Logistics transportation vehicle motion situation danger sensing and monitoring method and system

Abstract

The invention discloses a logistics transportation vehicle motion situation danger sensing and monitoring method and system. The method comprises the steps that S1, equipment protection early warning is carried out before a transportation vehicle driving task; s2, constructing an operation out-of-control early warning model to perform operation out-of-control early warning in the transport vehicle driving task, acquiring working state data of monitoring equipment corresponding to the equipment monitoring module set at the research moment in the transport vehicle driving task in real time, and inputting the working state data into the operation out-of-control early warning model to obtain accident items and accident grades predicted at the research moment; and S3, arranging an obstacle monitoring device at the monitoring position of the transport vehicle, performing target sensing on obstacles around the transport vehicle, performing real-time measurement to obtain distance data from the monitoring position to the obstacles around the transport vehicle, and performing collision risk judgment. According to the invention, danger is sensed in advance through data processing of historical monitoring data, early notification and early warning before dangerous accidents occur are realized, and safe driving of logistics vehicle drivers and logistics transportation safety management are facilitated.

Description

Technical Field

[0001] The present invention relates to the field of risk management and control of logistics vehicles, and in particular to a method and system for perceiving and monitoring the movement hazard of logistics transportation vehicles. Background Art

[0002] With the development of road and railway vehicle equipment technology and logistics transportation technology, the use of vehicles for logistics transportation is an important part of logistics (including cargo transportation logistics). Therefore, logistics vehicles (mainly trains on railways, trucks on roads, and also airplanes for air transportation and ships for water transportation) are important facilities for realizing current logistics transportation. Freight logistics transportation will frequently use logistics vehicles (logistics vehicles refer to power vehicles involved in cargo logistics transportation). Logistics vehicles rely on drivers to drive and travel. Generally, a driving recorder is installed on the vehicle to record the driving process, and front and rear cameras (some will also install cameras on the left and right sides) to detect the surroundings of the vehicle. Due to the long driving distance and complex driving conditions of logistics vehicles (the road environment has different levels of highways or mountain roads, and different sections or weather will cause depressions, etc.), the safe driving of logistics vehicles becomes very important. The traditional method is to install monitoring equipment such as cameras and radars on the vehicle, and then monitor the environment around the vehicle to feedback the surrounding conditions of the vehicle to the driver (for example, the reversing radar image shows the camera situation behind the reversing). Although the traditional method has played a certain role in safety auxiliary operations, it still has many shortcomings. With the development of big data processing technology, how to collect process data based on the monitoring equipment or modules configured on logistics vehicles, archive the data to form historical monitoring data and use it for accident warning prompts or equipment warning prompts, and perceive danger in advance through data processing of historical monitoring data (especially abnormal data changes before the danger occurs). To achieve advance notification and warning before dangerous accidents occur in various specific transportation scenarios, it will be beneficial to the safe driving of logistics vehicle drivers and the safety management of logistics transportation, and it is also an important direction of logistics transportation research. Summary of the invention

[0003] The purpose of the present invention is to solve the technical problems pointed out by the background technology, and to provide a method and system for perceiving and monitoring the movement situation of logistics transportation vehicles; to construct a set of equipment monitoring modules for each equipment of the transportation vehicle, to monitor the equipment status of each equipment and to issue equipment early warning prompts; to obtain the working status data of the equipment monitoring module set corresponding to the monitoring equipment at the study moment in the transportation vehicle driving task in real time based on the equipment monitoring module set, and to input the operation out of control early warning model to obtain the accident items and accident classification predicted at the study moment; to set a monitoring position at the edge of the transportation vehicle contour and / or the protruding position, to set an obstacle monitoring device at the monitoring position and to perform target perception of obstacles around the transportation vehicle, to obtain the distance data of the monitoring position from the obstacles around the transportation vehicle in real time by measuring, and to make real-time judgment of collision risk through the collision and scratch algorithm model.

[0004] The purpose of the present invention is achieved through the following technical solutions:

[0005] A method for sensing and monitoring the movement situation danger of a logistics transportation vehicle, the method comprising:

[0006] S1. Equipment protection warning before the transport vehicle travels: Set the equipment set required to be monitored for the transport vehicle. The equipment set includes several equipment required to be monitored in the transport vehicle. Set equipment monitoring modules for each equipment in the transport vehicle and combine them to form an equipment monitoring module set. The equipment monitoring modules in the equipment monitoring module set monitor the equipment status of the equipment in the equipment set and issue equipment warning prompts.

[0007] S2. Construct an operation out of control warning model to warn about operation out of control in the driving task of transport vehicles, collect an operation out of control accident sample database, the operation out of control accident sample database includes accident labels and working status data sets corresponding to the accident labels, the working status data in the working status data sets are obtained by monitoring the working status of the equipment in the equipment set using the equipment monitoring module set during the driving of the transport vehicle, and the accident label includes accident items and accident classifications corresponding to the accident items; input the operation out of control accident sample data in the operation out of control accident sample database into the operation out of control warning model for model training, obtain the working status data of the monitoring equipment corresponding to the equipment monitoring module set at the research moment in the driving task of the transport vehicle in real time, and input it into the operation out of control warning model to obtain the accident items and accident classifications predicted at the research moment;

[0008] S3. Set a monitoring position at the edge of the transport vehicle contour or / and the raised position, set an obstacle monitoring device at the monitoring position and use it to sense the obstacles around the transport vehicle, measure in real time to obtain the distance data between the monitoring position and the obstacles around the transport vehicle, build a collision and scratch algorithm model and make the following collision risk judgment:

[0009] S31. If the distance value measured by the obstacle monitoring device at the monitoring position is less than the threshold value A, and the distance value gradually decreases within the range of the threshold value A, a collision risk warning is issued and continuous tracking and monitoring is performed; if during the continuous tracking and monitoring process, the monitored transport vehicle speed is greater than V1, the transport vehicle is linked to perform emergency braking and / or emergency control processing.

[0010] In order to better realize the logistics transportation vehicle movement situation hazard perception monitoring method of the present invention, a multi-point algorithm model for transportation vehicle collision and scratching and a multi-point accident database for collision and scratching are constructed. The multi-point accident database for collision and scratching is based on the accident label and the obstacle distance data obtained by each monitoring position. The multi-point accident data of collision and scratching in the operation out of control accident sample database is input into the multi-point algorithm model for transportation vehicle collision and scratching for model training; the obstacle distance data obtained by each monitoring position of the transportation vehicle is obtained in real time and input into the multi-point algorithm model for transportation vehicle collision and scratching to evaluate and obtain predicted accidents and accident probabilities.

[0011] Preferably, step S3 also includes the following method:

[0012] S32. A plurality of displacement sensors for monitoring the sliding or movement of the loaded cargo are arranged inside the transport vehicle. If the displacement sensors detect that the loaded cargo moves and the accumulated movement distance is greater than a threshold value B, an early warning of the movement of the loaded cargo is issued.

[0013] Preferably, in step S1, the equipment in the equipment set includes a braking system, a steering system, an engine, an electrical system, a suspension system, a fuel system, an air-conditioning system, a cooling system, an exhaust system, and tires. The equipment monitoring modules in the equipment monitoring module set collect or monitor content including equipment service life, theoretical braking distance, current braking distance, last maintenance time, maintenance times, maintenance frequency, equipment brand, and equipment model. The equipment monitoring modules in the equipment monitoring module set set equipment safety standards to perform safety judgments and issue equipment early warning prompts; in step S2, the equipment monitoring modules in the equipment monitoring module set collect or monitor content also including the packaging and fixing method of the transported items, whether a special vehicle is used for escort, the radiation value / stability value / danger value of the transported items, whether the route has been traveled / investigated, daily acceleration and deceleration frequency, weather, temperature, longitude and latitude differences across the operating route, use environment of the transport vehicle, uphill / downhill, average mileage load, and current load. The equipment monitoring module for the tire includes a tire pressure sensor, and the equipment monitoring module for the engine includes a vibration sensor.

[0014] Preferably, the accident items in the accident label include brake failure, flat tire, turning failure, vehicle inability to start, insufficient power, slipping, line failure, battery failure, ABS / electronic stability failure, suspension / spring / shock absorber failure, water pump failure, and heat dissipation failure, and the accident classification is classified into high, medium, and low.

[0015] Preferably, the cumulative risk assessment of the equipment set of the transport vehicle in step S1 is performed according to the following method:

[0016] The risk value of a device set is equal to the sum of the risk values ​​of each device in the device set;

[0017] The equipment risk value is obtained as follows:

[0018] Equipment risk value = severity score of equipment-related failure x (failure probability + risk precursor recognition degree), where the severity score of equipment-related failure is a severity score calculated based on the equipment status of the equipment monitored by the equipment monitoring module, and the failure probability is the probability of occurrence based on the equipment status of the equipment monitored by the equipment monitoring module. The risk precursor recognition degree formula is as follows: Where y represents the sign recognition degree, x represents the current observation value, a represents the preset total value, and b represents the correction number.

[0019] Preferably, the operation out-of-control warning model uses the following method to evaluate the cumulative operation risk of the transport vehicle in step S2 according to the accident items:

[0020] The cumulative operational risk value of an accident item = the equipment risk value associated with the accident item + the working status risk value of the monitoring equipment corresponding to the accident item. The working status risk value of the monitoring equipment corresponding to the accident item is a risk value extracted and calculated based on the working status data of the monitoring equipment corresponding to the accident item. The operational out-of-control warning model has a database for risk value assessment of associated equipment according to accident items.

[0021] Preferably, the device monitoring modules in the device monitoring module set are networked by wired or / and wireless communication, and the wireless communication methods include hotspot, Lora, NB-IOT, and Internet of Things short-range communication. In step S31, the distance value measured by a single obstacle monitoring device is as follows:

[0022] The distance value D between the obstacle monitoring device and the target obstacle F , S1 represents the detection distance obtained by the obstacle monitoring device when detecting the target obstacle at time T1, S2 represents the detection distance obtained by the obstacle monitoring device when detecting the target obstacle at time T2, time T1 and time T2 are adjacent time points and are micro time periods, and d represents the distance moved by the transport vehicle in the time period from time T1 to time T2.

[0023] A logistics transport vehicle motion situation hazard perception monitoring system includes a transport vehicle, a pre-travel mission equipment protection warning system, an operation out of control warning model and a collision and scratch algorithm model. The transport vehicle has or is provided with a set of equipment that needs to be monitored, and the equipment set includes several equipment that needs to be monitored in the transport vehicle. The pre-travel mission equipment protection warning system includes an equipment monitoring module set, and the equipment monitoring module set includes several equipment monitoring modules for monitoring equipment in the transport vehicle. The pre-travel mission equipment protection warning system uses the equipment monitoring module set to monitor the equipment status of the equipment in the equipment set and issues equipment warning prompts; the operation out of control warning model includes an operation out of control accident sample database, and the operation out of control accident sample database includes accident labels and working status data sets corresponding to the accident labels. The working status data in the working status data set is The accident labels obtained by monitoring the working status of the equipment in the equipment set using the equipment monitoring module set during the driving of the transport vehicle include accident items and accident classifications corresponding to the accident items; the operation out of control accident sample data in the operation out of control accident sample database is input into the operation out of control warning model for model training, the working status data of the monitoring equipment corresponding to the equipment monitoring module set at the research moment in the transport vehicle driving task is obtained in real time and input into the operation out of control warning model to obtain the accident items and accident classifications predicted at the research moment; a monitoring position is set at the edge and / or raised position of the transport vehicle contour, an obstacle monitoring device is set at the monitoring position and used for target perception of obstacles around the transport vehicle, and real-time measurement is performed to obtain the distance data between the monitoring position and the obstacles around the transport vehicle, and the collision and scratching algorithm model obtains the distance data between the monitoring position and the obstacles around the transport vehicle and performs the following collision risk judgment:

[0024] If the distance value measured by the obstacle monitoring device at the monitoring position is less than the threshold value A, and the distance value gradually decreases within the range of threshold value A, a collision and scratch danger warning is issued and continuous tracking and monitoring is carried out; if during the continuous tracking and monitoring process, the monitored transport vehicle speed is greater than V1, the transport vehicle is linked to emergency braking and / or emergency control processing.

[0025] Preferably, a number of displacement sensors for monitoring the sliding or movement of loaded goods are arranged inside the transport vehicle. If the displacement sensor detects that the loaded goods are moving and the cumulative movement distance is greater than a threshold value B, an early warning of the movement of the loaded goods is issued. The collision and scratch algorithm model also includes a multi-point algorithm model for collision and scratching of the transport vehicle. The multi-point algorithm model for collision and scratching of the transport vehicle has a multi-point accident database for collision and scratching. The multi-point accident database for collision and scratching inputs the multi-point accident data in the operation out of control accident sample database into the multi-point algorithm model for collision and scratching of the transport vehicle for model training, and obtains the obstacle distance data obtained at each monitoring position of the transport vehicle in real time and inputs it into the multi-point algorithm model for collision and scratching of the transport vehicle for evaluation to obtain predicted accidents and accident probabilities.

[0026] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0027] (1) The present invention constructs an equipment monitoring module set for each equipment of the transport vehicle, monitors the equipment status of each equipment and issues equipment warning prompts; the present invention obtains the working status data of the equipment monitoring module set corresponding to the monitoring equipment at the study time during the transport vehicle driving task in real time based on the equipment monitoring module set and inputs the operation out of control warning model to obtain the accident items and accident classification predicted at the study time; the present invention sets a monitoring position at the edge of the transport vehicle contour and / or the protruding position, sets an obstacle monitoring device at the monitoring position and uses it to perceive the obstacles around the transport vehicle, obtains the distance data of the monitoring position from the obstacles around the transport vehicle in real time by measuring, and makes real-time judgment of the collision risk through the collision and scratch algorithm model.

[0028] (2) The multi-point algorithm model for collision and scratching of transport vehicles of the present invention is trained through a multi-point accident database for collision and scratching. The obstacle distance data obtained from each monitoring position of the transport vehicle is acquired in real time and input into the multi-point algorithm model for collision and scratching of transport vehicles to evaluate and obtain predicted accidents and accident probabilities.

[0029] (3) The present invention can perceive danger in advance through data processing of historical monitoring data, and realize advance notification and warning before dangerous accidents occur, which is beneficial to the safe driving of logistics vehicle drivers and the safety management of logistics transportation. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a method flow chart of the method for monitoring the movement situation danger of logistics transportation vehicles of the present invention;

[0031] Figure 2 This is a principle structure diagram of the logistics transportation vehicle movement situation danger perception monitoring system of the present invention;

[0032] Figure 3 It is a schematic diagram of an obstacle monitoring device being arranged at a contour edge and / or a raised position of a transport vehicle in an embodiment;

[0033] Figure 4 for Figure 3 Schematic diagram of a transport vehicle equipped with an obstacle monitoring device to sense a target and measure the distance value from the target obstacle. DETAILED DESCRIPTION

[0034] The present invention is further described in detail below in conjunction with embodiments:

[0035] Example

[0036] like Figure 1 As shown, a method for sensing and monitoring the movement danger of a logistics transportation vehicle comprises:

[0037] S1. Equipment protection warning before the transport vehicle travels: Set up a set of equipment that needs to be monitored for the transport vehicle. The equipment set includes several equipment that needs to be monitored in the transport vehicle. Set up equipment monitoring modules for each equipment in the transport vehicle and combine them to form an equipment monitoring module set; the equipment monitoring module (such as a sensor) plus a common quick installation method (including but not limited to screws, hooks, etc.) can be installed on the transport vehicle (such as a vehicle or train, etc.) and the transported items (various large and small pieces, metal and non-metal parts, etc.); thereby realizing convenient wireless deployment and application of local risk control object areas in different environments, different states, different states, and different locations. The equipment monitoring module in the equipment monitoring module set monitors the equipment status of the equipment in the equipment set and issues an equipment warning prompt. The equipment monitoring module in the equipment monitoring module set uses wired or / and wireless communication methods for communication networking, and the wireless communication methods include hotspot methods, Lora, NB-IOT, and short-distance communication methods of the Internet of Things; the present invention can construct a remote server for big data processing, and the transport vehicle communicates with the remote server through a remote mobile communication method.

[0038] S2. Construct an operation out of control warning model to warn of operation out of control in the driving mission of transport vehicles, collect a sample database of operation out of control accidents, the sample database of operation out of control accidents includes accident labels and working status data sets corresponding to the accident labels, the working status data in the working status data sets are obtained by monitoring the working status of the equipment in the equipment set using the equipment monitoring module set during the driving of the transport vehicle, and the accident labels include accident items and accident classifications corresponding to the accident items; input the operation out of control accident sample data in the sample database of operation out of control accidents into the operation out of control warning model for model training, obtain the working status data of the monitoring equipment corresponding to the equipment monitoring module set at the study moment in the driving mission of the transport vehicle in real time, and input them into the operation out of control warning model to obtain the accident items and accident classifications predicted at the study moment.

[0039] In some embodiments, the accident items in the accident label include brake failure, flat tire, turning failure, vehicle inability to start, insufficient power, slipping, line failure, battery failure, ABS / electronic stability failure, suspension / spring / shock absorber failure, water pump failure, and heat dissipation failure. The present invention classifies the accident items and grades the accidents according to the accident items. The accident grades are classified into high, medium, and low.

[0040] S3. Set a monitoring position at the edge of the transport vehicle contour and / or the raised position, set an obstacle monitoring device at the monitoring position (the obstacle monitoring device uses a ranging radar or a microwave radar) and use it to sense the obstacles around the transport vehicle, obtain the distance data of the monitoring position from the obstacles around the transport vehicle in real time, build a collision and scratch algorithm model and make the following collision risk judgment (pre-judge whether a collision will occur in the current situation):

[0041] S31. If the distance value measured by the obstacle monitoring device at the monitoring position is less than the threshold value A, and the distance value gradually decreases within the threshold value A, a collision and scratching danger warning is issued and continuous tracking and monitoring is performed; if the speed of the monitored transport vehicle is greater than V1 during continuous tracking and monitoring, the transport vehicle is controlled to perform emergency braking or / and emergency control processing. In some embodiments, the distance value measured by a single obstacle monitoring device is as follows:

[0042] The distance value D between the obstacle monitoring device and the target obstacle F , S1 represents the detection distance obtained by the obstacle monitoring device when detecting the target obstacle at time T1, S2 represents the detection distance obtained by the obstacle monitoring device when detecting the target obstacle at time T2, time T1 and time T2 are adjacent time points and are micro-time periods, and d represents the distance moved by the transport vehicle in the time period from time T1 to time T2.

[0043] See also Figure 3 An obstacle monitoring device is installed at the installation position outside the transport vehicle (because the loaded cargo may be large-scale transport parts with irregular shapes, part of which may be displaced from the side), and the driving direction is taken as an example of the relative displacement direction. The obstacle monitoring device moves with the transport vehicle. The obstacle monitoring device detects the target "T" at position A (i.e. Figure 4 The obstacle monitoring device measures the distance between position A and the perceived target as S1. The obstacle monitoring device moves with the transport vehicle for a short period of time (the corresponding distance is d, d = V*t, V represents the real-time calculated speed of the vehicle, and t represents the short period of time the vehicle moves) to position B. At this time, the obstacle monitoring device measures the distance between position B and the perceived target as S2, which can be obtained. Figure 4 The middle angle α, Then we can get D F =S1sinα. The distance between the future object and the perceived target can be judged at each instant (micro time t). When this predicted distance keeps decreasing, or even reaches 0 or a negative value, it can be judged in advance that the movement will have a scratch (when it is 0) or a collision (when it is a negative value), and ultimately achieve early warning.

[0044] In some embodiments, in addition to the above-mentioned single monitoring position real-time collision risk judgment during driving, the present invention can also combine multiple monitoring positions to form a multi-point early warning collision risk judgment during driving, as follows: for transport vehicles, a multi-point algorithm model for transport vehicle collisions and scratches and a multi-point accident database for collisions and scratches can be constructed, and historical data of accidents caused by obstacle distance data collected by multiple points are used as sample data for model learning and training, and then based on the currently collected obstacle distance data, accidents and accident probabilities are predicted (especially based on the multi-point obstacle distance data before the accident with historical data of accidents, model learning and training are performed to facilitate early judgment of the probability of accidents caused by currently collected obstacle distance data). The collision and scratches multi-point accident database inputs the collision and scratches multi-point accident data in the operation out of control accident sample database into the transport vehicle collision and scratches multi-point algorithm model for model training according to the accident label and the obstacle distance data obtained by each monitoring position; the obstacle distance data obtained by each monitoring position of the transport vehicle is obtained in real time and input into the transport vehicle collision and scratches multi-point algorithm model to evaluate the predicted accidents and accident probabilities. The present invention collects data in real time, trains a large model, issues early warning and provides real-time guidance for correction. Specifically, when an abnormal problem or even an accident occurs (such as scratching, displacement of transported items or even collision during transportation), the data before the accident or problem (corresponding to changes in sensor data) can be retrieved for learning, thereby realizing the function of early judgment and warning of approaching danger, and ultimately improving the level of risk management and control of vehicles and their transported items during transportation.

[0045] The present invention also includes the following method:

[0046] S32. Several displacement sensors are set inside the transport vehicle to monitor the sliding or movement of the loaded goods. If the displacement sensor detects that the loaded goods are moving and the cumulative movement distance is greater than the threshold value B, an early warning of the movement of the loaded goods is issued. The loaded goods are fixed inside the transport vehicle. If the transport distance of the transport vehicle is long, the loaded goods may move due to road bumps or loose fixation, which is difficult to check during transportation. Therefore, this application specifically sets displacement sensors at key positions of the loaded goods to monitor the cumulative movement distance of the loaded goods and issue early warning prompts (to prevent the loaded goods from moving or colliding inside the vehicle, causing damage to the goods).

[0047] In this embodiment, examples of monitored equipment in transport vehicles are as follows: the equipment in the equipment set includes braking system, steering system, engine, electrical system, suspension system, fuel system, air conditioning system, cooling system, exhaust system, and tires. The equipment monitoring module in the equipment monitoring module set collects or monitors content including equipment service life, theoretical braking distance, current braking distance, last maintenance time, maintenance times, maintenance frequency, equipment brand, and equipment model. The equipment monitoring modules in the equipment monitoring module set set equipment safety standards to perform safety judgments and issue equipment early warning prompts (judging the probability of equipment failure through equipment maintenance, vehicle failure big data, and related component usage environment data, so that personnel can make predictions). Examples of corresponding equipment monitoring modules are as follows: The equipment monitoring modules in the equipment monitoring module set also collect or monitor content including the packaging and fixing method of the transported items, whether a special vehicle is used for escort, the radiation value / stability value / danger value of the transported items, whether the route has been traveled / investigated, daily acceleration and deceleration frequency, weather, temperature (based on speed, temperature, driving operation response acceleration and environmental monitoring data, determine the controlled state of the vehicle and alarm for uncontrolled equipment), the difference in longitude and latitude across the operating route, the use environment of the transport vehicle, uphill / downhill, average mileage load, and current load. The equipment monitoring module for tires includes a tire pressure sensor, and the equipment monitoring module for engines includes a vibration sensor. The equipment listed above is the core monitoring equipment in transport vehicles. There is an equipment monitoring module corresponding to each equipment (the equipment monitoring module can be a separate monitoring sensor, such as a tire pressure sensor, a vibration sensor, etc., or it can be a data acquisition module placed in the transport vehicle to collect the braking distance and fuel volume of the braking system. It can also be a data acquisition module for collecting or collecting weather, temperature, etc.). In actual use, each equipment monitoring module in the equipment monitoring module set collects and obtains the required factor data (such as the packaging and fixing method of the transported items listed above, the radiation value / stability value / danger value of the transported items, weather, temperature, the difference in longitude and latitude across the operating route, the average course load, the current load, etc. Based on the above data, learn and extract related factors). The equipment monitoring module set jointly collects and obtains a factor database for orderly accident risk and safety assessment, which is convenient for subsequent model training and prediction.At the software level, through the continuous collection and training of data, a big data early perception system is eventually formed that integrates changes in various sensors and input condition data. The data will become more and more accurate in judging future dangerous situations. At the same time, at the hardware and external system docking level, it integrates multiple (Beidou positioning position, displacement, multi-axis acceleration, vibration, radar, tilt, vision (driver, vehicle), pressure (tire pressure, surface squeezing force of objects) or temperature (temperature changes before an accident, brake pads, differences in winter and summer), etc.) sensors and external systems and information (such as satellite remote sensing data (early warning), wind speed and direction (source weather data) data, meteorological radar temperature and humidity weather data, mountain landslide risk data, etc.). These systems will continue to evolve, and with the development of technology, new technical methods and sensor hardware will continue to emerge. These will allow the entire system to evolve and grow continuously, and eventually form a sustainable evolving and expanding system.

[0048] According to the analysis of the causes of serious and major road traffic accidents from 2007 to 2019 (Wu Yuanwei, Information Research Institute of the Ministry of Emergency Management, China Safety Production Magazine): there were 228 serious and major road traffic accidents during this period. The number one cause was improper operation or improper emergency handling in 85 cases, accounting for 37.3%; various vehicle safety hazards, including poor braking, tire blowouts, and special system problems, totaled 71 cases, accounting for 31.1%; special road conditions such as steep slopes, sharp turns, continuous bends, near water or cliffs, or slippery roads after rain and snow, and easy to ice sections accounted for 29 cases, accounting for 12.7%. In order to find out the root cause, the following measures can be taken to perceive and avoid it in advance:

[0049] Drivers cannot perceive the surroundings, bad weather, road conditions, intersection conditions or the environment, and cannot make accurate judgments: for example, in the transportation of large items, they cannot make sufficient judgments on materials extending outside the vehicle body, their own fatigue driving conditions, invisible environmental risks such as clouds and fog, and pedestrians in blind spots and blind spots. Visual sensing equipment can be added to assist the driver in making a more accurate understanding and judgment of the current transportation situation through the present invention; the network weather can be connected, including the landslide and mud-rock flow data in local mountains for more accurate prediction and judgment;

[0050] Insufficient understanding of the vehicle's own safety hazards, usage consumption and wear and tear can also be achieved through sensors, such as multi-directional acceleration sensors that can be installed in various places on the vehicle, including the front, body and load, combined with instructions and feedback data from brake sensors, relative displacement acceleration of the load, brake pad temperature sensor, Beidou positioning data, temperature and humidity, tire pressure, weather forecast, history and current turning / braking controlled curve changes, whether the vehicle can be controlled at certain speeds, certain distances, and certain slippery ground conditions, as well as the radius of sharp turns can be predicted in advance; through the prediction results of the present invention, before driving, combined with the vehicle's own parameters (including illegal overloading under actual acceleration), recommendations on driving speed, driving route and loading tonnage are made; during driving, warnings are given before each risk positioning point such as mountain roads and routes, and accurate reminders of the recommended acceleration and deceleration speed control can be made; after driving is completed, possible component wear and consumption conditions are prompted.

[0051] In some embodiments, the present invention can also perform cumulative risk assessment on the equipment set of the transport vehicle in step S1 according to the following method:

[0052] The risk value of a device set is equal to the sum of the risk values ​​of each device in the device set;

[0053] The equipment risk value is obtained as follows:

[0054] Equipment risk value = severity score of equipment-related failure x (failure probability + risk precursor recognition degree), where the severity score of equipment-related failure is a severity score calculated based on the equipment status of the equipment monitored by the equipment monitoring module, and the failure probability is the probability of occurrence based on the equipment status of the equipment monitored by the equipment monitoring module. The risk precursor recognition degree formula is as follows: Where y represents the sign recognition degree, x represents the current observation value (in terms of service life, the service life of a certain equipment is 3 years and 1095 days, and it has been used for 852 days. Therefore, the current observation value of the equipment in the service life item review is 852 days), a represents the preset total value (in terms of service life, the service life of a certain equipment is 3 years and 1095 days, and it has been used for 852 days. Therefore, the preset total value of the equipment in the service life item review is 1095 days), and b represents the correction number (or the cumulative number of abnormal risks), which is calculated using the following formula: b=δ*n, δ represents the risk value of a single abnormality (the value range is 0.01~0.2), and n represents the cumulative number of abnormalities. In this way, the cumulative risk of abnormalities detected by the equipment can be calculated.

[0055] The present invention can collect data, train a large model, give early warning and guide correction in real time; specifically, when a fault occurs (accident reporting is performed in the supporting software), the collection and extraction module will automatically retrieve the parameters of the corresponding key components for learning, so as to correct the fault possibility (foreshadowing frequency) according to its actual description; at the same time, the changes in other related parameters are incorporated into the risk sign recognition judgment parameters; if the brake failure is registered in the system, the system extracts the life data in its TPM configuration parameters, and obtains the possible fault possibility ratio of the parts in the form of fault time ÷∑ part working time; then the time curve of the braking to stopping operation during this period is compared horizontally, and the risk sign recognition parameter of the vehicle with a similar change curve is corrected (such as the initial possibility parameter is 1, which is modified to the current "failure time ÷∑ part working time" calculation result), and finally the possibility parameter of the problem is adjusted, the risk level indicator is updated, and the model is evolved. During the use of transport vehicles, especially before losing control, the data related to the out-of-control state such as brakes and tire pressure will change gradually, rather than suddenly. The probability of vehicle out-of-control is different for different brands, weather, temperature, environment and routes. The model of the present invention collects and enters all possible parameters, and according to the risk composition formula F (x) =ΣX is used for identification and learning, and all related data are captured and recorded.

[0056] In some embodiments, in addition to the operation out of control warning, the present invention can also perform operation cumulative risk assessment. The operation out of control warning model uses the following method to evaluate the operation cumulative risk of the transport vehicle in step S2 according to the accident item:

[0057] The cumulative operational risk value of an accident item = the equipment risk value associated with the accident item + the working status risk value of the monitoring equipment corresponding to the accident item. The working status risk value of the monitoring equipment corresponding to the accident item is a risk value extracted and calculated based on the working status data of the monitoring equipment corresponding to the accident item. The operational out-of-control warning model has a database for risk value assessment of associated equipment according to accident items.

[0058] like Figure 2As shown, a logistics transport vehicle movement situation hazard perception monitoring system includes a transport vehicle, a pre-travel mission equipment protection warning system, an operation out of control warning model and a collision and scratch algorithm model. The transport vehicle has or is provided with a set of equipment that needs to be monitored, and the equipment set includes several equipment that needs to be monitored in the transport vehicle. The pre-travel mission equipment protection warning system includes an equipment monitoring module set, and the equipment monitoring module set includes several equipment monitoring modules for monitoring equipment in the transport vehicle. The pre-travel mission equipment protection warning system uses the equipment monitoring module set to monitor the equipment status of the equipment in the equipment set and issues equipment warning prompts; the operation out of control warning model includes an operation out of control accident sample database, and the operation out of control accident sample database includes accident labels and working status data sets corresponding to the accident labels, and the working status data in the working status data sets The data is obtained by using the equipment monitoring module set to monitor the working status of the equipment in the equipment set during the driving of the transport vehicle, and the accident label includes the accident item and the accident classification corresponding to the accident item; the operation out of control accident sample data in the operation out of control accident sample database is input into the operation out of control warning model for model training, and the working status data of the monitoring equipment corresponding to the equipment monitoring module set at the research moment in the transport vehicle driving task is obtained in real time and input into the operation out of control warning model to obtain the accident item and accident classification predicted at the research moment; a monitoring position is set at the edge and / or protruding position of the transport vehicle contour, and an obstacle monitoring device is set at the monitoring position and used for target perception of obstacles around the transport vehicle, and real-time measurement is performed to obtain the distance data of the monitoring position from the obstacles around the transport vehicle, and the collision and scratching algorithm model obtains the distance data of the monitoring position from the obstacles around the transport vehicle and performs the following collision risk judgment:

[0059] If the distance value measured by the obstacle monitoring device at the monitoring position is less than the threshold value A, and the distance value gradually decreases within the range of threshold value A, a collision and scratch danger warning is issued and continuous tracking and monitoring is carried out; if during the continuous tracking and monitoring process, the monitored transport vehicle speed is greater than V1, the transport vehicle is linked to emergency braking and / or emergency control processing.

[0060] A number of displacement sensors are arranged inside the transport vehicle to monitor the sliding or movement of loaded goods (the loaded goods may be large-scale transported parts with irregular shapes). If the displacement sensor detects that the loaded goods are moving and the cumulative movement distance is greater than a threshold value B, an early warning of the movement of the loaded goods is issued; the collision and scratch algorithm model also includes a multi-point algorithm model for collision and scratching of transport vehicles. The multi-point algorithm model for collision and scratching of transport vehicles has a multi-point accident database for collision and scratching. The multi-point accident database for collision and scratching inputs the multi-point accident data in the operation out of control accident sample database into the multi-point algorithm model for collision and scratching of transport vehicles for model training, and obtains the obstacle distance data obtained at each monitoring position of the transport vehicle in real time and inputs it into the multi-point algorithm model for collision and scratching of transport vehicles for evaluation to obtain predicted accidents and accident probabilities.

[0061] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for sensing and monitoring the movement danger of logistics transportation vehicles, characterized by: The methods include: S1. Equipment protection warning before the transport vehicle travels: Set the equipment set required to be monitored for the transport vehicle. The equipment set includes several equipment required to be monitored in the transport vehicle. Set equipment monitoring modules for each equipment in the transport vehicle and combine them to form an equipment monitoring module set. The equipment monitoring modules in the equipment monitoring module set monitor the equipment status of the equipment in the equipment set and issue equipment warning prompts. S2. Construct an operation out of control warning model to warn about operation out of control in the driving task of transport vehicles, collect an operation out of control accident sample database, the operation out of control accident sample database includes accident labels and working status data sets corresponding to the accident labels, the working status data in the working status data sets are obtained by monitoring the working status of the equipment in the equipment set using the equipment monitoring module set during the driving of the transport vehicle, and the accident label includes accident items and accident classifications corresponding to the accident items; input the operation out of control accident sample data in the operation out of control accident sample database into the operation out of control warning model for model training, obtain the working status data of the monitoring equipment corresponding to the equipment monitoring module set at the research moment in the driving task of the transport vehicle in real time, and input it into the operation out of control warning model to obtain the accident items and accident classifications predicted at the research moment; S3. Set a monitoring position at the edge of the transport vehicle contour or / and the raised position, set an obstacle monitoring device at the monitoring position and use it to sense the obstacles around the transport vehicle, measure in real time to obtain the distance data between the monitoring position and the obstacles around the transport vehicle, build a collision and scratch algorithm model and make the following collision risk judgment: S31. If the distance value measured by the obstacle monitoring device at the monitoring position is less than the threshold value A, and the distance value gradually decreases within the range of the threshold value A, a collision risk warning is issued and continuous tracking and monitoring is performed; if during the continuous tracking and monitoring process, the monitored transport vehicle speed is greater than V1, the transport vehicle is linked to perform emergency braking and / or emergency control processing.

2. The method for monitoring the movement danger of logistics transportation vehicles according to claim 1 is characterized by: A multi-point algorithm model for transport vehicle collisions and a multi-point accident database for collisions and scrapes are constructed. The multi-point accident database for collisions and scrapes is based on the accident labels and the obstacle distance data obtained from each monitoring position. The multi-point accident data from the operation out of control accident sample database is input into the multi-point algorithm model for transport vehicle collisions and scrapes for model training. The obstacle distance data obtained from each monitoring position of the transport vehicle is obtained in real time and input into the multi-point algorithm model for transport vehicle collisions and scrapes for evaluation to obtain predicted accidents and accident probabilities.

3. The method for monitoring the movement danger of logistics transportation vehicles according to claim 1 is characterized in that: Step S3 also includes the following method: S32. A plurality of displacement sensors for monitoring the sliding or movement of the loaded cargo are arranged inside the transport vehicle. If the displacement sensors detect that the loaded cargo moves and the accumulated movement distance is greater than a threshold value B, an early warning of the movement of the loaded cargo is issued.

4. The method for monitoring the movement danger of logistics transportation vehicles according to claim 1 is characterized in that: In step S1, the equipment in the equipment set includes a braking system, a steering system, an engine, an electrical system, a suspension system, a fuel system, an air-conditioning system, a cooling system, an exhaust system, and tires. The equipment monitoring modules in the equipment monitoring module set collect or monitor content including equipment service life, theoretical braking distance, current braking distance, last maintenance time, maintenance times, maintenance frequency, equipment brand, and equipment model. The equipment monitoring modules in the equipment monitoring module set set equipment safety standards to perform safety judgments and issue equipment early warning prompts; in step S2, the equipment monitoring modules in the equipment monitoring module set collect or monitor content including the packaging and fixing method of the transported items, whether a special vehicle is used for escort, the radiation value / stability value / danger value of the transported items, whether the route has been traveled / investigated, daily acceleration and deceleration frequency, weather, temperature, longitude and latitude differences across the operating route, use environment of the transport vehicle, uphill / downhill, average mileage load, and current load. The equipment monitoring module for tires includes a tire pressure sensor, and the equipment monitoring module for engines includes a vibration sensor.

5. The method for monitoring the movement danger of logistics transportation vehicles according to claim 1 is characterized by: The accident items in the accident label include brake failure, flat tire, turning failure, vehicle inability to start, insufficient power, slipping, line failure, battery failure, ABS / electronic stability failure, suspension / spring / shock absorber failure, water pump failure, and cooling failure. The accident levels are classified into high, medium, and low.

6. The method for monitoring the movement danger of logistics transportation vehicles according to claim 1 is characterized by: The cumulative risk assessment of the equipment set of the transport vehicle in step S1 is performed as follows: The risk value of a device set is equal to the sum of the risk values ​​of each device in the device set; The equipment risk value is obtained as follows: Equipment risk value = severity score of equipment-related failure x (failure probability + risk precursor recognition degree), where the severity score of equipment-related failure is a severity score calculated based on the equipment status of the equipment monitored by the equipment monitoring module, and the failure probability is the probability of occurrence based on the equipment status of the equipment monitored by the equipment monitoring module. The risk precursor recognition degree formula is as follows: Where y represents the sign recognition degree, x represents the current observation value, a represents the preset total value, and b represents the correction number.

7. The method for monitoring the movement danger of logistics transportation vehicles according to claim 6 is characterized by: The operation out-of-control warning model uses the following method to evaluate the cumulative operation risk of the transport vehicle in step S2 according to the accident items: The cumulative operational risk value of an accident item = the equipment risk value associated with the accident item + the working status risk value of the monitoring equipment corresponding to the accident item. The working status risk value of the monitoring equipment corresponding to the accident item is a risk value extracted and calculated based on the working status data of the monitoring equipment corresponding to the accident item. The operational out-of-control warning model has a database for risk value assessment of associated equipment according to accident items.

8. The method for monitoring the movement danger of logistics transportation vehicles according to claim 1 is characterized by: The device monitoring modules in the device monitoring module set are networked by wired or / and wireless communication, and the wireless communication methods include hotspot, Lora, NB-IOT, and short-distance communication of the Internet of Things. In step S31, the distance value measured by a single obstacle monitoring device is as follows: The distance value D between the obstacle monitoring device and the target obstacle F , S1 represents the detection distance obtained by the obstacle monitoring device when detecting the target obstacle at time T1, S2 represents the detection distance obtained by the obstacle monitoring device when detecting the target obstacle at time T2, time T1 and time T2 are adjacent time points and are micro time periods, and d represents the distance moved by the transport vehicle in the time period from time T1 to time T2.

9. A logistics transport vehicle movement situation hazard perception monitoring system, including a transport vehicle, characterized in that: It also includes a pre-driving mission equipment protection warning system, an operation out of control warning model and a collision and scratch algorithm model. The transport vehicle has or is provided with a set of equipment that needs to be monitored. The equipment set includes several equipment that needs to be monitored in the transport vehicle. The pre-driving mission equipment protection warning system includes an equipment monitoring module set. The equipment monitoring module set includes several equipment monitoring modules for monitoring equipment in the transport vehicle. The pre-driving mission equipment protection warning system uses the equipment monitoring module set to monitor the equipment status of the equipment in the equipment set and issue equipment warning prompts; the operation out of control warning model includes an operation out of control accident sample database. The operation out of control accident sample database includes accident labels and working status data sets corresponding to the accident labels. The working status data in the working status data set is the equipment status of the equipment in the driving of the transport vehicle. The monitoring module set monitors the working status of the equipment in the equipment set, and the accident label includes the accident item and the accident classification corresponding to the accident item; the operation out of control accident sample data in the operation out of control accident sample database is input into the operation out of control warning model for model training, and the working status data of the monitoring equipment corresponding to the equipment monitoring module set at the research moment in the transport vehicle driving task is obtained in real time and input into the operation out of control warning model to obtain the accident item and accident classification predicted at the research moment; the monitoring position is set at the edge and / or protruding position of the transport vehicle contour, and an obstacle monitoring device is set at the monitoring position and used for target perception of obstacles around the transport vehicle, and real-time measurement is performed to obtain the distance data of the monitoring position from the obstacles around the transport vehicle, and the collision and scratching algorithm model obtains the distance data of the monitoring position from the obstacles around the transport vehicle and performs the following collision risk judgment: If the distance value measured by the obstacle monitoring device at the monitoring position is less than the threshold value A, and the distance value gradually decreases within the range of threshold value A, a collision and scratch danger warning is issued and continuous tracking and monitoring is carried out; if during the continuous tracking and monitoring process, the monitored transport vehicle speed is greater than V1, the transport vehicle is linked to emergency braking and / or emergency control processing.

10. The logistics transportation vehicle movement situation danger perception monitoring system according to claim 9 is characterized in that: A number of displacement sensors are arranged inside the transport vehicle for monitoring the sliding or movement of the loaded cargo. If the displacement sensor detects that the loaded cargo moves and the cumulative movement distance is greater than a threshold value B, an early warning of the movement of the loaded cargo is issued. The collision and scratch algorithm model also includes a multi-point algorithm model for collision and scratching of the transport vehicle. The multi-point algorithm model for collision and scratching of the transport vehicle has a multi-point accident database for collision and scratching. The multi-point accident database for collision and scratching inputs the multi-point accident data in the collision and scratching sample database into the multi-point algorithm model for collision and scratching of the transport vehicle for model training according to the accident label and the obstacle distance data obtained by monitoring each monitoring position. The obstacle distance data obtained by each monitoring position of the transport vehicle is obtained in real time and input into the multi-point algorithm model for collision and scratching of the transport vehicle for evaluation to obtain predicted accidents and accident probabilities.

Images