Powder and particle material transportation semitrailer safety supervision system based on neural network
Through the safety supervision system of powder-particle material transportation semi-trailer based on neural network, integrating a variety of information collectors and neural network technologies, the problem of difficult to effectively supervise semi-trailer transportation safety in the existing technology is solved, and road and vehicle information is realized. Score assignment and safety prediction improve the effect of transportation safety and driving reminders.
Patent Information
- Application Number
- CN202411378034.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-30
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
It is difficult for the existing technology to effectively summarize and supervise various factors affecting road traffic safety, especially traffic accidents caused by semi-trailer transportation, resulting in poor safety supervision results.
Adopt a safety supervision system for powder particle material transportation semi-trailer based on neural network, and integrates a variety of information collectors such as high-definition cameras, microwave detectors, geomagnetic detectors, etc. through information acquisition modules, information analysis modules, information storage modules and information prediction modules. Use neural network technology to assign and weighted calculations to road and vehicle information, predict future safety status, and distribute safety information to vehicles and command centers.
A comprehensive collection and layered score of factors affecting road traffic safety has been achieved, and safe driving can be directly evaluated and reminded of future transportation safety status, improved transportation safety, and issued early warnings to drivers and command centers in a timely manner.
Smart Images

Figure CN118940191A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of road safety monitoring, and more specifically to a safety monitoring system for a powder and particle material transport semitrailer based on a neural network. Background Art
[0002] Road traffic safety refers to the state in which casualties or property losses can be controlled within an acceptable level during traffic activities. Traffic safety means that the possibility of loss of people or property is acceptable; if this possibility exceeds the acceptable level, it is unsafe. As a dynamic open system, the safety of the road traffic system is constrained by both internal factors of the system and the interference of the external environment of the system, and is closely related to factors such as people, vehicles and road environment. The unreliability, imbalance and instability of any factor in the system may lead to conflicts and contradictions, and produce unsafe factors or unsafe conditions.
[0003] All road traffic participants will affect traffic safety through their own behavior, but when it comes to the intensity of traffic safety incidents, large vehicles represented by semi-trailers are the first to be affected. Traffic accidents caused by semi-trailers are often accompanied by the loss of life and a large amount of public and private property. How to summarize and compile statistics on various factors that affect road traffic safety and improve safety supervision is an issue that needs to be urgently addressed. Summary of the invention
[0004] In order to solve the above problems and overcome the shortcomings of the prior art, the present invention provides a safety monitoring system for a powder and particle material transport trailer based on a neural network.
[0005] To achieve the above-mentioned purpose, the present invention provides a safety monitoring system for a powder material transport trailer based on a neural network, comprising: An information acquisition module, used to acquire road information including traffic flow data, road facility data, and road history data within a predetermined road section, vehicle information including vehicle driving status data and driver driving habit data, and meteorological data including weather conditions, temperature conditions, and visibility conditions within a predetermined road section; An information analysis module, which is connected to the information acquisition module, uses neural network technology to summarize, score and weight the acquired information to determine the total safety score including the road safety value and the vehicle safety value; An information storage module, which is connected to the information acquisition module and the information analysis module, stores information according to the mapping relationship between road information, vehicle information and total safety score, and constructs a deep database and a shallow database respectively based on the preset time segment as the division standard; The information prediction module is connected to the information analysis module and the information storage module. It uses neural network technology to extract key features of data in the shallow database, predicts and judges the security status of the future preset time period, and gives a corresponding total security score; The information distribution module is connected to the information analysis module and the information prediction module to distribute the current total safety score and the total safety score for a preset time period in the future to designated vehicles and the command center.
[0006] Furthermore, the information acquisition module at least includes A high-definition camera is installed on a predetermined road section and is used to collect high-definition image information within the predetermined road section; A microwave detector installed on a predetermined road section is used to detect the number of vehicles, vehicle speed, traffic composition and lane occupancy in the predetermined road section; The geomagnetic detector is installed on the predetermined road section to identify vehicles on the lane; Visibility detection equipment, used to detect visibility information of a predetermined road section; The weather warning interface module is used to access the national meteorological system to obtain weather information for the predetermined road section.
[0007] Furthermore, the traffic flow data includes the number of vehicles, vehicle speeds, traffic composition and lane occupancy within a predetermined road section; Road facility data include road type, road speed limit, number of road lanes, road slope, road flatness, road skid resistance, number of street lights, road marking conditions and emergency equipment conditions; Road history data includes the number of historical traffic accidents, traffic accident severity and high-incidence sections of traffic accidents within the predetermined road section; Vehicle driving status data includes vehicle tire pressure status, vehicle brake health status, vehicle load and vehicle actual speed; The driver's driving habit data includes the speed distribution, acceleration at intersections, braking force, lane change frequency and overtaking frequency of the driver of a specific vehicle during the entire driving of the predetermined road section.
[0008] Furthermore, the information acquisition module uploads the information to the information analysis module, uses neural network technology to extract and classify the features of the data information, and performs weighted scoring processing on different types of data according to the preset scoring strategy: The influence level is divided into four levels according to preset rules: no influence, slight influence, general influence and major influence, with the weights preset as 0%, 15%, 30% and 55% respectively; According to the preset stratification rules, various types of data are divided into various influence levels, the number of data types contained in each influence level is counted, and the total score of the influence level is evenly assigned to various types of data; The scoring levels of various types of data are divided into four levels: excellent, general, caution and dangerous according to preset thresholds. The scores obtained for each scoring level are 100%, 60%, 30% and 10% of the scores assigned to the data of this type.
[0009] Furthermore, the influence level division is managed in a floating manner according to the law of information change in the deep database. Within a preset period of time, when the proportion of a certain data type in the proportion of traffic accident factors falls below a threshold, it will be downgraded from its influence level to the next level; when the proportion of a certain data type in the proportion of traffic accident factors rises above a threshold, it will be upgraded from its influence level to the previous level; Within the preset period of time, when the proportion of a certain influence level in the proportion of the levels of factors causing traffic accidents drops below the threshold, the weight of that influence level will be reduced by 3%. At the same time, if the proportion of other influence levels in the proportion of the levels of factors causing traffic accidents rises above the threshold, the 3% weight will be fully assigned to that influence level. If no other influence level rises above the threshold, the 3% weight will be divided according to the current weight ratios of other influence levels and assigned to other influence levels respectively.
[0010] Furthermore, the proportion of the types of factors causing traffic accidents refers to: the ratio of the number of times a certain type of data is assigned a dangerous scoring level when a traffic accident occurs to the total number of traffic accidents within a preset period of time; The proportion of the level of factors causing traffic accidents refers to: within a preset period of time, the proportion of all types of data in a certain influence level that are assigned to dangerous levels to the total number of data types.
[0011] Furthermore, the deep database contains all the data since the system started running. The deep database is in a dormant state to save energy loss and bandwidth. After a transport is completed, all the data of the transport is packaged and stored. The shallow database contains all data within the preset cycle time before this transportation, and the shallow database is in an activated state for real-time reading and storage.
[0012] Furthermore, based on the data in the shallow database, neural network technology is used to extract time features, habit features and scoring features to form a dual-track curve graph model, which is then fed into the neural network for training to predict the total safety score of this transportation at a certain time node in the future. When the predicted habit score and total safety score are lower than the safety threshold, early warning reminder information is issued to the designated vehicle and command center in advance.
[0013] Furthermore, the habit feature is a score assigned to the driving habit category data of a specific vehicle driver.
[0014] Furthermore, the deep database is awakened according to a preset period to extract relevant data to calculate the proportion of types of factors causing traffic accidents and the proportion of levels of factors causing traffic accidents. After adjusting the data type level division and influence level weights, the adjustments are recorded in the change log.
[0015] The beneficial effects of the present invention are: Through a variety of information collectors, various factors that affect road traffic safety are comprehensively collected and stratified according to the degree of impact on traffic safety. Then, the factors are classified and scored through preset and floating scoring methods to obtain a final total safety score. This can not only directly evaluate and remind safe driving, but also comprehensively summarize and analyze various factors.
[0016] By collecting and storing data during the continuous transportation process and aggregating them into neural networks for training, it will be possible to predict and remind future safe transportation driving.
[0017] It can collect, summarize and analyze drivers' driving habit data, remind drivers during the transportation process, and improve transportation safety. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Attached Figure 1 It is a system structure diagram of the present invention; Attached Figure 2 It is the first transport double track curve diagram of the present invention; Attached Figure 3 It is the N-1th transport double-track curve diagram of the present invention; Attached Figure 4 It is the Nth transport double-track curve prediction diagram of the present invention. DETAILED DESCRIPTION
[0019] In order to make the purpose, technical scheme and advantages of the present invention more clear, the following will be combined with the appended claims of the present invention. Figure 1 ~Attached Figure 4 , the present invention is described in more detail.
[0020] The neural network-based powder and particle material transport semi-trailer safety supervision system provided by the present invention comprises: An information acquisition module, used to acquire road information including traffic flow data, road facility data, and road history data within a predetermined road section, vehicle information including vehicle driving status data and driver driving habit data, and meteorological data including weather conditions, temperature conditions, and visibility conditions within a predetermined road section; The information acquisition module at least includes a high-definition camera, which is installed on a predetermined road section and is used to collect high-definition image information within the predetermined road section; a microwave detector, which is installed on a predetermined road section and is used to detect the number of vehicles, vehicle speed, traffic composition and lane occupancy within the predetermined road section; a geomagnetic detector, which is installed on a predetermined road section and is used to identify vehicles on lanes; a visibility detection device, which is used to detect visibility information of the predetermined road section; and a meteorological warning interface module, which is used to access the national meteorological system to obtain meteorological information of the predetermined road section.
[0021] Traffic flow data includes the number of vehicles, vehicle speed, traffic composition and lane occupancy within the predetermined road section. The number of vehicles has a greater impact on the vehicle speed, a greater impact on the traffic safety intensity, and a smaller impact on the probability of traffic safety. Comprehensive consideration can be determined as a light impact level. Similarly, vehicle speed has a greater impact on the traffic safety intensity and the probability of traffic safety. In particular, speeding is an important cause of traffic accidents. Comprehensive consideration can be used to locate the heavy impact level. According to the same strategy, the influence level of each data type is manually divided when the system is deployed.
[0022] Road facility data include road type, road speed limit, number of road lanes, road slope, road flatness, road anti-skid, number of street lights, road sign line conditions and emergency equipment conditions. For example, if the road type is a national highway, due to the large number of residents around and multiple intersections, it can be comprehensively judged as a general impact level. According to the same strategy, the influence level of each data type is manually divided when the system is deployed; Road historical data include the number of historical traffic accidents, the severity of traffic accidents and sections with high incidence of traffic accidents within the predetermined sections. According to the same strategy, the influence level of each data type is manually divided when the system is deployed. It is worth noting that road historical data can be collected in a variety of ways. One is to collect it through information acquisition devices on authoritative online platforms, the second is to collect it through information acquisition devices on news platforms, and the third is to collect traffic accidents that occurred or recorded during the operation of this system.
[0023] The vehicle driving status data includes the vehicle tire pressure status, vehicle brake health status, vehicle load and vehicle actual speed. According to the same strategy, the influence level of each data type is manually divided when the system is deployed. It is worth noting that the vehicle driving status data can be sensed, collected and transmitted through sensors installed on the vehicle and weighed at the factory.
[0024] Driver driving habit data include the speed distribution, acceleration at intersections, braking force, lane change frequency and overtaking frequency of a specific vehicle driver during the entire driving process of a predetermined road section. The driver's driving habit data can be sensed, collected and transmitted through sensors installed on the vehicle. According to the same strategy, the influence level of each data type is manually divided when the system is deployed.
[0025] The information acquisition module uploads the information to the information analysis module, uses neural network technology to extract and classify the features of the data information, and performs weighted scoring processing on different types of data according to the preset scoring strategy: The neural network processes the data information and decomposes the signal into wavelet coefficients of different scales and frequency bands. The wavelet transform provides local information of the signal in the time domain and frequency domain, so it has better time-frequency resolution. The wavelet coefficients are thresholded to suppress noise by setting smaller coefficients to zero. The threshold is usually calculated based on the energy or statistical properties of the coefficients. The processed wavelet coefficients are transformed back to the original signal, and the inverse transform produces a denoised signal.
[0026] The denoised signal can be used as more reliable and accurate data for subsequent data analysis, habit recognition and other applications.
[0027] The influence level is divided into four levels according to preset rules: no influence level, slight influence level, general influence level and major influence level. The weights are preset as 0%, 10-20%, 25-40% and 50-65% respectively. The weights of each level can be adjusted according to actual conditions, but the sum of the weights of the four levels must be 100%; According to the preset stratification rules, various types of data are divided into various influence levels. Each influence level contains multiple data types. The number of data types contained in each influence level is counted N, and the total score of the influence level is evenly assigned to each type of data. The score of each type of data is changed to 100 / N; The scoring levels of various types of data are divided into four levels: excellent, general, caution and dangerous according to preset thresholds. The scores obtained for each scoring level are 100~90%, 60~70%, 30~40% and 5~10% of the scores assigned to that type of data. This means that when a certain type of data is judged to be at a certain level, the score obtained is 100 / N*the percentage of that level. The percentage of each level can be adjusted according to actual conditions, but the sum of the percentages of the four levels must be 100%.
[0028] When the system is deployed, the data type hierarchy is divided by big data statistical analysis and empirical analysis, which has certain limitations. During the operation of the system, by collecting and comparing operating data, the data is fed into the neural network for training, and thus floating management is performed, so as to gradually adjust the matching degree of the system to the actual situation and improve the accuracy of safety reminders.
[0029] The influence level division is managed on a floating basis according to the information change rules in the deep database. The deep database is woken up at a preset period to extract relevant data to calculate the types and levels of factors that cause traffic accidents. After adjusting the data type level division and influence level weights, the adjustments are recorded in the change log.
[0030] Within the preset period, when the proportion of a certain data type in the proportion of traffic accident factors falls below the threshold, it will be downgraded from its influence level to the next level; when the proportion of a certain data type in the proportion of traffic accident factors rises above the threshold, it will be upgraded from its influence level to the previous level; The proportion of types of traffic accident causing factors involved here refers to the ratio of the number of times a certain type of data is assigned a dangerous scoring level when a traffic accident occurs to the total number of traffic accidents within a preset period of time. For example, the road slope data was originally at a light impact level. Within a six-month period, the road slope data was judged to be at a dangerous level when a traffic accident occurred. This means that this long downhill slope may be one of the important factors causing the accident. If the data is judged to be at a dangerous level many times within six months and its ratio exceeds the threshold, then it will be upgraded from a light impact level to a general impact level.
[0031] In addition to the adjustment of the data type level, the weight of each influence level is also adjusted according to the actual situation of transportation: Within the preset period of time, when the proportion of a certain influence level in the proportion of the levels of factors causing traffic accidents drops below the threshold, the weight of that influence level will be reduced by 3%. At the same time, if the proportion of other influence levels in the proportion of the levels of factors causing traffic accidents rises above the threshold, the 3% weight will be fully assigned to that influence level. If no other influence level rises above the threshold, the 3% weight will be divided according to the current weight ratios of other influence levels and assigned to other influence levels respectively.
[0032] The proportion of the level of factors causing traffic accidents refers to: the proportion of the number of all types of data in a certain influence level that are assigned to dangerous levels within the preset period of time to the total number of data types. For example, the general influence level contains 11 data types. Within a half-year period, when a traffic accident occurs, the number of data that is assigned to a dangerous level is 1, and the ratio is 9%, which is significantly lower than the preset threshold. That means that the influence of the general influence level on safety accidents is smaller than the preset weight. In this case, 3% is subtracted from its standard weight of 35%, that is, its weight is adjusted to 32%. The proportion of the levels of factors causing traffic accidents in the light influence level and the heavy influence level do not exceed the threshold. In this case, according to the ratio between the two, that is, 15%:55% (3:11), the 3% is divided into two parts and added to the weights of the light influence level and the heavy influence level respectively.
[0033] It should be clarified that the above-mentioned time period such as half a year, the above-mentioned weight such as 55%, and the above-mentioned adjustment range such as 3% need to be assigned values according to actual conditions when the system is actually deployed. The above-mentioned data are only obtained for illustration in the embodiment, and any other values are within the protection scope of the present invention.
[0034] An information storage module, which is connected to the information acquisition module and the information analysis module, stores information according to the mapping relationship between road information, vehicle information and total safety score, and constructs a deep database and a shallow database respectively based on the preset time segment as the division standard; The deep database contains all the data from the beginning of the system operation. The deep database is in a dormant state to save energy loss and bandwidth. After a transportation is completed, all the data of the transportation is packaged and stored. The above data is classified to form a mapping relationship. When the database is awakened to extract information, the corresponding related data is quickly extracted according to the mapping relationship; The shallow database contains all the data within the preset cycle time before this transportation. The shallow database is in an activated state for real-time reading and storage. The cycle time can be set according to the actual situation. In order to ensure the accuracy of the prediction, the cycle is not shorter than 3 months or 60 transportations. After completing the 4th month or the 61st transportation, the oldest data packet in the shallow database will be transferred to the deep database, and enter the dormant state after the mapping relationship is established.
[0035] The information prediction module is connected to the information analysis module and the information storage module. It uses neural network technology to extract key features of data in the shallow database, predicts and judges the security status of the future preset time period, and gives a corresponding total security score; Based on the data in the shallow database, neural network technology is used to extract time features, habit features and scoring features to form a dual-track curve map model, in which the horizontal axis is the time used for a single transport and the vertical axis is the obtained score, so that the corresponding score floating curve can be drawn. The map includes a total safety score curve and a habit score curve, in which a total safety score threshold and a habit score threshold are set. The two thresholds are determined through experience and big data analysis, and the part of the curve below the threshold is the accident-prone interval.
[0036] All transportation data are summarized and mapped, and then fed into a neural network for training to predict the total safety score of this transportation at a certain time point in the future. When the predicted habit score and total safety score are lower than the safety threshold, early warning reminders are sent to designated vehicles and command centers in advance.
[0037] The collected data is used as the input parameter of the neural network. According to the preset threshold, judgment criteria and other requirements, the specific data type to be processed is determined. According to the nature of the data type and the characteristics of the data, the appropriate neural network type is selected, such as fully connected neural network, convolutional neural network (CNN), recurrent neural network (RNN), etc. According to the selected network type, the structure of the network is determined, including the depth (number of layers) and width (number of nodes in each layer) of the network. For deep neural networks, the depth and width of the network can usually be increased by increasing the number of hidden layers and the number of nodes; select a suitable activation function in each network layer, and select it according to the performance and convergence speed of the network; according to the nature of the data type and the output type of the network, select the loss function, such as the mean square error loss function, the cross entropy loss function, etc. ; Preprocess the prepared data, including normalization, standardization, data smoothing and other operations to improve the training effect and convergence speed of the neural network; Use appropriate feature extraction methods, such as statistical methods, frequency domain analysis methods or time domain analysis methods, to extract key features from the data, such as the mean, variance, spectrum energy, etc. of the data; Use the prepared training data set to train the designed neural network, and continuously adjust the network parameters through the back propagation algorithm and optimizer so that the network can learn the patterns and laws in the data; Based on the data in the historical transportation process as data samples, train the model to identify the characteristic patterns under these different states; Use the trained model to classify or regress the newly collected data to predict the possible intervals of various types of data in a specific time period in the future; Based on the results of the model output, an initial judgment is made on the data scoring within each influence level. If the data characteristics and patterns are consistent with the characteristic status of past historical big data, the predicted data scoring is judged to be normal. If there are abnormal characteristics or patterns, go to the previous step, re-assign the data within each influence level and make an initial judgment.
[0038] The information distribution module is connected to the information analysis module and the information prediction module to distribute the current total safety score and the total safety score for a preset time period in the future to designated vehicles and the command center.
[0039] Embodiment 1 is as follows: When the system is deployed, the total security score is set to 100 points, the weight of the no-impact level is 0%, the weight of the slight impact level is 15%, the weight of the general impact level is 35%, and the weight of the severe impact level is 55%; Among them, the no-impact level includes the number of street lights and the status of sign lines, with a total of two sets of data, and the score of each type of data is 50.000 points; The light impact level includes five groups of data, including road type, road speed limit, number of road lanes, emergency equipment conditions, and temperature conditions. The score of each type of data is 20,000 points. The general impact levels include weather conditions, road slope, road flatness, road skid resistance, number of historical traffic accidents, traffic accident severity, lane change frequency and overtaking frequency, a total of 8 groups of data, and the score of each type of data is 12.500 points; The severe impact level includes visibility status, sections with high incidence of traffic accidents, vehicle tire pressure status, vehicle brake health status, vehicle load, vehicle actual speed, braking force, and acceleration at intersections, a total of 8 groups of data, and the score of each type of data is 12,500 points.
[0040] The level of each data type is determined by comparing the characteristic value with the standard value. For example, the vehicle speed status uses the current road speed limit as the baseline. Speeds within the range of 80% to 100% of the speed limit are excellent, speeds exceeding the speed limit by up to 105% are fair, speeds exceeding the speed limit by 105% to 110% are cautionary, and speeds exceeding the speed limit by 110% are dangerous. Other data types are preset in the same way.
[0041] Among them, the excellent score percentage is 100%, the general score percentage is 60%, the attention score percentage is 300%, and the dangerous score percentage is 10%.
[0042] By determining the level to which various types of data belong, calculating their specific scores, and summing up the scores for each level of influence, the total safety score is obtained through weighted calculation.
[0043] Assume that the total safety score is calculated 10 minutes after the vehicle departs: No impact level:
[0044] Mild impact level:
[0045] General impact level:
[0046] Severe impact level:
[0047] Total safety score:
[0048] At this time, the total safety score of 65.75 is lower than the safety threshold score of 70 points. The safety reminder information will be transmitted to the vehicle display, and the data type currently at the attention level will be displayed to provide special reminders to the driver; the safety reminder information will be synchronously transmitted to the command center, and the staff of the command center will contact the vehicle driver to manually remind him to drive safely until the data type score at the attention level is improved to the excellent or general level and the total safety score rises above the safety threshold score; At this time, the information prediction module predicts the total safety score and driving score for the next 1 minute, 5 minutes, 10 minutes and 30 minutes, and transmits them synchronously to the vehicle display and command center to provide early warning for future safe driving.
[0049] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the above-mentioned boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the system diagram of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. The optional embodiment is expected, wherein the order of various operations is changed and the sub-operation described as a part of a larger operation is performed independently.
[0050] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified to the contrary, one or more of the above-mentioned functions and / or features can be integrated into a single physical device and / or software module, or one or more functions and / or features can be implemented in a separate physical device or software module. It can also be understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present invention. More specifically, in view of the properties, functions and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention without excessive experimentation using ordinary techniques. It can also be understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention.
[0051] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the above methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0052] The logic and / or steps represented in the system diagram or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0053] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and editable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the above-mentioned program is printed, since the above-mentioned program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or processing in other suitable ways as necessary, and then stored in a computer memory.
[0054] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0055] In the above description of this specification, the description with reference to the terms "one embodiment / example", "another embodiment / example" or "certain embodiments / examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.
[0056] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention.
[0057] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention, and these equivalent modifications or substitutions are all included in the scope defined in this application.
Claims
1. A safety monitoring system for a powder and particle material transport trailer based on a neural network, characterized in that: include An information acquisition module, used to acquire road information including traffic flow data, road facility data, and road history data within a predetermined road section, vehicle information including vehicle driving status data and driver driving habit data, and meteorological data including weather conditions, temperature conditions, and visibility conditions within a predetermined road section; An information analysis module, which is connected to the information acquisition module, uses neural network technology to summarize, score and weight the acquired information to determine the total safety score including the road safety value and the vehicle safety value; An information storage module, which is connected to the information acquisition module and the information analysis module, stores information according to the mapping relationship between road information, vehicle information and total safety score, and constructs a deep database and a shallow database respectively based on the preset time segment as the division standard; The information prediction module is connected to the information analysis module and the information storage module. It uses neural network technology to extract key features of data in the shallow database, predicts and judges the security status of the future preset time period, and gives a corresponding total security score; The information distribution module is connected to the information analysis module and the information prediction module to distribute the current total safety score and the total safety score for a preset time period in the future to designated vehicles and the command center.
2. The safety supervision system for powder and particle material transport trailer based on neural network according to claim 1 is characterized in that: The information acquisition module at least includes A high-definition camera is installed on a predetermined road section and is used to collect high-definition image information within the predetermined road section; A microwave detector installed on a predetermined road section is used to detect the number of vehicles, vehicle speed, traffic composition and lane occupancy in the predetermined road section; The geomagnetic detector is installed on the predetermined road section to identify vehicles on the lane; Visibility detection equipment, used to detect visibility information of a predetermined road section; The weather warning interface module is used to access the national meteorological system to obtain weather information for the predetermined road section.
3. The safety monitoring system for powder and particle material transport trailer based on neural network according to claim 2 is characterized in that: The traffic flow data includes the number of vehicles, vehicle speed, traffic composition and lane occupancy within a predetermined road section; Road facility data include road type, road speed limit, number of road lanes, road slope, road flatness, road skid resistance, number of street lights, road marking conditions and emergency equipment conditions; Road history data includes the number of historical traffic accidents, traffic accident severity and high-incidence sections of traffic accidents within the predetermined road section; Vehicle driving status data includes vehicle tire pressure status, vehicle brake health status, vehicle load and vehicle actual speed; The driver's driving habit data includes the speed distribution, acceleration at intersections, braking force, lane change frequency and overtaking frequency of the driver of a specific vehicle during the entire driving of the predetermined road section.
4. The safety monitoring system for powder and particle material transport trailer based on neural network according to claim 1 is characterized in that: The information acquisition module uploads the information to the information analysis module, uses neural network technology to extract and classify the features of the data information, and performs weighted scoring processing on different types of data according to the preset scoring strategy: The influence level is divided into four levels according to preset rules: no influence, slight influence, general influence and major influence, with the weights preset as 0%, 15%, 30% and 55% respectively; According to the preset stratification rules, various types of data are divided into various influence levels, the number of data types contained in each influence level is counted, and the total score of the influence level is evenly assigned to various types of data; The scoring levels of various types of data are divided into four levels: excellent, general, caution and dangerous according to preset thresholds. The scores obtained for each scoring level are 100%, 60%, 30% and 10% of the scores assigned to the data of this type.
5. The safety monitoring system for powder and particle material transport semitrailer based on neural network according to claim 4 is characterized in that: The influence level division is managed in a floating manner according to the law of information change in the deep database. Within a preset period of time, when the proportion of a certain data type in the proportion of traffic accident factors falls below a threshold, it will be downgraded from its influence level to the next level. When the proportion of a certain data type in the proportion of traffic accident factors rises above a threshold, it will be upgraded from its influence level to the previous level. Within the preset period of time, when the proportion of a certain influence level in the proportion of the levels of factors causing traffic accidents drops below the threshold, the weight of that influence level will be reduced by 3%. At the same time, if the proportion of other influence levels in the proportion of the levels of factors causing traffic accidents rises above the threshold, the 3% weight will be fully assigned to that influence level. If no other influence level rises above the threshold, the 3% weight will be divided according to the current weight ratios of other influence levels and assigned to other influence levels respectively.
6. The safety monitoring system for powder and particle material transport semitrailer based on neural network according to claim 5 is characterized in that: The proportion of the types of factors causing traffic accidents refers to: the ratio of the number of times a certain type of data is assigned a dangerous scoring level when a traffic accident occurs to the total number of traffic accidents within a preset period of time; The proportion of the level of factors causing traffic accidents refers to: within a preset period of time, the proportion of all types of data in a certain influence level that are assigned to dangerous levels to the total number of data types.
7. The safety monitoring system for powder and particle material transport trailer based on neural network according to claim 1 is characterized in that: The deep database contains all the data since the system started running. The deep database is in a dormant state to save energy loss and bandwidth. After a transport is completed, all the data of the transport is packaged and stored. The shallow database contains all data within a preset period of time before this transportation, and the shallow database is in an activated state for real-time reading and storage.
8. The safety monitoring system for powder and particle material transport semitrailer based on neural network according to claim 7 is characterized in that: Based on the data in the shallow database, neural network technology is used to extract time features, habit features and scoring features to form a dual-track curve graph model, which is then fed into the neural network for training to predict the total safety score of this transportation at a certain time node in the future. When the predicted habit score and total safety score are lower than the safety threshold, early warning reminder information is issued to designated vehicles and command centers in advance.
9. The safety monitoring system for powder and particle material transport semitrailer based on neural network according to claim 8 is characterized in that: The habit feature is a score assigned to the driving habit type data of a specific vehicle driver.
10. The safety monitoring system for powder and granular material transport trailer based on neural network according to claim 5 is characterized in that: The deep database is woken up according to the preset cycle to extract relevant data to calculate the type ratio of traffic accident cause factors and the level ratio of traffic accident cause factors. After adjusting the data type level division and influence level weight, the adjustment items are recorded in the change log.
Citation Information
Patent Citations
Body and rule-based logistic transport prewarning method
CN102722803A
Early warning method for logistics transportation vehicle
CN109409768A
Road risk assessment method based on analytic hierarchy process
CN110705850A
Highway safety vehicle-mounted intelligent system based on machine learning principle and working process
CN111784873A
Mine car safety evaluation and prediction method, terminal equipment and storage medium
CN112668786A
