Passenger vehicle door opening monitoring system and method

Abstract

The present invention provides a passenger vehicle door opening monitoring system and method, wherein the system includes: a millimeter wave radar module, an infrared human body sensor module, and a control module; when the approach distance is less than or equal to a preset distance threshold, the infrared human body sensor module is controlled to detect whether the approaching object is a living thing; when it is a living thing, the door of the passenger vehicle is controlled to be prohibited from opening; a door opening collision probability map is also constructed; when the passenger vehicle maintains a target state for more than a preset time, the vehicle position of the passenger vehicle is obtained; the door opening collision probability corresponding to the vehicle position is determined from the collision probability map; the radar detection frequency corresponding to the door opening collision probability is determined from a preset radar detection frequency library; the millimeter wave radar module is controlled to detect the approach distance of the approaching object behind the rear of the vehicle at the radar detection frequency. The data processing load of the system is reduced, the impact of bad weather on mobile biological detection is reduced, and the applicability of the system is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile safety control, and in particular to a passenger vehicle door opening monitoring system and method. Background Art

[0002] At present, there are two main solutions to avoid collision accidents caused by vehicle door opening: one is to use cameras and millimeter-wave radar to detect objects behind. When the distance between the moving object and the vehicle body is less than the set threshold, the result is notified to the occupants through a display or voice prompt, and the occupants take action to prevent collision. The other is a solution combining ultrasonic sensors, which uses a combination of multiple ultrasonic sensors to detect objects behind, determine the location of the objects, and issue a warning of door opening based on the distance.

[0003] However, for the former solution of combining camera with millimeter wave radar, the image data signal collected by the camera needs to be processed. On the one hand, an algorithm needs to be built to identify objects, which makes the software more complicated and increases the data load of the vehicle ECU; on the other hand, the camera recognition effect is greatly affected in environments such as night, rainy days, and snowy days. For the latter solution of combining ultrasonic sensors, although the cost is lower, the recognition and detection efficiency of moving objects is low, and it is impossible to identify moving creatures behind and prevent the door from opening in the first time.

[0004] Secondly, both solutions control the detection hardware to work continuously. However, the traffic volume and other factors on different streets are different, and the probability of colliding with pedestrians or oncoming vehicles when opening the door is also different. If the detection hardware is controlled for continuous monitoring, the service life of the detection hardware may be reduced and power consumption may increase.

[0005] Therefore, a solution is urgently needed. Summary of the invention

[0006] One of the purposes of the present invention is to provide a passenger vehicle door opening monitoring system, which adopts a solution of millimeter-wave radar in combination with infrared human body sensor, which can not only solve the complex problem of big data image processing and reduce the data processing load of the system, but also reduce the impact of bad weather on mobile biological detection, and can also achieve efficient detection and prevent the opening of the door at a low cost to prevent collision accidents; introduce a door opening collision probability map, quickly determine the door opening collision probability, thereby determining the radar detection frequency corresponding to the door opening collision probability, and control the millimeter-wave radar module to detect the approach distance of the object approaching the rear of the vehicle at the radar detection frequency, which greatly improves the applicability of the system.

[0007] An embodiment of the present invention provides a passenger vehicle door opening monitoring system, comprising:

[0008] At least one millimeter wave radar module is arranged at the rear of the passenger vehicle;

[0009] At least one infrared human body sensor module is arranged at the rearview mirror of the passenger vehicle;

[0010] A control module is connected to the millimeter wave radar module and the infrared human body sensor module, and is used to include:

[0011] Controlling the millimeter wave radar module to detect the approach distance of an object approaching behind the rear of the vehicle;

[0012] When the approaching distance is less than or equal to a preset distance threshold, controlling the infrared human body sensor module to detect whether the approaching object is a living being;

[0013] When it is a living thing, the doors of the passenger vehicle are prohibited from opening;

[0014] The control module is also used to include:

[0015] Construct a door opening collision probability map;

[0016] When the passenger vehicle maintains the target state for more than a preset time, the vehicle position of the passenger vehicle is obtained; the target state includes: the vehicle speed is zero and the passengers have not gotten off the vehicle;

[0017] Determining a door opening collision probability corresponding to the vehicle position from a collision probability map;

[0018] Determine the radar detection frequency corresponding to the door opening collision probability from a preset radar detection frequency library;

[0019] The millimeter wave radar module is controlled to detect the approach distance of an object approaching behind the rear of the vehicle at the radar detection frequency.

[0020] Preferably, the control module constructs a door opening collision probability map, including:

[0021] Obtain multiple door opening collision events from the big data platform;

[0022] Obtaining multiple locations of each of the door opening collision events;

[0023] Retrieve the preset city traffic map;

[0024] Determine a plurality of first roadside parking locations corresponding to each of the incident locations from the city traffic map;

[0025] Assigning a preset first target probability to each of the first roadside parking positions;

[0026] determining a plurality of second roadside parking locations other than the first roadside parking location from the city traffic map;

[0027] Traversing the second roadside parking positions in sequence;

[0028] Determining from the city traffic map whether the first roadside parking position exists on the street where the traversed second roadside parking position is located;

[0029] When it exists, assigning the first target probability to the traversed second roadside parking position; when it does not exist, respectively acquiring the first position environment information of the traversed second roadside parking position and the second position environment information of the first roadside parking position from the city traffic map;

[0030] Calculating the maximum similarity between the first position environment information and the second position environment information;

[0031] Determining a second target probability corresponding to the maximum similarity from a preset probability library;

[0032] Assigning the second target probability to the traversed second roadside parking position;

[0033] The urban traffic map that assigns the first target probability to the first roadside parking position and the second target probability to the second roadside parking position is used as a door opening collision probability map.

[0034] Preferably, the control module obtains multiple door opening collision events from the big data platform, including:

[0035] Obtain the guarantee value of the big data platform for the trustworthy guarantee of each event element of the door opening collision event;

[0036] When the guarantee value of the trustworthy guarantee for each event element of the door opening collision event by the big data platform is greater than or equal to the guarantee threshold, the door opening collision event is obtained.

[0037] Preferably, the control module obtains a plurality of door opening collision events from a big data platform, and further includes:

[0038] When at least one of the guarantee values ​​of the big data platform for the trustworthy guarantee of each event element of the door opening collision event is less than the guarantee threshold, the scene and source of the door opening collision event are obtained;

[0039] Get the generation type of the door opening collision event generated by the generation source in the generation scene;

[0040] When the generation source generates a door opening collision event in the generation scene with the generation type being active, traverse each event element in the door opening collision event in sequence;

[0041] Each time the traversal is performed, the scene association relationship and the element extraction strategy corresponding to the element type of the traversed event element are determined from the scene association relationship library; the element extraction rules include: multiple groups of one-to-one corresponding extraction execution rules and execution priorities;

[0042] Execute the corresponding extraction execution rules in other scenes that have scene association relationships with the generated scene in order from the highest to the lowest execution priority;

[0043] When a replacement event element is extracted, the execution of the extraction execution rule is stopped, and the traversed event element in the door opening collision event is replaced with the replacement event element;

[0044] After traversing each event element in the door opening collision event, obtain the door opening collision event in which the event element is replaced with the replacement event element;

[0045] When the generation type of the door opening collision event generated by the generation source in the generation scene is passive generation, obtaining multimodal information related to the door opening collision event newly generated by the generation source in the generation scene;

[0046] Construct information description vectors of multimodal information;

[0047] Determine the quasi-acquisition value corresponding to the information description vector from the quasi-acquisition value library;

[0048] When the quasi-acquisition value is greater than or equal to the quasi-acquisition value threshold, a door opening collision event is acquired.

[0049] Preferably, the timing at which the control module controls the millimeter wave radar module to detect the approaching distance of an object approaching behind the rear of the vehicle includes:

[0050] After the passenger vehicle stops, the passengers are ready to get off; at this time, the vehicle speed sensor detects that the speed of the passenger vehicle is 0;

[0051] The passengers are ready to get on the vehicle; at this time, the vehicle lock status identifier detects that the vehicle lock is in the open state.

[0052] Preferably, the distance threshold includes: 5 meters.

[0053] Preferably, the control module is also used to control the door to open when the passenger opens the door twice in succession.

[0054] An embodiment of the present invention provides a method for monitoring the opening of a door of a passenger vehicle, comprising:

[0055] Controlling a millimeter wave radar module disposed at the rear of a passenger vehicle to detect an approaching distance of an object approaching behind the rear of the vehicle;

[0056] When the approaching distance is less than or equal to a preset distance threshold, an infrared human body sensor module arranged at a rearview mirror of the passenger vehicle detects whether the approaching object is a living being;

[0057] When it is a living thing, the doors of the passenger vehicle are prohibited from opening;

[0058] The method further comprises:

[0059] Construct a door opening collision probability map;

[0060] When the passenger vehicle maintains the target state for more than a preset time, the vehicle position of the passenger vehicle is obtained; the target state includes: the vehicle speed is zero and the passengers have not gotten off the vehicle;

[0061] Determining a door opening collision probability corresponding to the vehicle position from a collision probability map;

[0062] Determine the radar detection frequency corresponding to the door opening collision probability from a preset radar detection frequency library;

[0063] Turning on the millimeter wave radar module;

[0064] The millimeter wave radar module is controlled to detect the approach distance of an object approaching behind the rear of the vehicle at the radar detection frequency.

[0065] Preferably, the constructing of the door opening collision probability map comprises:

[0066] Obtain multiple door opening collision events from the big data platform;

[0067] Obtaining multiple locations of each of the door opening collision events;

[0068] Retrieve the preset city traffic map;

[0069] Determine a plurality of first roadside parking locations corresponding to each of the incident locations from the city traffic map;

[0070] Assigning a preset first target probability to each of the first roadside parking positions;

[0071] determining a plurality of second roadside parking locations other than the first roadside parking location from the city traffic map;

[0072] Traversing the second roadside parking positions in sequence;

[0073] Determining from the city traffic map whether the first roadside parking position exists on the street where the traversed second roadside parking position is located;

[0074] When it exists, assigning the first target probability to the traversed second roadside parking position; when it does not exist, respectively acquiring the first position environment information of the traversed second roadside parking position and the second position environment information of the first roadside parking position from the city traffic map;

[0075] Calculating the maximum similarity between the first position environment information and the second position environment information;

[0076] Determining a second target probability corresponding to the maximum similarity from a preset probability library;

[0077] Assigning the second target probability to the traversed second roadside parking position;

[0078] The urban traffic map that assigns the first target probability to the first roadside parking position and the second target probability to the second roadside parking position is used as a door opening collision probability map.

[0079] Preferably, the acquiring of multiple door opening collision events from the big data platform includes:

[0080] Obtain the guarantee value of the big data platform for the trustworthy guarantee of each event element of the door opening collision event;

[0081] When the guarantee value of the trustworthy guarantee for each event element of the door opening collision event by the big data platform is greater than or equal to the guarantee threshold, the door opening collision event is obtained.

[0082] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0083] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0085] Figure 1 Schematic diagram of the installation positions of the millimeter wave radar and infrared human body sensor in an embodiment of the present invention;

[0086] Figure 2 Schematic diagram of the electrical structure of a door opening monitoring system for a passenger vehicle in an embodiment of the present invention. DETAILED DESCRIPTION

[0087] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0088] The embodiment of the present invention provides a passenger vehicle door opening monitoring system, such as Figure 1 As shown, including:

[0089] At least one millimeter wave radar module is arranged at the rear of the passenger vehicle;

[0090] At least one infrared human body sensor module is arranged at the rearview mirror of the passenger vehicle;

[0091] A control module is connected to the millimeter wave radar module and the infrared human body sensor module, and is used to include:

[0092] Controlling the millimeter wave radar module to detect the approach distance of an object approaching behind the rear of the vehicle;

[0093] When the approaching distance is less than or equal to a preset distance threshold, controlling the infrared human body sensor module to detect whether the approaching object is a living being;

[0094] When it is a living thing, the doors of the passenger vehicle are prohibited from opening;

[0095] The control module is also used to include:

[0096] Construct a door-opening collision probability map; the door-opening collision probability map contains the door-opening collision probabilities corresponding to different roadside parking locations in the city; the door-opening collision probability is the probability of a collision between a parked car with a door opened and a pedestrian, oncoming vehicle, etc.;

[0097] When the passenger vehicle maintains the target state for more than a preset time, the vehicle position of the passenger vehicle is obtained; the target state includes: the vehicle speed is zero and the passengers have not gotten off the vehicle; the preset time may be, for example, 3 minutes; when the passenger vehicle maintains the target state for more than a preset time, it means that the vehicle has been parked, but there are passengers who have not gotten off the vehicle, for example, the driver / passenger is waiting for someone in the vehicle, etc.;

[0098] Determining a door opening collision probability corresponding to the vehicle position from a collision probability map;

[0099] Determine the radar detection frequency corresponding to the door opening collision probability from a preset radar detection frequency library; the radar detection frequency library contains radar detection frequencies corresponding to different door opening collision probabilities, and the greater the door opening collision probability, the more intensive monitoring is required, and the higher the radar detection frequency;

[0100] The millimeter wave radar module is controlled to detect the approach distance of an object approaching behind the rear of the vehicle at the radar detection frequency.

[0101] In the specific implementation, the millimeter wave radar is set at the rear of the car body, the first millimeter wave radar is set at the left rear edge of the car body; the second millimeter wave radar is set at the right rear edge of the car body; the millimeter wave radar is used to measure the condition of the target approaching from the rear and detect the distance of the moving object from the door. The infrared human body sensor is set under the rearview mirror of the car, the first infrared human body sensor is set under the left rearview mirror of the car body, and the second infrared human body sensor is set under the right rearview mirror of the car body. The infrared human body sensor is used to determine whether the moving object approaching the door is a living thing. The installation position is as follows: Figure 1 shown. Figure 1 Mark 1 is the installation position of the infrared human body sensor, and mark 2 is the installation position of the millimeter wave radar.

[0102] There are two states for door opening: State 1 is when the car stops and the speed is zero, the passenger opens the door to get out; State 2 is when there is no one in the car and the person outside needs to open the door to get in.

[0103] State 1 is when the vehicle speed sensor (built-in, installed on the wheel hub) detects that the vehicle speed is 0; State 2 is when the vehicle lock state identifier (built-in, installed on the vehicle lock) detects that the vehicle lock is in the open state. When at least one of the above states is activated, the door opening monitoring system function is activated. The millimeter wave radar is used to detect whether the moving object has reached the threshold set by the controller. The returned signal is sent to the processor for signal processing. The time difference between the ultrasonic radar transmission signal and the received signal and the speed of the sound wave are used to obtain the distance between the target object to be measured and the door. When the distance reaches the set threshold (the control system sets the safety threshold to 5 meters), the controller processes the signal received by the infrared human body sensor. Through the signal judgment, if the target object is a living thing, the control processor controls the vehicle lock control switch to lock the vehicle lock and sound an alarm. If you want to open the door, you need to manually unlock it twice in a row before you can open the door. Figure 2 Schematic diagram of the electrical structure of a door opening monitoring system for a passenger vehicle in an embodiment of the present invention.

[0104] The solution of using millimeter wave radar with infrared human body sensor can not only solve the complex problem of big data image processing and reduce the data processing load of the system, but also reduce the impact of bad weather on mobile biological detection. It can also achieve efficient detection and prevent the opening of car doors to prevent collision accidents at a low cost. The solution of using millimeter wave radar with infrared human body sensor detects the road conditions outside, and restricts or warns the car to open the door when it may collide with pedestrians or cyclists outside at the moment of opening the door. It has a certain effect on preventing and reducing traffic accidents caused by sudden door opening.

[0105] The millimeter-wave radar module is controlled to detect the approach distance of the object approaching behind the rear of the vehicle at the radar detection frequency; there is a special usage scenario for passenger vehicles: the driver parks the car on the side of the road and waits for someone / the driver parks the car on the side of the road and gets off, leaving the passengers in the car waiting for someone. At this time, someone in the car may get off at any time. Generally, the millimeter-wave radar module is controlled to perform continuous monitoring. However, this is not reasonable. The traffic volume and other factors on different streets are different, and the probability of colliding with pedestrians and oncoming vehicles when opening the door is different. If the millimeter-wave radar module is controlled to perform continuous monitoring, the service life of the millimeter-wave radar module may be reduced and the power consumption may be increased. The embodiment of the present invention can solve this problem, introduce a door opening collision probability map, quickly determine the door opening collision probability, thereby determining the radar detection frequency corresponding to the door opening collision probability, and control the millimeter-wave radar module to detect the approach distance of the object approaching behind the rear of the vehicle at the radar detection frequency, which greatly improves the applicability of the system.

[0106] In one embodiment, the control module constructs a door opening collision probability map, including:

[0107] Obtain multiple door opening collision events from the big data platform; the big data platform is a big data information platform; the door opening collision events are "door opening killing" events reported in the news and shared by drivers / the public;

[0108] Obtaining multiple locations of each of the door opening collision events;

[0109] Retrieve the preset city traffic map;

[0110] Determine a plurality of first roadside parking locations corresponding to each of the incident locations from the city traffic map;

[0111] Assigning a preset first target probability to each of the first roadside parking positions; the first target probability may be, for example, 70%; if a door opening collision event has occurred at the first roadside parking position in the past, the first target probability is directly assigned;

[0112] determining a plurality of second roadside parking locations other than the first roadside parking location from the city traffic map;

[0113] Traversing the second roadside parking positions in sequence;

[0114] Determining from the city traffic map whether the first roadside parking position exists on the street where the traversed second roadside parking position is located;

[0115] When it exists, assigning the first target probability to the traversed second roadside parking position; if there is a first roadside parking position on the same street as the second roadside parking position, directly assigning the first target probability to the second roadside parking position;

[0116] When it does not exist, the first position environment information of the second roadside parking position traversed and the second position environment information of the first roadside parking position are respectively obtained from the urban traffic map; the first position environment information includes: the daily average traffic flow, lane width, number of residents near the street, etc. of the street where the second roadside parking position is located; the second position environment information includes: the daily average traffic flow, lane width, number of residents near the street, etc. of the street where the first roadside parking position is located;

[0117] Calculating the maximum similarity between the first location environment information and the second location environment information; similarity calculation is within the scope of the prior art and will not be described in detail;

[0118] Determine the second target probability corresponding to the maximum similarity from a preset probability library; the probability library has second target probabilities corresponding to different similarities, and the greater the similarity, the greater the possibility that the second roadside parking position sends a door opening collision event at the first roadside parking position, and the greater the corresponding second target probability;

[0119] Assigning the second target probability to the traversed second roadside parking position;

[0120] The urban traffic map that assigns the first target probability to the first roadside parking position and the second target probability to the second roadside parking position is used as a door opening collision probability map.

[0121] When constructing the door-opening collision probability map, the embodiment of the present invention reasonably determines the probability of a collision between a vehicle with an open door and a pedestrian or an oncoming vehicle at each roadside parking location in the city, further improving the applicability of the system and making it more intelligent. In addition, the door-opening collision probability map is constructed online without the need to install relevant hardware equipment in the passenger vehicle, which improves convenience.

[0122] In one embodiment, the control module obtains multiple door opening collision events from a big data platform, including:

[0123] Obtain the guarantee value of the big data platform for the trustworthy guarantee of each event element of the door opening collision event; the event elements include: the time, location, type, etc. of the door opening collision event; when the big data platform provides the door opening collision event, it will trustworthy guarantee each event element and set the guarantee value. The larger the guarantee value, the greater the guarantee degree;

[0124] When the guarantee value of the big data platform for each event element of the door opening collision event is greater than or equal to the guarantee threshold, the door opening collision event is obtained. The guarantee threshold can be, for example, 75; when the guarantee value is greater than or equal to the guarantee threshold, it means that the big data platform has a high degree of guarantee for the event element, and the corresponding event element is credible; when the guarantee value of the big data platform for each event element of the door opening collision event is greater than or equal to the guarantee threshold, it means that the door opening collision event is completely credible and is obtained.

[0125] In this application, the event quality of the door opening collision event directly determines the construction quality of the door opening collision probability map. If the event quality of the door opening collision event is low, it will affect the determination of the first target probability and the second target probability, resulting in determination errors, which seriously reduces the accuracy of the door opening collision probability map in determining the collision probability; secondly, big data technology is becoming more and more popular, but the quality of data provided by big data technology is uneven, and a solution is urgently needed. The embodiment of the present invention obtains the guarantee value of the big data platform for each event element of the door opening collision event. When the guarantee value of the big data platform for each event element of the door opening collision event is greater than or equal to the guarantee threshold, the door opening collision event is obtained, which improves the quality and accuracy of the acquisition of the door opening collision event, and improves the applicability of this application using big data technology.

[0126] In one embodiment, the control module obtains multiple door opening collision events from a big data platform, and further includes:

[0127] When at least one of the guarantee values ​​of the big data platform for the trustworthy guarantee of each event element of the door-opening collision event is less than the guarantee threshold, the generation scene and generation source of the door-opening collision event are obtained; when at least one of the guarantee values ​​of the big data platform for the trustworthy guarantee of each event element of the door-opening collision event is less than the guarantee threshold, it indicates that the door-opening collision event may not be trustworthy; the generation scene may be: car friends exchange forum, news platform, etc., and correspondingly, the generation source may be: car owner user, news reporting party, etc.;

[0128] Obtain the type of door opening collision event generated by the source in the generation scenario; the generation type is divided into active generation and passive generation. Active generation is, for example, a car owner shares an event where he was about to open the car door in the city and encountered a car coming from behind, and passive generation is, for example, the vehicle manufacturer obtains an event monitored by the vehicle computer that the driver encountered a car coming from behind when opening the door, etc.;

[0129] When the generation source generates a door opening collision event in the generation scene with the generation type being active, traverse each event element in the door opening collision event in sequence;

[0130] Each time traversal is performed, the scene association relationship and the element extraction strategy corresponding to the element type of the traversed event element are determined from the scene association relationship library; the element extraction rules include: multiple sets of one-to-one corresponding extraction execution rules and execution priorities; the element types include: time, location, type, etc.; when the door opening collision event is not credible, its event elements need to be replaced; scene association relationships and element extraction strategies corresponding to different element types are introduced, and based on the element extraction strategy, replacement event elements of the replacement event elements can be extracted from other scenes that have scene association relationships with the generated scene, for example: if the element type is the location of the incident, then the scene association relationship is a scene that is generated after the generated scene and is the same as the source of information sharing (after the car owner user shares in the car friends forum where he was about to open the car door in the city and encountered a sudden car from behind, the location of the incident will be mentioned in the subsequent sharing post), the element extraction strategy includes multiple sets of one-to-one corresponding extraction execution rules and execution priorities, for example: the extraction execution rule is to extract the location of the incident published by the source of the generation, and the corresponding execution priority is 3, and for example: the extraction execution rule is to extract the location of the incident commented by the commentator, and the corresponding execution priority is 2;

[0131] Execute the corresponding extraction execution rules in other scenes that have scene association relationships with the generated scene in order from the highest to the lowest execution priority;

[0132] When a replacement event element is extracted, the execution of the extraction execution rule is stopped, and the traversed event element in the door opening collision event is replaced with the replacement event element; when a replacement event element is extracted, the execution of the extraction execution rule is stopped, for example: after extracting the location of the incident where the car owner shared in the car owner forum that he was about to open the car door and encountered a sudden car from behind, the execution of the extraction execution rule is stopped;

[0133] After traversing each event element in the door opening collision event, obtain the door opening collision event in which the event element is replaced with the replacement event element;

[0134] When the generation type of the door opening collision event generated by the generation source in the generation scene is passive generation, obtaining multimodal information related to the door opening collision event newly generated by the generation source in the generation scene; the multimodal information includes: the number of door opening collision events provided by the generation source in history, authentication information of the generation source, etc.;

[0135] Construct information description vectors of multimodal information;

[0136] Determine the quasi-acquisition value corresponding to the information description vector from the quasi-acquisition value library; the quasi-acquisition value library contains quasi-acquisition values ​​corresponding to different information description vectors. The higher the credibility of the source represented by the multimodal information, the greater the quasi-acquisition value corresponding to the constructed information description vector. For example, the more door opening collision events the source has provided in history and the more comprehensive the authentication information of the source, the higher the credibility of the source and the greater the quasi-acquisition value corresponding to the information description vector.

[0137] When the quasi-acquisition value is greater than or equal to the quasi-acquisition value threshold, the door opening collision event is acquired. The quasi-acquisition value threshold may be, for example, 80; when the quasi-acquisition value is greater than or equal to the quasi-acquisition value threshold, the door opening collision event is acquired.

[0138] When the embodiment of the present invention obtains multiple door opening collision events from a big data platform, it first determines the guarantee status of the big data platform for the door opening collision events, and determines whether it is necessary to perform event element replacement / generated source trusted verification on the door opening collision events based on the guarantee status, thereby reducing the resource usage of the system and improving the event acquisition efficiency. Secondly, when it is necessary to perform event element replacement / generated source trusted verification on the door opening collision events, event element replacement / generated source trusted verification is performed separately according to the different generation types of the door opening collision events generated by the generation sources in the generation scenarios, thereby improving the accuracy of event acquisition; secondly, the scene association relationship and element extraction strategy corresponding to the element type are introduced to improve the efficiency of determining the replacement event elements.

[0139] In one embodiment, the timing at which the control module controls the millimeter wave radar module to detect the approaching distance of an object behind the rear of the vehicle includes:

[0140] After the passenger vehicle stops, the passengers are ready to get off; at this time, the vehicle speed sensor detects that the speed of the passenger vehicle is 0;

[0141] The passengers are ready to get on the vehicle; at this time, the vehicle lock status identifier detects that the vehicle lock is in the open state.

[0142] This embodiment corresponds to the two states of the vehicle door being opened.

[0143] In one embodiment, the control module is also used to control the door to open when the passenger opens the door twice in succession.

[0144] When a passenger opens the door twice in succession, the door can be opened for him.

[0145] An embodiment of the present invention provides a method for monitoring the opening of a door of a passenger vehicle, comprising:

[0146] Controlling a millimeter wave radar module disposed at the rear of a passenger vehicle to detect an approaching distance of an object approaching behind the rear of the vehicle;

[0147] When the approaching distance is less than or equal to a preset distance threshold, an infrared human body sensor module arranged at a rearview mirror of the passenger vehicle detects whether the approaching object is a living being;

[0148] When it is a living thing, the doors of the passenger vehicle are prohibited from opening;

[0149] The method further comprises:

[0150] Construct a door opening collision probability map;

[0151] When the passenger vehicle maintains the target state for more than a preset time, the vehicle position of the passenger vehicle is obtained; the target state includes: the vehicle speed is zero and the passengers have not gotten off the vehicle;

[0152] Determining a door opening collision probability corresponding to the vehicle position from a collision probability map;

[0153] Determine the radar detection frequency corresponding to the door opening collision probability from a preset radar detection frequency library;

[0154] Turning on the millimeter wave radar module;

[0155] The millimeter wave radar module is controlled to detect the approach distance of an object approaching behind the rear of the vehicle at the radar detection frequency.

[0156] The step of constructing a door opening collision probability map includes:

[0157] Obtain multiple door opening collision events from the big data platform;

[0158] Obtaining multiple locations of each of the door opening collision events;

[0159] Retrieve the preset city traffic map;

[0160] Determine a plurality of first roadside parking locations corresponding to each of the incident locations from the city traffic map;

[0161] Assigning a preset first target probability to each of the first roadside parking positions;

[0162] determining a plurality of second roadside parking locations other than the first roadside parking location from the city traffic map;

[0163] Traversing the second roadside parking positions in sequence;

[0164] Determining from the city traffic map whether the first roadside parking position exists on the street where the traversed second roadside parking position is located;

[0165] When it exists, assigning the first target probability to the traversed second roadside parking position; when it does not exist, respectively acquiring the first position environment information of the traversed second roadside parking position and the second position environment information of the first roadside parking position from the city traffic map;

[0166] Calculating the maximum similarity between the first position environment information and the second position environment information;

[0167] Determining a second target probability corresponding to the maximum similarity from a preset probability library;

[0168] Assigning the second target probability to the traversed second roadside parking position;

[0169] The urban traffic map that assigns the first target probability to the first roadside parking position and the second target probability to the second roadside parking position is used as a door opening collision probability map.

[0170] The obtaining of multiple door opening collision events from the big data platform includes:

[0171] Obtain the guarantee value of the big data platform for the trustworthy guarantee of each event element of the door opening collision event;

[0172] When the guarantee value of the trustworthy guarantee for each event element of the door opening collision event by the big data platform is greater than or equal to the guarantee threshold, the door opening collision event is obtained.

[0173] Multiple door opening collision events are obtained from the big data platform, including:

[0174] When at least one of the guarantee values ​​of the big data platform for the trustworthy guarantee of each event element of the door opening collision event is less than the guarantee threshold, the scene and source of the door opening collision event are obtained;

[0175] Get the generation type of the door opening collision event generated by the generation source in the generation scene;

[0176] When the generation source generates a door opening collision event in the generation scene with the generation type being active, traverse each event element in the door opening collision event in sequence;

[0177] Each time the traversal is performed, the scene association relationship and the element extraction strategy corresponding to the element type of the traversed event element are determined from the scene association relationship library; the element extraction rules include: multiple groups of one-to-one corresponding extraction execution rules and execution priorities;

[0178] Execute the corresponding extraction execution rules in other scenes that have scene association relationships with the generated scene in order from the highest to the lowest execution priority;

[0179] When a replacement event element is extracted, the execution of the extraction execution rule is stopped, and the traversed event element in the door opening collision event is replaced with the replacement event element;

[0180] After traversing each event element in the door opening collision event, obtain the door opening collision event in which the event element is replaced with the replacement event element;

[0181] When the generation type of the door opening collision event generated by the generation source in the generation scene is passive generation, obtaining multimodal information related to the door opening collision event newly generated by the generation source in the generation scene;

[0182] Construct information description vectors of multimodal information;

[0183] Determine the quasi-acquisition value corresponding to the information description vector from the quasi-acquisition value library;

[0184] When the quasi-acquisition value is greater than or equal to the quasi-acquisition value threshold, a door opening collision event is acquired.

[0185] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A passenger vehicle door opening monitoring system, characterized in that: include: At least one millimeter wave radar module is arranged at the rear of the passenger vehicle; At least one infrared human body sensor module is arranged at the rearview mirror of the passenger vehicle; A control module is connected to the millimeter wave radar module and the infrared human body sensor module, and is used to include: Controlling the millimeter wave radar module to detect the approach distance of an object approaching behind the rear of the vehicle; When the approaching distance is less than or equal to a preset distance threshold, controlling the infrared human body sensor module to detect whether the approaching object is a living being; When it is a living thing, the doors of the passenger vehicle are prohibited from opening; The control module is also used to include: Construct a door opening collision probability map; When the passenger vehicle maintains the target state for more than a preset time, the vehicle position of the passenger vehicle is obtained; the target state includes: the vehicle speed is zero and the passengers have not gotten off the vehicle; Determining a door opening collision probability corresponding to the vehicle position from a collision probability map; Determine the radar detection frequency corresponding to the door opening collision probability from a preset radar detection frequency library; Controlling the millimeter wave radar module to detect the approach distance of an object approaching behind the rear of the vehicle at the radar detection frequency; The control module constructs a door opening collision probability map, including: Obtain multiple door-opening collision events from the big data platform; the door-opening collision events at least include: door-opening collision events reported in the news and shared by drivers / the public; Obtaining multiple locations of each of the door opening collision events; Retrieve the preset city traffic map; Determine a plurality of first roadside parking locations corresponding to each of the incident locations from the city traffic map; Assigning a preset first target probability to each of the first roadside parking positions; determining a plurality of second roadside parking locations other than the first roadside parking location from the city traffic map; Traversing the second roadside parking positions in sequence; Determining from the city traffic map whether the first roadside parking position exists on the street where the traversed second roadside parking position is located; When it exists, assign the first target probability to the traversed second roadside parking position; when it does not exist, respectively obtain the first position environment information of the traversed second roadside parking position and the second position environment information of the first roadside parking position from the urban traffic map; the first position environment information includes at least: the daily average traffic volume, lane width, and number of residents near the street where the second roadside parking position is located; the second position environment information includes at least: the daily average traffic volume, lane width, and number of residents near the street where the first roadside parking position is located; Calculating the maximum similarity between the first position environment information and the second position environment information; Determine the second target probability corresponding to the maximum similarity from a preset probability library; the probability library has second target probabilities corresponding to different similarities, and the greater the similarity, the greater the possibility that the second roadside parking position sends a door opening collision event at the first roadside parking position, and the greater the corresponding second target probability; Assigning the second target probability to the traversed second roadside parking position; Using the urban traffic map that assigns the first target probability to the first roadside parking position and the second target probability to the second roadside parking position as a door opening collision probability map; The control module obtains multiple door opening collision events from the big data platform, including: Obtain the guarantee value of the big data platform for the reliable guarantee of each event element of the door opening collision event; the event elements at least include: the time, location and type of the door opening collision event; When the guarantee value of the trustworthy guarantee of each event element of the door opening collision event by the big data platform is greater than or equal to the guarantee threshold, the door opening collision event is obtained; The control module obtains multiple door opening collision events from the big data platform, and also includes: When at least one of the guarantee values ​​of the big data platform for the reliable guarantee of each event element of the door opening collision event is less than the guarantee threshold, the scene and source of the door opening collision event are obtained; the scene at least includes: a car owner exchange forum, a news platform; the source at least includes: a car owner user, a news reporting party; Get the generation type of the door opening collision event generated by the generation source in the generation scene; When the generation source generates the door opening collision event in the generation scene as active generation, each event element in the door opening collision event is traversed in sequence; active generation at least includes: the car owner user shares the event where he / she encounters a sudden car coming from behind when he / she is about to open the car door in the city; Each time the traversal is performed, the scene association relationship and the element extraction strategy corresponding to the element type of the traversed event element are determined from the scene association relationship library; the element extraction rules include: multiple groups of one-to-one corresponding extraction execution rules and execution priorities; Execute corresponding extraction execution rules in other scenes that have scene association relationships with the generated scene in order from high to low execution priority; When a replacement event element is extracted, the execution of the extraction execution rule is stopped, and the traversed event element in the door opening collision event is replaced with the replacement event element; After traversing each event element in the door opening collision event, obtain the door opening collision event in which the event element is replaced with the replacement event element; When the generation type of the door opening collision event generated by the generation source in the generation scene is passive generation, multimodal information related to the door opening collision event newly generated by the generation source in the generation scene is obtained; the passive generation at least includes: the vehicle manufacturer obtains the event of the driver encountering a rear vehicle when opening the door monitored by the vehicle computer; the multimodal information at least includes: the number of door opening collision events provided by the generation source in history, and the authentication information of the generation source; Construct information description vectors of multimodal information; Determine a quasi-acquisition value corresponding to the information description vector from a quasi-acquisition value library; the quasi-acquisition value library contains quasi-acquisition values ​​corresponding to different information description vectors, and the higher the credibility of the source of the multimodal information representative, the greater the quasi-acquisition value corresponding to the constructed information description vector; When the quasi-acquisition value is greater than or equal to the quasi-acquisition value threshold, a door opening collision event is acquired.

2. A passenger vehicle door opening monitoring system as claimed in claim 1, characterized in that: The timing for the control module to control the millimeter wave radar module to detect the approaching distance of an object behind the rear of the vehicle includes: After the passenger vehicle stops, the passengers are ready to get off; at this time, the vehicle speed sensor detects that the speed of the passenger vehicle is 0; The passengers are ready to get on the vehicle; at this time, the vehicle lock status identifier detects that the vehicle lock is in the open state.

3. A passenger vehicle door opening monitoring system as claimed in claim 1, characterized in that: The distance threshold includes: 5 meters.

4. A passenger vehicle door opening monitoring system as claimed in claim 1, characterized in that: The control module is also used to control the door to open when the passenger opens the door twice in succession.

5. A method for monitoring the opening of a passenger vehicle door, characterized in that: include: Controlling a millimeter wave radar module disposed at the rear of a passenger vehicle to detect an approaching distance of an object approaching behind the rear of the vehicle; When the approaching distance is less than or equal to a preset distance threshold, an infrared human body sensor module arranged at a rearview mirror of the passenger vehicle detects whether the approaching object is a living being; When it is a living thing, the doors of the passenger vehicle are prohibited from opening; The method further comprises: Construct a door opening collision probability map; When the passenger vehicle maintains the target state for more than a preset time, the vehicle position of the passenger vehicle is obtained; the target state includes: the vehicle speed is zero and the passengers have not gotten off the vehicle; Determining a door opening collision probability corresponding to the vehicle position from a collision probability map; Determine the radar detection frequency corresponding to the door opening collision probability from a preset radar detection frequency library; Turning on the millimeter wave radar module; Controlling the millimeter wave radar module to detect the approach distance of an object approaching behind the rear of the vehicle at the radar detection frequency; The step of constructing a door opening collision probability map includes: Obtain multiple door-opening collision events from the big data platform; the door-opening collision events at least include: door-opening collision events reported in the news and shared by drivers / the public; Obtaining multiple locations of each of the door opening collision events; Retrieve the preset city traffic map; Determine a plurality of first roadside parking locations corresponding to each of the incident locations from the city traffic map; Assigning a preset first target probability to each of the first roadside parking positions; determining a plurality of second roadside parking locations other than the first roadside parking location from the city traffic map; Traversing the second roadside parking positions in sequence; Determining from the city traffic map whether the first roadside parking position exists on the street where the traversed second roadside parking position is located; When it exists, assign the first target probability to the traversed second roadside parking position; when it does not exist, respectively obtain the first position environment information of the traversed second roadside parking position and the second position environment information of the first roadside parking position from the urban traffic map; the first position environment information includes at least: the daily average traffic volume, lane width, and number of residents near the street where the second roadside parking position is located; the second position environment information includes at least: the daily average traffic volume, lane width, and number of residents near the street where the first roadside parking position is located; Calculating the maximum similarity between the first position environment information and the second position environment information; Determine the second target probability corresponding to the maximum similarity from a preset probability library; the probability library has second target probabilities corresponding to different similarities, and the greater the similarity, the greater the possibility that the second roadside parking position sends a door opening collision event at the first roadside parking position, and the greater the corresponding second target probability; Assigning the second target probability to the traversed second roadside parking position; Using the urban traffic map that assigns the first target probability to the first roadside parking position and the second target probability to the second roadside parking position as a door opening collision probability map; The obtaining of multiple door opening collision events from the big data platform includes: Obtain the guarantee value of the big data platform for the trustworthy guarantee of each event element of the door opening collision event; the event elements at least include: the time, location and type of the door opening collision event; When the guarantee value of the trustworthy guarantee of each event element of the door opening collision event by the big data platform is greater than or equal to the guarantee threshold, the door opening collision event is obtained; The acquiring of multiple door opening collision events from the big data platform also includes: When at least one of the guarantee values ​​of the big data platform for the reliable guarantee of each event element of the door opening collision event is less than the guarantee threshold, the scene and source of the door opening collision event are obtained; the scene at least includes: a car owner exchange forum, a news platform; the source at least includes: a car owner user, a news reporting party; Get the generation type of the door opening collision event generated by the generation source in the generation scene; When the generation source generates the door opening collision event in the generation scene as active generation, each event element in the door opening collision event is traversed in sequence; active generation at least includes: the car owner user shares the event where he / she encounters a sudden car coming from behind when he / she is about to open the car door in the city; Each time the traversal is performed, the scene association relationship and the element extraction strategy corresponding to the element type of the traversed event element are determined from the scene association relationship library; the element extraction rules include: multiple groups of one-to-one corresponding extraction execution rules and execution priorities; Execute the corresponding extraction execution rules in other scenes that have scene association relationships with the generated scene in order from the highest to the lowest execution priority; When a replacement event element is extracted, the execution of the extraction execution rule is stopped, and the traversed event element in the door opening collision event is replaced with the replacement event element; After traversing each event element in the door opening collision event, obtain the door opening collision event in which the event element is replaced with the replacement event element; When the generation type of the door opening collision event generated by the generation source in the generation scene is passive generation, multimodal information related to the door opening collision event newly generated by the generation source in the generation scene is obtained; the passive generation at least includes: the vehicle manufacturer obtains the event of the driver encountering a rear vehicle when opening the door monitored by the vehicle computer; the multimodal information at least includes: the number of door opening collision events provided by the generation source in history, and the authentication information of the generation source; Construct information description vectors of multimodal information; Determine a quasi-acquisition value corresponding to the information description vector from a quasi-acquisition value library; the quasi-acquisition value library contains quasi-acquisition values ​​corresponding to different information description vectors, and the higher the credibility of the source of the multimodal information representative, the greater the quasi-acquisition value corresponding to the constructed information description vector; When the quasi-acquisition value is greater than or equal to the quasi-acquisition value threshold, a door opening collision event is acquired.

Images