Mining truck driving risk dynamic early warning method, system, equipment and medium

Abstract

The embodiment of the invention relates to the field of open pit coal mine transportation safety, and discloses a dynamic early warning method, system and device for the driving risk of a mining truck and a medium. The dynamic early warning method for the driving risk of the mining truck comprises the steps that driving information of a vehicle is collected in real time based on a vehicle-mounted positioning system, the driving information comprises driving time, a geographic position, a driving state and topographic features; extracting key risk features for reflecting driving risks based on the driving information, wherein the key risk features include dangerous driving behavior features and terrain-related features; and constructing a risk prediction model based on the key risk features so as to carry out classified early warning on the driving risk of the vehicle. According to the technical scheme disclosed by the invention, the limitation of a single data dimension is broken through, the evaluation comprehensiveness is improved, and real-time perception, multi-dimensional analysis and accurate early warning of the driving risk are realized.

Description

Technical Field

[0001] This application relates to the field of open-pit coal mine transportation safety, and in particular to a method, system, equipment and medium for dynamic early warning of driving risks of mining trucks. Background Technology

[0002] Existing monitoring of driving behavior of open-pit coal mine transport trucks relies on manual inspections and simple onboard equipment, which cannot analyze driving behavior and predict potential risks in real time and comprehensively. At the same time, there is a lack of systematic methods to deeply mine vehicle driving data (such as location data), making it difficult to achieve accurate analysis of driving behavior and effective risk warning. Summary of the Invention

[0003] The purpose of this application is to provide at least one method, system, device and medium for dynamic early warning of driving risks of mining trucks, which can at least realize dynamic classification and early warning of driving risks in special environments of mining areas, timely detect dangerous driving behaviors and effectively reduce accident risks.

[0004] To address the aforementioned technical problems, at least one embodiment of this application provides a dynamic early warning method for driving risks of mining trucks, comprising: The vehicle's driving information is collected in real time based on the vehicle positioning system. The driving information includes driving time, geographical location, driving status and terrain features. Based on the driving information, key risk features are extracted to reflect driving risks. These key risk features include dangerous driving behavior features and terrain-related features. A risk prediction model is constructed based on the key risk characteristics to classify and warn of the driving risks of the vehicle.

[0005] In some optional embodiments, key risk features reflecting driving risk are extracted based on the driving information, including: The driving information is preprocessed to retain valid driving information; the valid driving information includes: timestamp, longitude, latitude, driving speed, heading angle, and altitude; The dangerous driving behavior characteristics and the terrain-related characteristics are calculated based on effective driving information; the dangerous driving behavior characteristics include: number of rapid accelerations, number of rapid decelerations, number of speedings, and number of sharp turns; the terrain-related characteristics include: mean altitude, mean slope, and standard deviation of slope.

[0006] In some alternative embodiments, the preprocessing includes: Remove adjacent duplicate points with the same longitude and latitude; and / or, Remove any outliers where the driving speed is negative, missing, or outside a reasonable range.

[0007] In some optional embodiments, the dangerous driving behavior characteristics are calculated based on valid driving information, including: The driving characteristics of the vehicle are calculated based on effective driving information, and the driving characteristics include at least: acceleration, rate of change of heading angle, horizontal displacement, gradient and altitude; Based on the aforementioned driving characteristics, dangerous driving behaviors are identified, including: rapid acceleration, rapid deceleration, speeding, and sharp turns. The number of dangerous driving behaviors is counted based on the aforementioned dangerous driving behaviors, and this count is used as a characteristic of dangerous driving behaviors.

[0008] In some alternative embodiments, identifying dangerous driving behavior based on the driving characteristics includes: The criteria for identifying dangerous driving behaviors are as follows: When the acceleration is greater than 3 m / s², it is identified as rapid acceleration; When the acceleration is less than -3 m / s², it is identified as rapid deceleration; When the driving speed in a straight section is greater than 40 km / h, or the driving speed in a curved section is greater than 25 km / h, it is considered speeding. When the rate of change of the heading angle is greater than 5° / s and the speed is greater than 9km / h, it is identified as a sharp turn.

[0009] In some optional embodiments, a risk prediction model is constructed based on the key risk characteristics to classify and warn of the driving risks of the vehicle, including: A risk feature matrix is ​​constructed based on the dangerous driving behavior characteristics and the terrain-related features; The risk feature matrix is ​​used as input to a risk prediction model to predict the driving risk of the vehicle and assign risk level labels; the risk prediction model includes a random forest model.

[0010] In some alternative embodiments, it also includes: The system can visualize the classification and warning results of the driving risks of the vehicle; and / or provide real-time alerts for vehicles with high risk levels.

[0011] At least one embodiment of this application also provides a dynamic early warning system for driving risks of mining trucks, including: The data acquisition module is used to collect vehicle driving information in real time based on the vehicle positioning system. The driving information includes driving time, geographical location, driving status and terrain. The extraction module is used to extract key risk features reflecting driving risks based on the driving information. The key risk features include dangerous driving behavior features and terrain-related features. The early warning module is used to construct a risk prediction model based on the key risk characteristics in order to classify and warn about the driving risks of the vehicle.

[0012] At least one embodiment of this application also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the above-described dynamic early warning method for driving risks of mining trucks.

[0013] At least one embodiment of this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described dynamic early warning method for driving risks of mining trucks.

[0014] The embodiments of this application provide a dynamic early warning method, system, device, and medium for driving risks of mining trucks. By collecting vehicle driving information in real time through an on-board positioning system, key risk features are extracted. These key risk features include dangerous driving behavior features and terrain-related features. A risk prediction model is constructed based on these features, which can accurately predict driving risks and classify and warn of them. This breaks through the limitations of a single data dimension, improves the comprehensiveness of the assessment, and realizes real-time perception, multi-dimensional analysis, and accurate early warning of driving risks. Attached Figure Description

[0015] One or more embodiments are illustrated by way of example with reference to the accompanying drawings, and these illustrative descriptions do not constitute a limitation on the embodiments.

[0016] Figure 1 This is a flowchart of a dynamic early warning method for driving risks of mining trucks provided in one embodiment of this application; Figure 2 This is a flowchart of a dynamic early warning method for driving risks of mining trucks provided in one embodiment of this application; Figure 3 This is a flowchart of a dynamic early warning method for driving risks of mining trucks provided in one embodiment of this application; Figure 4 This is a distribution map of different types of dangerous driving behaviors in the steps of the dynamic early warning method for driving risks of mining trucks provided in one embodiment of this application; Figure 5 This is a spatial probability density map of dangerous driving behaviors in the dynamic early warning method for driving risks of mining trucks provided in an embodiment of this application; Figure 6This is a schematic diagram illustrating the importance of key risk characteristics in the dynamic early warning method for driving risks of mining trucks provided in an embodiment of this application for predicting driving risks; Figure 7 This is a schematic diagram showing the information of the top 10 high-risk vehicles in the dynamic early warning method for driving risks of mining trucks provided in an embodiment of this application. Figure 8 This is a schematic diagram of a dynamic early warning system for driving risks of mining trucks provided in another embodiment of this application; Figure 9 This is a schematic diagram of the structure of an electronic device provided in another embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the various embodiments of this application to help readers better understand this application. However, the technical solutions claimed in this application can be implemented even without these technical details and various changes and modifications based on the following embodiments. The division of the various embodiments below is for the convenience of description and should not constitute any limitation on the specific implementation of this application. The various embodiments can be combined with and referenced by each other without contradiction.

[0018] To facilitate understanding of the embodiments of this application, relevant content regarding open-pit mining and transportation will be introduced first.

[0019] In open-pit coal mining, mining trucks are the core equipment for transporting ore and waste rock, and the safety of their driving directly affects mine production efficiency and personnel safety. Traditional methods of monitoring driving risks mainly rely on manual inspections and simple onboard equipment, which have the following limitations: (1) Insufficient real-time capability: Unable to capture vehicle driving status in real time, resulting in delayed detection of dangerous behaviors; (2) Single analysis dimension: Risk is judged only by basic data such as speed, ignoring the influence of multiple dimensions such as terrain and driving habits; (3) Low accuracy of early warning: lack of systematic data processing and model support, and risk assessment is subjective and one-sided.

[0020] With the development of vehicle positioning technology, it has become possible to achieve refined management by collecting data such as time, location, and speed. However, existing technologies have not yet formed a systematic approach, making it difficult to meet the needs of dynamic and accurate risk control.

[0021] This application aims to address the problems of poor real-time performance, incomplete feature extraction, and low accuracy of early warning in existing monitoring of driving risks for mining trucks. It provides a dynamic early warning method for driving risks of mining trucks based on an onboard positioning system, which enables real-time perception, multi-dimensional analysis, and accurate early warning of driving risks.

[0022] This application proposes a dynamic early warning method for driving risks of mining trucks based on an on-board positioning system. By collecting vehicle driving information in real time through the on-board positioning system, extracting key risk features and constructing a risk prediction model, it achieves accurate early warning of driving risks, forming a complete closed loop of "data collection-feature extraction-risk early warning" and improving the level of safety management and control in mining transportation.

[0023] The following is a detailed description of the implementation details of the dynamic early warning method for driving risks of mining trucks in this embodiment. The following content is only for the convenience of understanding and is not necessary for implementing this solution.

[0024] Example 1: The dynamic early warning method for driving risks of mining trucks in this embodiment can be applied to electronic devices with communication, computing, and data storage capabilities. Its specific process can be as follows: Figure 1 As shown, it includes: Step 101: Collect vehicle driving information in real time based on the vehicle positioning system. The driving information includes driving time, geographical location, driving status and terrain features.

[0025] Because vehicle-mounted positioning systems can reliably provide high-precision spatiotemporal and status data in open-pit coal mine environments, and are simple to deploy and cost-effective, they are more suitable for large-scale applications in complex mining environments compared to relying on manual inspections or multi-sensor fusion (such as adding millimeter-wave radar). This reduces system complexity and implementation costs. Therefore, each mining truck is equipped with at least one vehicle-mounted positioning system. Thus, vehicle driving information can be collected in real time directly based on the vehicle-mounted positioning system, such as GPS or Beidou positioning. Moreover, the driving time, geographical location, driving status, and terrain feature information collected by the vehicle-mounted positioning system already cover all the key data required for subsequent calculation of driving characteristics, identification of dangerous driving behaviors, and construction of risk feature matrices. Therefore, no additional equipment is needed to supplement the information.

[0026] Among them, driving time is the specific time (time stamp) of data collection, which can be used to calculate the rate of change of vehicle driving status. For example, acceleration needs to be calculated by the speed difference at different time points. It can also locate the specific time when dangerous behavior occurs, which is convenient for tracing and analysis.

[0027] Geographic location data, including longitude and latitude, reflects the real-time geographical location of a vehicle and is core data for determining its driving trajectory. On one hand, it can be used to clean up invalid data, such as removing duplicate points with identical latitude and longitude; on the other hand, it can mark the locations of dangerous driving behaviors on a map, such as locations of rapid acceleration and speeding, and then draw dangerous behavior distribution maps and risk level heat maps to visually display high-risk areas.

[0028] Driving status, including driving speed and heading angle (direction), is a key indicator for identifying various dangerous behaviors. For example, driving speed can be used to identify dangerous driving behaviors such as rapid acceleration and speeding; the rate of change of heading angle (speed of change of direction) at adjacent time points reflects the degree of steering aggression of the vehicle, thus aiding in the judgment of dangerous driving behaviors.

[0029] Terrain features include elevation. The gradient of a vehicle can be calculated by the elevation difference (elevation difference) and horizontal displacement (calculated from latitude and longitude) between two adjacent points. Furthermore, the mean gradient, standard deviation of gradient, and other terrain features can be obtained to help predict driving risks.

[0030] In this step, for each vehicle, the vehicle positioning system collects the vehicle's driving information in real time, which can ensure that dangerous behaviors and risk changes are captured in a timely manner, so as to provide core data support for subsequent driving behavior analysis and risk prediction.

[0031] Step 102: Extract key risk features based on driving information to reflect driving risks. Key risk features include dangerous driving behavior features and terrain-related features.

[0032] Traditional methods rely solely on single-dimensional data, such as driving speed, which fails to comprehensively reflect the influencing factors of driving risks. These factors include the synergistic effects of dangerous behaviors like rapid acceleration and terrain conditions such as slope, leading to incomplete and inaccurate risk assessments that are insufficient for effective risk prediction and early warning. Therefore, this step in the embodiment processes the collected driving information to extract two types of key risk features: dangerous driving behavior features and terrain-related features. These features are then fused and analyzed.

[0033] The characteristics of dangerous driving behavior include: the number of times of rapid acceleration, rapid deceleration, speeding, and sharp turns. Terrain-related characteristics include: mean altitude, mean slope, and standard deviation of slope.

[0034] In this step, dangerous driving behavior features such as the number of times of rapid acceleration, rapid deceleration, speeding, and sharp turns are extracted to accurately quantify the driver's operational risks. At the same time, terrain-related features such as average altitude, average slope, and slope standard deviation are extracted to reflect the impact of the driving environment on risks, breaking through the limitations of traditional single feature analysis.

[0035] Step 103: Construct a risk prediction model based on key risk characteristics to classify and warn of vehicle driving risks.

[0036] In this step, key risk characteristics such as dangerous driving behavior features and terrain-related features constitute the core dimensions of driving risk assessment and serve as input parameters for the risk prediction model. This allows for a comprehensive quantitative assessment of driving risks by integrating multiple factors, enabling the risk prediction model to consider the synergistic effects of driving behavior and the terrain environment. This facilitates the classification and early warning of vehicle driving risks, improving the accuracy of risk level prediction. Ultimately, this achieves quantitative grading and effective early warning of driving risks, meeting the needs of mine safety management for precise risk control.

[0037] The dynamic early warning method for driving risks of mining trucks provided in this application collects vehicle driving information in real time through an on-board positioning system and extracts key risk features, including dangerous driving behavior features and terrain-related features. A risk prediction model is constructed based on dangerous driving behavior features and terrain-related features, which can accurately predict driving risks and classify and warn of driving risks. This method breaks through the limitations of a single data dimension, improves the comprehensiveness of assessment, and realizes real-time perception, multi-dimensional analysis and accurate early warning of driving risks.

[0038] Example 2: According to an exemplary embodiment, most of the content of the dynamic early warning method for driving risks of mining trucks in this embodiment is the same as that in the above embodiments. The difference between this embodiment and the above embodiments is that this embodiment is a further explanation of step 102 in the above embodiments, and its specific process can be as follows: Figure 2 As shown.

[0039] In this embodiment, key risk features reflecting driving risks are extracted based on driving information, including: Step 201: Preprocess the driving information and retain the valid driving information; the valid driving information includes: timestamp, longitude, latitude, driving speed, heading angle, and altitude.

[0040] In this step, consecutive data points with the same longitude and latitude in the driving information can be identified and deleted. These data points are usually caused by signal drift or brief periods of stillness in the positioning equipment and are considered redundant information with no practical analytical value; therefore, adjacent duplicate points should be removed. For driving speed data, three types of outliers can be eliminated: points with negative driving speeds (physically impossible, indicating data acquisition errors); points with missing driving speeds (incomplete data records unusable for subsequent calculations); and points with driving speeds exceeding reasonable ranges, such as extreme values ​​far exceeding the speed limits on mining roads, which are considered abnormal interference data.

[0041] After the above processing, a standardized dataset containing six fields—timestamp, longitude, latitude, driving speed, heading angle, and altitude—is retained as the basis for subsequent feature extraction and risk analysis.

[0042] In this step, by removing adjacent duplicate points and outliers, interference from redundant information and erroneous data is eliminated, ensuring that the retained driving information is authentic and valid. This provides reliable raw data for subsequent calculations of driving characteristics such as acceleration, rate of change of heading angle, and gradient. Moreover, removing redundant duplicate points reduces the amount of data, lowers the computational cost of subsequent feature extraction and model calculation, improves the system's real-time processing efficiency, and better meets the needs of dynamic monitoring in mining areas.

[0043] In addition, in this embodiment, the retained valid data (valid driving information) is used as a prerequisite for extracting dangerous driving behavior features and terrain-related features, so as to avoid feature calculation deviations caused by outliers through preprocessed data. For example, incorrect driving speed values ​​will lead to incorrect acceleration calculations, which will affect the identification of rapid acceleration / deceleration.

[0044] Step 202: Calculate dangerous driving behavior characteristics and terrain-related features based on effective driving information.

[0045] In this step, behaviors such as rapid acceleration, rapid deceleration, speeding, and sharp turns can be identified through quantitative analysis, and the frequency of their occurrence can be counted to obtain characteristics of dangerous driving behaviors. Based on altitude and geographical location, the mean altitude, mean slope, and standard deviation of slope can be calculated to obtain terrain-related features.

[0046] In one embodiment, calculating dangerous driving behavior characteristics based on valid driving information includes: The driving characteristics of the vehicle are calculated based on effective driving information. The driving characteristics include at least: acceleration, rate of change of heading angle, horizontal displacement, gradient and altitude. Based on driving characteristics, dangerous driving behaviors are identified, including: rapid acceleration, rapid deceleration, speeding, and sharp turns. The frequency of dangerous driving behaviors is counted based on the number of dangerous driving behaviors, which is used as a characteristic of dangerous driving behavior.

[0047] In this embodiment, based on the preprocessed driving information, relevant driving features such as acceleration, rate of change of heading angle, horizontal displacement, and slope are calculated. Combined with driving speed, time series analysis, and threshold judgment, four typical dangerous driving behaviors such as rapid acceleration, rapid deceleration, speeding, and sharp turning, as well as terrain-related features, are identified.

[0048] One method is to calculate the time difference by taking the velocities (v1, v2) at two consecutive time points (t1, t2). t = t2 - t1, unit: s, calculate acceleration a using the following formula: a = (v2 - v1) / t; Unit: m / s², positive values ​​indicate acceleration, negative values ​​indicate deceleration.

[0049] The calculation process for the rate of change of heading angle may include: taking the heading angles (θ1, θ2) at two consecutive time points (t1, t2), in degrees, and calculating the time difference. t = t2 - t1, unit: seconds, and calculate the difference in heading angle between adjacent time points. θ: If |θ2-θ1|≤180°, then θ=|θ2-θ1|; if |θ2-θ1|>180°, then θ = 360° - |θ2 - θ1|, corrected to the minimum angle difference; The formula for the rate of change of heading angle r is as follows: r= θ / t; unit: ° / s.

[0050] Horizontal displacement d is the horizontal distance (unit: m) calculated using the spherical distance formula based on the longitude and latitude of two consecutive points (A and B), reflecting the distance the vehicle moves in the horizontal direction.

[0051] The slope is calculated by taking the elevation (h1, h2) of two consecutive points (A, B) in meters, and the horizontal displacement. d (unit: m), calculate the altitude difference The slope s is calculated using the formula h = h2 - h1: s=arctan( h / d); Altitude is obtained by directly extracting altitude data (h, unit: m) from valid driving information for subsequent calculation of terrain-related features.

[0052] Driving characteristics may also include: driving speed in straight sections and driving speed in curved sections; Based on the driving characteristics calculated above, and combined with preset thresholds, four types of dangerous driving behaviors are identified. The identification conditions for dangerous driving behaviors are as follows: When the calculated acceleration a > 3 m / s², it is determined as one rapid acceleration; When the calculated acceleration a < -3m / s², it is determined as one rapid deceleration.

[0053] You can first determine the road segment type by the rate of change of heading angle: a rate of change of heading angle r < 1° / s indicates a straight road, and r ≥ 1° / s indicates a curve. When a vehicle is in a straight section of road, a speed exceeding 40 km / h is considered one instance of speeding. When a vehicle is in a curved area, if the driving speed is greater than 25 km / h, it is considered as one instance of speeding. When the rate of change of heading angle r > 5° / s and the driving speed at the corresponding time point > 9km / h, it is judged as a sharp turn.

[0054] By traversing all continuous valid driving information data points, the cumulative count of each type of dangerous driving behavior identified above is obtained: the number of times of rapid acceleration, the number of times of rapid deceleration, the number of times of speeding, and the number of times of sharp turns. These four data together constitute the characteristics of dangerous driving behavior and serve as input parameters for subsequent risk prediction models.

[0055] In one embodiment, calculating terrain-related features based on valid driving information includes: Extract valid altitude data and calculate the average altitude based on the valid altitude data; Calculate the slope of a single road segment, and calculate the average slope of all road segments. Obtain the mean slope and calculate the standard deviation of the slope.

[0056] Specifically, all altitude data (unit: m) within the target time period can be extracted from the preprocessed valid driving information and denoted as h1, h2, ..., h n Where n is the total number of altitude data points. The mean altitude is calculated using the following formula: Average altitude = (h1 + h2 + ... + h) n ) / n.

[0057] The process of calculating the average slope may include the following steps: For two consecutive points (denoted as point A and point B) in the valid driving information: extract the elevation data h of the two points. _A and h _B Calculate the altitude difference h=h _B -h _A Unit: m; Horizontal displacement is calculated using the spherical distance formula based on the longitude and latitude of two points. d (unit: m, reflecting the horizontal distance between two points); calculate the slope of a single road segment using the formula: slope s = arctan( h / d) The radians can be converted to degrees. Then, the above process is repeated for all consecutive points within the target time period to obtain the slope data s1, s2, ..., s of all road segments.m Where m is the total number of road segments, the slope of all road segments is statistically analyzed, and finally, the average slope μ is calculated using the following formula: μ = (s1 + s2 + ... + s m ) / m.

[0058] The above process is used to obtain the slope data s1, s2, ..., s of all road segments. m The slope mean μ is used to calculate the slope standard deviation using the following formula. This data reflects the dispersion of the slope data: Slope standard deviation = -μ) 2 / m]; in, -μ) 2 It is the sum of squares of the differences between each slope value and the mean slope, where m is the total number of road segments.

[0059] Through the above process, three indicators of terrain-related features are finally obtained: mean elevation, mean slope, and standard deviation of slope, which serve as input parameters for subsequent risk prediction models.

[0060] Example 3: According to an exemplary embodiment, most of the content of the dynamic early warning method for driving risks of mining trucks in this embodiment is the same as that in the above embodiments. The difference between this embodiment and the above embodiments is that this embodiment is a further explanation of step 103 in the above embodiments, and its specific process can be as follows: Figure 3 As shown.

[0061] In this embodiment, a risk prediction model is constructed based on key risk characteristics to classify and warn of vehicle driving risks, including: Step 301: Construct a risk feature matrix based on dangerous driving behavior characteristics and terrain-related features.

[0062] Because existing technologies often rely on single features or fragmented data to comprehensively reflect the multidimensional influencing factors of driving risks, risk prediction models suffer from non-standard inputs and low prediction accuracy. Therefore, this application addresses this issue by integrating the synergistic effect of driver operating habits and terrain environment—that is, disparate dangerous driving behavior features and terrain-related features—into a structured matrix. This standardizes the data format, providing a standardized input carrier for subsequent risk prediction models and resolving the problem of fragmented data preventing direct modeling in existing technologies.

[0063] In this step, because the risk feature matrix encompasses two core factors—driver operation (dangerous driving behavior) and driving environment (terrain)—the risk prediction model can comprehensively analyze the synergistic impact of both on driving risk. This overcomes the limitations of traditional single-feature assessments and enhances the comprehensiveness of risk prediction. Furthermore, the standardized risk feature matrix reduces the complexity of data preprocessing for the risk prediction model. Simultaneously, the synergistic input of multiple features, including dangerous driving behavior characteristics and terrain-related features, provides the risk prediction model with richer risk correlation information, helping it to more accurately classify risk levels and laying a data foundation for subsequent classification and early warning.

[0064] Step 302: Use the risk feature matrix as input to the risk prediction model to predict the driving risk of the vehicle and assign risk level labels; the risk prediction model includes a random forest model.

[0065] In this step, firstly, the labeled data can be constructed as the target variable of the risk prediction model: each sample can be labeled with a corresponding risk level label of high, medium, or low. The risk level label can be determined based on historical accident data, manual assessment, or preset rules, such as a threshold for the total number of dangerous behaviors, to form a training dataset of risk feature matrix-risk label.

[0066] Next, you can set the parameters of the random forest model, including the number of decision trees, evaluation metrics, and other parameters.

[0067] The number of decision trees can be set to 100, which is a validated optimal parameter for balancing model performance and computational efficiency. The out-of-bag (OOB) error rate can be used to evaluate model performance; it requires no additional validation set and calculates the error using samples not used in decision tree training. Other parameters can be set to the standard settings for the random forest algorithm by default.

[0068] A risk feature matrix is ​​constructed as input to the risk prediction model. This risk feature matrix includes dangerous driving behavior characteristics and terrain-related characteristics, such as seven features: the number of rapid accelerations, the number of rapid decelerations, the number of speeding incidents, the number of sharp turns, the mean altitude, the mean slope, and the standard deviation of the slope. For example, a specific driving period of a single truck can be used as a sample unit, with each sample corresponding to one row of data. The seven features can be arranged in a fixed order, such as the number of rapid accelerations, the number of rapid decelerations, the number of speeding incidents, the number of sharp turns, the mean altitude, the mean slope, and the standard deviation of the slope, as columns of the matrix to form a structured risk feature matrix.

[0069] For example, the risk feature matrix is ​​used as input to a random forest model. The model is iteratively trained using the random forest algorithm, and its performance is evaluated using the out-of-bag (OOB) error rate. This completes the training of the random forest model, which then predicts the driving risk of the vehicle and assigns a risk level label. The risk level label includes high, medium, and low, allowing for a clear distinction between the degree of driving risk. In this process, the random forest model, through ensemble learning of multiple decision trees, integrates the synergistic effects of dangerous driving behavior features and terrain-related features, avoiding the limitations of a single decision tree. Compared to traditional single-rule or simple models, it can more accurately classify risk levels (high, medium, low), providing a reliable basis for early warning. In other exemplary embodiments, more risk level labels can be set, such as low, low-medium, medium, medium-high, and high, without specific limitations.

[0070] In one embodiment, the process of predicting the driving risk of a vehicle through a risk prediction model further includes: analyzing and calculating the contribution of each key risk feature to the risk prediction result through a random forest model. In this process, during the training of the risk prediction model, feature importance analysis can be performed to calculate the contribution of each feature, such as the number of rapid accelerations, the number of speeding incidents, and the average slope, to the risk prediction results. This clarifies the influence weight of different key risk features on the driving risk prediction results, providing targeted scientific basis for open-pit coal mine transportation safety management. For example, by quantifying the contribution of each key risk feature, such as determining that the number of speeding incidents contributes the most, managers can formulate targeted control measures, such as strengthening speed limit monitoring on curves and increasing penalties for speeding, rather than allocating resources evenly to all features, significantly improving the efficiency and effectiveness of risk control. Furthermore, the changes in the contribution of each key risk feature can be tracked over a long period. For example, an increase in the contribution of the number of rapid decelerations over a certain period can reflect dynamic changes in driving behavior or terrain risks, providing data support for mine road planning, such as optimizing steep slope sections and driver training (such as targeted correction of rapid deceleration habits), thus achieving continuous optimization of safety management.

[0071] Example 4: According to an exemplary embodiment, most of the content of the dynamic early warning method for driving risks of mining trucks in this embodiment is the same as that in the above embodiments. The difference between this embodiment and the above embodiments is that, based on the above embodiments, this embodiment further includes: The system can visualize the classification and warning results of vehicle driving risks, and / or provide real-time alerts for vehicles with high risk levels.

[0072] In this embodiment, the classification and warning results of vehicle driving risks are visualized. Visualization methods include dangerous driving behavior distribution maps, risk level heatmaps, and information on high-risk vehicles. Risk heatmaps can be generated, using color depth to represent the level of risk in different areas, to display the classification and warning results of vehicle driving risks, facilitating viewing and interaction by management personnel.

[0073] In this step, such as Figure 4 As shown, different types of dangerous driving behaviors can be mapped, such as using scatter plots to mark the locations of behaviors like rapid acceleration / deceleration, thus representing the location of dangerous behaviors in scatter plot form. Figure 5 As shown, an exemplary spatial probability density map of dangerous driving behavior is presented. This method, based on the classification and warning results of vehicle driving risks, displays the spatial distribution of dangerous driving behaviors in a specific geographical area to represent the density of dangerous behaviors occurring in different regions. Figure 6 As shown, this example illustrates the importance of key risk features in driving risk prediction, demonstrating the significance of these features through bar charts. This visually displays the contribution of key risk features and clarifies which features have a greater impact on risk prediction. Figure 7 As shown, it can also display information on vehicles with high risk levels. For example, it can output a screenshot of a table ranking the top 10 vehicles with high risk levels, showing the top 10 vehicles with the highest risk levels.

[0074] In this embodiment, the risk level is output based on the risk prediction model and combined with visualization to provide managers with an intuitive and scientific basis for control.

[0075] In this embodiment, during the visualization of the classification and early warning results of vehicle driving risks, management personnel can be notified of vehicles with high risk levels through pop-up windows, sound and light alarms, etc. on the mining area management terminal to assist in timely intervention.

[0076] Example 5: Another embodiment of this application relates to a dynamic warning system for driving risks of mining trucks. The implementation details of this embodiment's dynamic warning system for driving risks of mining trucks are described below. The following implementation details are provided for ease of understanding and are not essential for implementing this solution. A schematic diagram of the dynamic warning system for driving risks of mining trucks in this embodiment can be seen as follows: Figure 8 As shown, the dynamic warning system for driving risks of this mining truck includes: The data acquisition module 100 is used to collect vehicle driving information in real time based on the vehicle positioning system. The driving information includes driving time, geographical location, driving status and terrain. The extraction module 200 is used to extract key risk features that reflect driving risks based on driving information. Key risk features include dangerous driving behavior features and terrain-related features. The early warning module 300 is used to build a risk prediction model based on key risk characteristics in order to classify and warn about the driving risks of vehicles.

[0077] In one embodiment, the acquisition module 100 includes an on-board positioning system, which is used to acquire vehicle driving information in real time.

[0078] In one embodiment, the extraction module 200 includes: The preprocessing module is used to preprocess the driving information and retain the valid driving information. The valid driving information includes: timestamp, longitude, latitude, driving speed, heading angle, and altitude. The calculation module is used to calculate dangerous driving behavior characteristics and terrain-related features based on effective driving information. Dangerous driving behavior characteristics include: number of rapid accelerations, number of rapid decelerations, number of speedings, and number of sharp turns. Terrain-related features include: mean altitude, mean slope, and standard deviation of slope.

[0079] In one embodiment, the preprocessing module is used to remove adjacent duplicate points with the same longitude and latitude; and / or to remove outlier points with negative driving speeds, missing points, or points outside a reasonable range.

[0080] In one embodiment, the computing module includes: The computing unit is used to calculate the driving characteristics of the vehicle based on effective driving information. The driving characteristics include at least: acceleration, rate of change of heading angle, horizontal displacement, gradient and altitude. The identification unit is used to identify dangerous driving behaviors based on driving characteristics, including: rapid acceleration, rapid deceleration, speeding, and sharp turns. The statistical unit is used to count the number of dangerous driving behaviors based on dangerous driving behavior, as a characteristic of dangerous driving behavior.

[0081] In one embodiment, the identification unit is further configured to identify dangerous driving behavior using the following identification conditions: When the acceleration is greater than 3 m / s², it is identified as rapid acceleration; When the acceleration is less than -3 m / s², it is identified as rapid deceleration; When the driving speed in a straight section is greater than 40 km / h, or the driving speed in a curved section is greater than 25 km / h, it is considered speeding. When the rate of change of heading angle is greater than 5° / s and the speed is greater than 9km / h, it is identified as a sharp turn.

[0082] In one embodiment, the early warning module 300 includes: The building unit is used to construct a risk feature matrix based on dangerous driving behavior characteristics and terrain-related features; The output unit is used to take the risk feature matrix as input to the risk prediction model, so as to predict the driving risk of the vehicle through the risk prediction model and assign risk level labels. In one embodiment, the output unit is also used to analyze and calculate the contribution of each key risk feature to the risk prediction result through a random forest model. In one embodiment, the dynamic early warning system for driving risks of mining trucks further includes: The display module is used to visually present the classification and warning results of vehicle driving risks; and / or, The alert module is used to provide real-time alerts for vehicles with high-risk levels.

[0083] It is worth mentioning that all modules involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this application, this embodiment does not introduce units that are not closely related to solving the technical problems proposed in this application; however, this does not mean that other units are absent in this embodiment.

[0084] Example 6: Another embodiment of this application relates to an electronic device, such as... Figure 9 As shown, it includes: at least one processor 901; and a memory 902 communicatively connected to the at least one processor 901; wherein the memory 902 stores instructions executable by the at least one processor 901, the instructions being executed by the at least one processor 901 to enable the at least one processor 901 to execute the dynamic early warning method for driving risks of mining trucks in the above embodiments.

[0085] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0086] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0087] Example 7: Another embodiment of this application relates to a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the method embodiments described above.

[0088] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0089] Those skilled in the art will understand that the above embodiments are specific embodiments for implementing this application, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of this application.

Claims

1. A dynamic early warning method for driving risk of a mine truck, characterized in that, The method comprises: collecting driving information of a vehicle in real time based on a vehicle positioning system, the driving information comprising driving time, geographical position, driving state and terrain characteristics; extracting key risk features reflecting driving risk based on the driving information, the key risk features comprising dangerous driving behavior features and terrain-related features; constructing a risk prediction model based on the key risk features to classify and warn of driving risk of the vehicle.

2. The method for dynamically warning the driving risk of the mining truck according to claim 1, characterized in that, The step of extracting key risk features reflecting driving risk based on the driving information comprises: preprocessing the driving information to retain valid driving information, the valid driving information comprising timestamp, longitude, latitude, driving speed, heading angle and altitude; calculating the dangerous driving behavior features and the terrain-related features based on the valid driving information, the dangerous driving behavior features comprising number of times of sudden acceleration, sudden deceleration, overspeed and sudden turning, and the terrain-related features comprising mean altitude, mean slope and slope standard deviation.

3. The method for dynamically warning the driving risk of the mining truck according to claim 2, characterized in that, The preprocessing comprises: removing adjacent repeated points with the same longitude and latitude; and / or removing abnormal points with negative, missing or unreasonable driving speed.

4. The mine truck driving risk dynamic early warning method according to claim 2 or 3, characterized in that, The step of calculating the dangerous driving behavior features based on the valid driving information comprises: calculating driving features of the vehicle based on the valid driving information, the driving features comprising acceleration, heading angle change rate, horizontal displacement, slope and altitude; identifying dangerous driving behaviors based on the driving features, the dangerous driving behaviors comprising sudden acceleration, sudden deceleration, overspeed and sudden turning; counting the number of times of the dangerous driving behaviors as the dangerous driving behavior features.

5. The method for dynamically warning the driving risk of the mining truck according to claim 4, characterized in that, The step of identifying dangerous driving behaviors based on the driving features comprises: the identification conditions of the dangerous driving behaviors are as follows: sudden acceleration is identified when the acceleration is greater than 3 m / s²; sudden deceleration is identified when the acceleration is less than -3 m / s²; overspeed is identified when the driving speed is greater than 40 km / h in a straight road area or the driving speed is greater than 25 km / h in a curve road area; sudden turning is identified when the heading angle change rate is greater than 5° / s and the speed is greater than 9 km / h.

6. The mine truck driving risk dynamic early warning method according to claim 5, characterized in that, The step of constructing a risk prediction model based on the key risk features to classify and warn of driving risk of the vehicle comprises: constructing a risk feature matrix based on the dangerous driving behavior features and the terrain-related features; inputting the risk feature matrix as an input of a risk prediction model to predict driving risk of the vehicle by the risk prediction model and assign a risk level label, the risk prediction model comprising a random forest model.

7. The mine truck driving risk dynamic early warning method according to claim 1 or 6, characterized in that, The method further comprises: visually displaying the classification and warning result of the driving risk of the vehicle; and / or, real-time prompting the vehicle with high risk level.

8. A dynamic pre-warning system for driving risk of a mine truck, characterized in that, The method comprises: a collecting module configured to collect driving information of a vehicle in real time based on a vehicle positioning system, the driving information comprising driving time, geographical position, driving state and terrain; an extracting module configured to extract key risk features reflecting driving risk based on the driving information, the key risk features comprising dangerous driving behavior features and terrain-related features; The early warning module is configured to construct a risk prediction model based on the key risk features to classify and warn the driving risk of the vehicle.

9. An electronic device, comprising: The method comprises: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method for dynamically warning the driving risk of the mining truck according to any one of claims 1 to 7.

10. A computer readable storage medium storing a computer program, characterized in that, The computer program is executed by the processor to implement the method for dynamically warning the driving risk of the mining truck according to any one of claims 1 to 7.

Images