physical quantities are measured for each loading unit. The benchmark measurement data includes: benchmark appearance image of the cargo, initial environmental physical parameters, and initial operating physical parameters of the carrier. All benchmark measurement data are bound to numbers, labels, and measurement timestamps and stored in the system benchmark database.

9. A method for intelligent monitoring of coal road transportation, applied to an intelligent monitoring system for coal road transportation as described in any one of claims 8, characterized in that, The process includes the following steps: Step S01: Taking the entire coal transportation chain—from the loading point, along the route, to the destination—as the axis, a benchmark data acquisition station is set up at the loading point. Several monitoring stations are evenly distributed along the route according to the transportation mileage. These monitoring stations integrate environmental physical quantity acquisition equipment, carrier operation physical quantity acquisition equipment, and cargo status acquisition equipment. Through these devices, continuous physical quantity measurement is achieved throughout the entire chain, resulting in full-chain time-series measurement data. Step S02: A four-level storage structure is constructed, consisting of vehicle number, loading unit, measurement data type, and measurement timestamp. A unique time-series measurement data chain is generated for each loading unit. The benchmark measurement data at the loading point and the measurement data at each station along the route are concatenated in chronological order to form a complete time-series measurement sequence from the benchmark measurement data to the destination measurement data. Step S03: The full-chain time-series measurement data is optimized and processed to convert the time-series measurement data into... Standardized measurement data with unified dimensions; Step S04: Establish time-series trend models for the physical characteristics of the standardized measurement data, extract core feature parameters reflecting the change law of physical quantities from the time-series trend models, and form time-series feature vectors for each type of measurement data; Step S05: Calculate the trend correlation of the time-series feature vectors of each type of measurement data through a multi-dimensional correlation trend algorithm, construct a physical quantity trend correlation matrix with the time-series feature vectors of each type of measurement data as a three-dimensional dimension, and reveal the inherent correlation of physical quantity measurement results in different dimensions; Step S06: Based on road transport safety standards, set dual thresholds for the time-series measurement data of the entire link. The dual thresholds include a single change threshold for physical quantities and a cumulative trend threshold for physical quantities. Combined with the strong correlation coefficient of the time-series measurement data judged above, if it exceeds the range set by the dual thresholds, it is judged as a gradual anomaly of physical quantities.

10. The intelligent monitoring method for coal road transportation according to claim 9, characterized in that, In step S02, an independent data folder is created according to the numbers, containing the measurement data of all loading units of the train. The data of each loading unit is classified and stored according to environmental physical quantity measurement data, carrier operation physical quantity measurement data, and cargo status measurement data, and finally a complete time-series measurement sequence from the baseline measurement data to the endpoint measurement data is formed.

Images