Accident severity estimation system for vehicle

Abstract

An accident severity estimation system for estimating the severity of an accident in a vehicle includes one or more microphones that capture a plurality of audio-based inputs indicative of linguistic and non-linguistic sounds emitted by one or more occupants of the vehicle. The accident severity estimation system also includes a vision system that captures a plurality of vision-based inputs representing image data indicative of the occupant; a motion-based input system that collects a plurality of motion-based inputs indicative of motion of the vehicle during the accident; a thermal event system that collects a plurality of thermal inputs indicative of thermal events within the vehicle; a propulsion system providing a state-based input of the propulsion system of the vehicle; and one or more controllers.

Description

Technical Field

[0001] The present disclosure relates to an accident severity estimation system for estimating the severity of an accident occurring to a vehicle, and the accident severity estimation system further selects one or more remedial actions to be performed based on the severity of the accident. Background Art

[0002] After a vehicle is involved in an incident (e.g., a collision), various remedial actions may be taken to assist the occupants of the vehicle. For example, if the vehicle is involved in a relatively minor collision that does not affect the occupants, non-emergency personnel may be contacted via the subscription-based vehicle communication system. Conversely, if the collision is more severe, emergency personnel may be contacted in addition to non-emergency personnel. In some cases, after a vehicle is involved in an accident, the doors may be unlocked or the windows may be lowered.

[0003] It should be understood that, depending on the specific details and severity of the collision, some types of remedial actions may be more appropriate or helpful than some other types of remedial actions performed after a vehicle accident. For example, for a relatively minor collision, contacting emergency personnel may not be particularly helpful or necessary. As another example, in certain circumstances, such as when the vehicle is in a body of water, lowering the vehicle's windows may not be particularly helpful. As another example, unlocking the vehicle's doors may be particularly useful, such as when smoke is present in the interior cabin.

[0004] Therefore, while current accident response systems achieve their intended purpose, there remains a need in the art for an improved method of determining remedial actions that may be performed following a vehicle collision. Summary of the invention

[0005] According to several aspects, an accident severity estimation system for estimating the severity of a vehicle accident is disclosed, including one or more microphones that capture multiple audio-based inputs indicating verbal and non-verbal sound inputs emitted by one or more occupants of the vehicle; a vision system that captures multiple vision-based inputs representing image data indicating the occupants; a motion-based input system that collects multiple motion-based inputs indicating the movement of the vehicle during the accident; a thermal event system that collects multiple thermal inputs indicating thermal events within the vehicle; a propulsion system that provides state-based inputs of a propulsion system of the vehicle; and one or more controllers that electronically communicate with the one or more microphones, the vision system, the motion-based input system, the thermal event system, and the propulsion system. The one or more controllers execute instructions to combine the multiple audio-based inputs based on at least a binary logic system to determine an audio-based indicator, combine the multiple vision-based inputs based on at least a binary logic system to determine a vision-based indicator, and combine the multiple thermal inputs based on at least a binary logic system to determine a thermal-based indicator. The one or more controllers combine the multiple motion-based inputs from the motion-based input system based on a weighted sum model to determine the motion-based indicator. The one or more controllers determine a propulsion-based indicator based on the state-based input. The one or more controllers estimate a severity of the accident based on the audio-based indicator, the visual-based indicator, the thermal-based indicator, the motion-based indicator, and the propulsion-based indicator.

[0006] In another aspect, one or more controllers execute instructions to estimate a severity of an accident involving a vehicle based on comparison values ​​of audio-based indicators, visual-based indicators, motion-based indicators, thermal-based indicators, and propulsion-based indicators stored in a lookup table.

[0007] In yet another aspect, one or more controllers execute instructions to estimate a severity of the accident based on a fuzzy logic severity estimation technique.

[0008] On the one hand, the fuzzy logic severity estimation technology includes assigning a corresponding fuzzy value to each indicator based on a membership function, estimating the corresponding fuzzy value of each indicator based on one or more fuzzy logic rules to determine a fuzzified output value, wherein the fuzzified output value indicates the severity of the accident and a true value corresponding to the severity, and converting the fuzzified output value into the severity of the accident based on the true value.

[0009] In another aspect, one or more controllers execute instructions to select one or more remedial actions based on a severity of the incident, a thermal event indicator, a water event indicator, and a root cause probability indicator.

[0010] In yet another aspect, the one or more remedial actions include one or more of: contacting emergency personnel, unlocking the vehicle's doors, lowering the vehicle's windows, sending an SOS signal to one or more vehicles within a predetermined radius of the vehicle, and contacting non-emergency personnel.

[0011] In one aspect, a propulsion system includes a battery pack that provides power to one or more electric motors.

[0012] In another aspect, the state-based input is a battery-based input that is indicative of the health of the battery pack.

[0013] In yet another aspect, a propulsion system includes an internal combustion engine.

[0014] In one aspect, the state-based input includes an engine- and position-based state input indicative of a state of the internal combustion engine and a position input indicative of a position of the vehicle.

[0015] In another aspect, audio-based input includes non-speech-based input and speech-based input.

[0016] In yet another aspect, the plurality of vision-based inputs includes in-vehicle inputs indicative of visual states of occupants within a cabin of the vehicle and out-vehicle inputs indicative of visual states of occupants outside the vehicle and outside the vehicle.

[0017] In one aspect, the heat input includes cabin heat input and engine compartment heat input.

[0018] In another aspect, the plurality of motion-based inputs includes one or more of: acceleration and deceleration inputs, crash angle inputs, rollover inputs, and speed change inputs.

[0019] In one aspect, the weighted sum model is expressed as:

[0020]

[0021] Among them, ω M1 represents the first weighting factor corresponding to the acceleration and deceleration input M1, ω M2 represents the second weighting factor corresponding to the collision angle input M2, ω M3 represents the third weight factor corresponding to the rollover input M3, ω M4 represents a fourth weighting factor corresponding to the speed change input M4.

[0022] On the other hand, a two-valued logic system is one of a binary logic system and a ternary logic system.

[0023] In yet another aspect, an accident severity estimation system for estimating the severity of a vehicle accident is disclosed, comprising: one or more microphones that capture a plurality of audio-based inputs indicating verbal and non-verbal sounds emitted by one or more occupants of the vehicle; a vision system that captures a plurality of vision-based inputs representing image data indicating the occupants; a motion-based input system that collects a plurality of motion-based inputs indicating the motion of the vehicle during the accident; a thermal event system that collects a plurality of thermal inputs indicating thermal events within the vehicle; a propulsion system that provides state-based inputs of a propulsion system of the vehicle; and one or more controllers that electronically communicate with the one or more microphones, the vision system, the motion-based input system, the thermal event system, and the propulsion system. The one or more controllers execute instructions to combine the plurality of audio-based inputs based on at least a binary logic system to determine an audio-based indicator, combine the plurality of vision-based inputs based on at least a binary logic system to determine a vision-based indicator, and combine the plurality of thermal inputs based on at least a binary logic system to determine a thermal-based indicator. The one or more controllers combine the plurality of motion-based inputs from the motion-based input system based on a weighted sum model to determine a motion-based indicator. One or more controllers determine a propulsion-based indicator based on the state-based input, and estimate a severity of the accident based on the audio-based indicator, the visual-based indicator, the thermal-based indicator, the motion-based indicator, and the propulsion-based indicator, and estimate the severity of the accident based on a fuzzy logic severity estimation technique.

[0024] On the other hand, the fuzzy logic severity estimation technology includes: assigning a corresponding fuzzy value to each indicator based on a membership function, estimating the corresponding fuzzy value of each indicator based on one or more fuzzy logic rules to determine a fuzzified output value, wherein the fuzzified output value indicates the severity of the accident and a true value corresponding to the severity; and converting the fuzzified output value into the severity of the accident based on the true value.

[0025] In yet another aspect, one or more controllers execute instructions to select one or more remedial actions based on a severity of the incident, a thermal event indicator, a water event indicator, and a root cause probability indicator.

[0026] In one aspect, an accident severity estimation system for estimating the severity of a vehicle accident is disclosed, comprising: one or more microphones that capture a plurality of audio-based inputs indicating verbal and non-verbal sounds emitted by one or more occupants of the vehicle; a vision system that captures a plurality of vision-based inputs representing image data indicating the occupants; a motion-based input system that collects a plurality of motion-based inputs indicating the motion of the vehicle during the accident; a thermal event system that collects a plurality of thermal inputs indicating thermal events within the vehicle; a propulsion system that provides state-based inputs of a propulsion system of the vehicle; and one or more controllers that electronically communicate with the one or more microphones, the vision system, the motion-based input system, the thermal event system, and the propulsion system. The one or more controllers execute instructions to combine the plurality of audio-based inputs based on at least a binary logic system to determine an audio-based indicator, combine the plurality of vision-based inputs based on at least a binary logic system to determine a vision-based indicator, and combine the plurality of thermal inputs based on at least a binary logic system to determine a thermal-based indicator. The one or more controllers combine the plurality of motion-based inputs from the motion-based input system based on a weighted sum model to determine the motion-based indicator. One or more controllers determine a propulsion-based indicator based on the state-based input, and estimate the severity of the accident based on the audio-based indicator, the vision-based indicator, the thermal-based indicator, the motion-based indicator, and the propulsion-based indicator, and estimate the severity of the accident based on a fuzzy logic severity estimation technique. The fuzzy logic severity estimation technique includes assigning a corresponding fuzzy value to each indicator based on a membership function, estimating the fuzzy value corresponding to each indicator based on one or more fuzzy logic rules to determine a fuzzified output value, wherein the fuzzified output value indicates the severity of the accident and a truth value corresponding to the severity, and converting the fuzzified output value to the severity of the accident based on the truth value.

[0027] Further areas of applicability will become apparent from the description provided herein.It should be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The drawings described herein are for illustration purposes only and are not intended to limit the scope of the present disclosure in any way.

[0029] Figure 1 A schematic diagram of a vehicle including a disclosed severity estimation system including one or more controllers in electronic communication with a microphone, a vision system, a motion-based indicator system, a thermal event system, a battery management system, and an engine status system is shown according to an exemplary embodiment;

[0030] Figure 2is a schematic diagram of one or more controllers that receive inputs from a microphone, a vision system, a motion-based indicator system, a thermal event system, a battery management system, and an engine status system according to an exemplary embodiment;

[0031] Figure 3 shows a ternary logical OR table for combining two audio-based inputs according to an exemplary embodiment;

[0032] Figure 4 shows a table indicating ternary logic for determining a battery-based indicator according to an exemplary embodiment;

[0033] Figure 5 According to an exemplary embodiment, Figure 1 A fuzzy logic severity estimation technique implemented by one or more controllers as shown; and

[0034] Figure 6 is a schematic diagram illustrating a cause-action diagram for selecting one or more remedial actions by one or more controllers according to an exemplary embodiment. DETAILED DESCRIPTION

[0035] The following description is merely exemplary in nature and is not intended to limit the present disclosure, application, or uses.

[0036] refer to Figure 1 , a vehicle 10 including the disclosed accident severity estimation system 12 is shown. As explained below, the accident severity estimation system 12 employs a multimodal approach to estimate the severity of an accident occurring with the vehicle 10. It should be understood that the vehicle 10 may be any type of vehicle, such as, but not limited to, a car, a truck, a sport utility vehicle, a van, or an RV. In the non-limiting embodiment shown in the figure, as Figure 1 As shown, the vehicle 10 includes one or more controllers 20 in electronic communication with one or more microphones 22 , a vision system 24 , a motion-based indicator system 26 , a thermal event system 28 , a battery management system 30 , and an engine status system 32 , which are located within an interior cabin 18 of the vehicle 10 .

[0037] As explained below, the one or more controllers 20 estimate the severity of an accident that occurs with the vehicle 10 and select one or more remedial actions based on the severity of the accident. In an embodiment, an accident may refer to an unintentional event (caused by the driver or the autonomous driving system of the vehicle 10) that occurs with the vehicle 10. By way of example only, an accident may refer to a rollover accident, a breakdown of the vehicle 10, a collision between the vehicle 10 and another vehicle in the surrounding environment, a collision between the vehicle 10 and an object in the surrounding environment (e.g., a traffic sign, a pole, or a tree).

[0038] One or more microphones 22 are located in the interior cabin 18 of the vehicle 10 and the exterior 16 of the vehicle 10 to capture audio signals. The visual system 24 includes one or more cameras 34 disposed in the interior cabin 18 of the vehicle 10 to capture image data representing the occupant 14, and one or more cameras 36 disposed along the exterior side 44 of the vehicle 10 to capture image data of the occupant 14 who places his or her limbs outside the interior cabin 18. For example, the occupant 14 may place his or her hands or arms outside the window. The motion-based indicator system 26 includes one or more diagnostic controllers that record motion-based data of the vehicle 10 during the accident. The motion-based data includes, for example, acceleration data and pressure measurements. In an embodiment, the motion-based indicator system 26 includes one or more sensing and diagnostic modules (SDMs) and / or one or more electronic data recorders (EDRs) that record motion-based data of the vehicle 10. The thermal event system 28 includes a plurality of sensors 38 that monitor the exterior 16 and interior cabin 18 of the vehicle 10 to determine temperature, and monitor the engine compartment to determine the engine compartment temperature.

[0039] The controller 20 receives state-based input from the propulsion system of the vehicle 10. It should be understood that in one embodiment, the propulsion system is a battery pack 40 that provides power to one or more electric motors 42, and the state-based input indicates the health of the battery pack 40. In one embodiment, the battery management system 30 is a wireless battery management system (wBMS). Alternatively, in another embodiment, the propulsion system is an internal combustion engine 48, and the state-based input indicates the state of the internal combustion engine 48 (e.g., on or off) and the vehicle state position to one or more controllers 20. The vehicle state position indicates the location of the vehicle 10, wherein the location may indicate whether the vehicle 10 is parked at the side of the road, on the road, or parked.

[0040] Figure 2 is a schematic diagram showing one or more controllers 20 that receive a plurality of audio-based inputs A, B from a microphone 22, a plurality of vision-based inputs C, D from a vision system 24, a plurality of motion-based inputs M1-M4 from a motion-based indicator system 26, a plurality of thermal inputs E, F from a thermal event system 28, a battery-based input K from a battery management system 30, and a plurality of engine- and position-based indicators G, H from an engine status system 32. Reference Figure 1 and Figure 2, a microphone 22 is located in the interior cabin 18 of the vehicle 10 and captures audio-based inputs A, B, the inputs A, B being indicative of verbal and non-verbal sounds made by the occupants 14 of the vehicle 10. The vision system 24 includes one or more cameras 34, 36 for capturing a plurality of vision-based inputs C, D, the inputs C, D representing image data indicative of the occupants 14. The motion-based indicator system 26 collects a plurality of motion-based inputs M1-M4, the inputs M1-M4 being indicative of the motion of the vehicle 10 during the accident. The thermal event system 28 collects thermal inputs E, F, the thermal inputs E, F being indicative of thermal events within the interior cabin 18 and the engine compartment of the vehicle 10. A thermal event refers to the presence of smoke or flames.

[0041] In one embodiment, the vehicle 10 is an electric vehicle, and the battery management system 30 provides a battery-based input K to the one or more controllers 20, the input K indicating the health of the battery pack 40 of the vehicle 10. Alternatively, in another embodiment, the vehicle 10 is propelled by an internal combustion engine 48, and the engine status system 32 provides a plurality of engine- and position-based inputs G, H to the one or more controllers 20, the inputs G, H indicating the status of the internal combustion engine 48 and the vehicle status position.

[0042] As explained below, the one or more controllers 20 combine the multiple audio-based inputs A, B from the microphones 22 based on at least a binary logic system to determine an audio-based indicator I audio In one embodiment, the binary logic system refers to a binary logic system or a ternary logic system. The one or more controllers 20 combine the vision-based inputs C and D from the vision system 24 based on at least the binary logic system to determine the vision-based indicator I vision The one or more controllers 20 combine the plurality of motion-based inputs M1-M4 from the motion-based indicator system 26 based on a weighted sum model to determine a motion-based indicator I mo tion The one or more controllers 20 combine the thermal inputs E, F from the thermal event system 28 based on at least a binary logic system to determine a thermal-based indicator I thermal In one embodiment, the propulsion system includes a battery pack 40 and an electric motor 42, and the one or more controllers 20 determine a propulsion-based indicator based on a battery-based input K from the battery pack 40 of the vehicle 10, which is a battery-based indicator I battery In another embodiment, the propulsion system includes an internal combustion engine 48, and the one or more controllers 20 combine multiple engine- and position-based inputs G, H from the internal combustion engine 48 based on at least a binary logic system to determine a propulsion-based indicator that is an engine- and position-based indicator I location .

[0043] As also explained below, the one or more controllers 20 may determine the audio-based indicator I audio , Vision-based indicators I vision , Motion-based indicators I motion 、Heat-based indicators I thermal 、Battery-based indicators I battery and engine and position based indicators I location To estimate the severity of the accident that occurred in vehicle 10 SL In one embodiment, one or more controllers 20 may be configured based on a lookup table 46 ( Figure 1 As shown) the severity of the accident occurring in the vehicle 10 is estimated A SL In one embodiment, the lookup table 46 is stored in a memory of one or more controllers 20. Based on the audio-based indicator I stored in the lookup table 46 audio , Vision-based indicators I vision , Motion-based indicators I motion 、Heat-based indicators I thermal 、Battery-based indicators I battery and engine and position based indicators I location The value of A is used to estimate the severity of the accident. sL Alternatively, in another embodiment, the severity of the accident occurring to the vehicle 10 is determined based on a fuzzy logic severity estimation technique. SL Severity A SL Indicates the severity level of the incident.

[0044] refer to Figure 1 and Figure 2 In the exemplary embodiment shown, the audio-based input includes a non-speech-based input A and a speech-based input B. The non-speech-based input A represents a non-verbal sound generated by the occupant 14, such as crying, groaning, breathing noise, or the sound of the occupant 14 hitting the interior of the vehicle 10. The speech-based input B represents a speech-based sound generated by the occupant 14, such as a request for help or an expression of an emotion such as fear, surprise, or shock.

[0045] In one embodiment, the one or more controllers 20 use three-value or ternary logic to combine the audio-based inputs A and B with each other to determine the audio-based indicator I. audio For example, Figure 3 As shown, a ternary logic OR table 50 is shown, where audio-based inputs A and B are combined based on the ternary OR-based logic. Table 50 indicates a truth value indicated by T, a false truth value indicated by F, and one truth value of an unknown truth value indicated by U. Figure 3In the example shown, when the truth value is false, it indicates that the microphone 22 does not detect any non-speech-based sound corresponding to the non-speech-based input A. Similarly, a false truth value also indicates that the microphone 22 does not detect any speech-based sound corresponding to the speech-based input B. An unknown truth value indicates that the microphone 22 does not detect a signal. It should be understood that although ternary logic is described, in another embodiment, binary logic can be used instead to combine the audio-based inputs A and B with each other to determine the audio-based indicator I. audio It should be understood that although the present disclosure only shows a ternary logic OR table 50 for combining audio-based inputs A, B, a ternary logic table may be provided to combine vision-based inputs C, D from the vision system 24, thermal inputs E, F from the thermal event system 28, and multiple engine- and position-based inputs G, H.

[0046] refer to Figure 1 and Figure 2 In the exemplary embodiment shown, the vision-based inputs C, D from the vision system 24 include an in-vehicle input C indicating a visual state of an occupant 14 in the interior cabin 18 of the vehicle 10. The in-vehicle input C may indicate an event, such as, but not limited to, evidence of a collision and consciousness of the occupant 14. The exterior input D indicates the exterior 16 of the vehicle 10 and the visual state of the occupant 14 outside the vehicle 10. For example, the exterior input D may indicate when the occupant 14 extends his or her hand out of the window of the vehicle 10. In a non-limiting embodiment, the one or more controllers 20 use a ternary logic OR table to combine the in-vehicle input C and the exterior input D with each other to determine the vision-based indicator I vision .

[0047] In the exemplary embodiment shown, the plurality of motion-based inputs include an acceleration and deceleration input M1, a collision angle input M2, a rollover input M3, and a speed change input M4. The acceleration and deceleration input M1 indicates the acceleration or deceleration that occurred during the accident for the vehicle 10. The collision angle input M2 indicates the collision angle and collision direction at which the vehicle 10 was hit during the accident. The rollover input M3 indicates whether the vehicle 10 rolled over during the accident. Finally, the speed change input M4 indicates the speed change of the vehicle 10 within a time window during the accident.

[0048] In one embodiment, the one or more controllers 20 combine the acceleration and deceleration input M1, the collision angle input M2, the rollover input M3, and the speed change input M4 based on a weighted sum model to determine a motion-based index I motion In one non-limiting embodiment, the weighted sum model is expressed in Equation 1 as:

[0049]

[0050] Wherein, the weighted sum model includes a unique weight factor for each motion-based input, and each weight factor is ordered according to the level of importance. In another embodiment, the weighted sum model includes a unique weight factor for each motion-based input, which indicates the degree of certainty quantized as a function of the variance. Specifically, ω M1 represents the first weighting factor corresponding to the acceleration and deceleration input M1, ω M2 represents the second weighting factor corresponding to the collision angle input M2, ω M3 represents the third weight factor corresponding to the rollover input M3, ω M4 represents a fourth weighting factor corresponding to the speed change input M4 (within the time window), wherein the sum of the first weighting factor, the second weighting factor, the third weighting factor and the fourth weighting factor is equal to 1.

[0051] refer to Figure 1 and Figure 2 In the exemplary embodiment shown, the thermal inputs E and F from the thermal event system 28 include an interior cabin temperature or thermal input E and an engine compartment thermal input F. The interior cabin thermal input E indicates that a thermal event exists in the interior cabin 18. The engine compartment input F indicates that a thermal event exists in the engine compartment of the vehicle 10. In one embodiment, the one or more controllers 20 use ternary logic to combine the multiple thermal inputs E and F with each other to determine a thermal-based indicator I thermal .

[0052] refer to Figure 1 and Figure 2 In the exemplary embodiment shown, the battery-based input K from the battery pack 40 of the vehicle 10 indicates the health of the battery pack 40 of the vehicle 10. Figure 4 As shown, in a non-limiting embodiment, a table 52 is shown, which indicates the method for determining the battery-based indicator I battery Since there is only a single input K, the one or more controllers 20 determine the battery-based indicator I based on the truth value of the battery-based input K. battery For example, if the battery management system 30 indicates that the battery pack 40 is not functioning properly, the truth value is false. If the battery management system 30 indicates that the battery pack 40 is functioning properly, the truth value is true. If no signal is received from the battery management system 30, the truth value is unknown.

[0053] refer to Figure 1 and Figure 2In the exemplary embodiment shown, the engine status system 32 provides a plurality of engine and position based inputs G, H, including an engine status input G and a position input H. The engine status input G indicates the state of the internal combustion engine 48 (e.g., on or off). The position input H indicates the position of the vehicle 10. In a non-limiting embodiment, one or more controllers 20 use a ternary logical AND table to combine the engine status input G and the position input H with each other to determine the engine and position based indicator I location .

[0054] refer to Figure 1 and Figure 2 In one embodiment, one or more controllers 20 perform a control operation based on the information stored in the lookup table 46 ( Figure 1 Audio-based indicators I audio , Vision-based indicators I vision , Motion-based indicators I motion 、Heat-based indicators I thermal 、Battery-based indicators I battery and engine and position based indicators I location The comparison value is used to estimate the severity of the accident that occurred in the vehicle 10. SL As an example, if corresponding to each indicator (i.e., audio-based indicator I audio , Vision-based indicators I vision , Motion-based indicators I motion 、Heat-based indicators I thermal 、Battery-based indicators I battery and engine and position based indicators I location ) is unknown, then one or more controllers 20 determine the severity A SL is unknown. If all values ​​of the indicator indicate that no event has occurred (e.g., False), then severity A SL Indicates that no event has occurred. If an indicator in the lookup table 46 indicates an event (e.g., true), then the severity level A SL The severity level of the incident is low. An example of a low severity incident is when the airbags do not deploy, the occupant 14 is normal and does not make sounds or eye movements indicative of an incident, and the motion-based indicator system 26 registers a higher than normal acceleration (e.g., 2 g-force). Finally, if at least two indicators in the lookup table indicate an incident, then severity level A is high. SL Indicating a high level of severity of the accident. An example of a high severity accident is when the airbag is deployed (first and second stages), no occupant motion is detected, and the motion-based indicator system 26 records an acceleration of approximately 30 g-force.

[0055] Or, in another embodiment, based on Figure 5 The fuzzy logic severity estimation technique shown is used to estimate the severity of the accident that occurred in the vehicle 10. SL . refer to Figure 5 The one or more controllers 20 include a fuzzification block 60, a fuzzy reasoning block 62, one or more fuzzy rule databases 64, and a defuzzification block 66. The fuzzification block 60 of the one or more controllers 20 generates a membership function for each index (i.e., an audio-based index I audio , Vision-based indicators I vision , Motion-based indicators I motion 、Heat-based indicators I thermal 、Battery-based indicators I battery and engine and position based indicators I location ) is assigned a corresponding fuzzy value. Specifically, the fuzzy value of each indicator represents the degree of truth, ranging from 0 to 1.0. Figure 5 In the example shown, the fuzziness value includes an audio-based fuzziness index F audio , Vision-based fuzzy index F vision , motion-based blur index F motion , thermal-based fuzzy index F thermal , battery-based fuzzy index F battery and the fuzzy index F based on the engine and position location .

[0056] The fuzzy reasoning block 62 of the one or more controllers 20 receives the corresponding fuzzy value of each indicator as input and estimates the corresponding fuzzy value of each indicator based on one or more fuzzy logic rules stored in one or more fuzzy rule databases 64 to determine a fuzzified output value. The fuzzified output value indicates the severity of the accident that occurred to the vehicle 10. SL And the severity A SL The defuzzification block 66 converts the fuzzified output value into a clear value based on the true value. Specifically, the clear value is the severity of the accident A SL Specifically, when the true value is at least the threshold value, the defuzzification block 66 determines the severity of the accident indicated by the fuzzified output value A SL is correct. In one embodiment, the threshold is fifty percent, or 0.5.

[0057] Figure 6 is a schematic diagram illustrating a cause-action diagram 100 for selecting one or more remedial actions 102. Figure 1 and Figure 6 In one embodiment, the one or more controllers 20 may determine the severity of an accident occurring with the vehicle 10 corresponding to one or more occupants 14 based on the severity A of the accident.SL to select one or more remedial actions 102. The one or more remedial actions 102 represent activities to help one or more occupants 14 of the vehicle 10 after the accident. Some examples of remedial actions 102 include, but are not limited to, contacting emergency personnel, unlocking the doors of the vehicle 10, lowering the windows of the vehicle 10, sending an SOS signal to one or more vehicles within a predetermined radius of the vehicle 10, and contacting non-emergency personnel.

[0058] In a non-limiting embodiment, one or more controllers 20 may determine the severity of an accident that has occurred with the vehicle 10. SL , thermal event index S0, water event index S1, and root cause probability index P(E|π) to select one or more remedial actions 102. Thermal event index S0 indicates that a thermal event has occurred in the interior cabin 18 and the engine compartment of the vehicle 10. Water event index S1 indicates that the vehicle 10 is in a body of water, such as a lake or a river. Figure 6 The cause-action diagram 100 shown in includes a hard / soft thresholding block 106 corresponding to a thermal event indicator S0, a hard / soft thresholding block 108 corresponding to a water event indicator S1, and a hard / soft thresholding block 110 for a root cause probability indicator P(E|π). In a non-limiting embodiment, when the thermal event indicator s0 is equal to or greater than 0.5, one or more controllers 20 determine that a thermal event has occurred. Similarly, when the water event indicator S1 is equal to or greater than 0.5, one or more controllers 20 determine that the vehicle 10 is in a body of water. The cause-action diagram 100 also includes a cause-action mapping block 112, which receives the thermal event indicator S0, the water event indicator S1, the root cause probability indicator P(E|π), and the severity A of the accident of the vehicle 10. SL , and perform cause-action mapping to determine the root cause of the incident.

[0059] The root cause probability index P(E|π) indicates the probability that the root cause of the accident is true. In a non-limiting embodiment, the root cause probability index P(E|π) is based on a water event probability P0, a thermal event probability P1, a rollover probability P2, a failure probability P3, and a high impact collision probability P5. The water event probability P0 indicates the probability that the root cause of the accident is because the vehicle 10 is in a body of water, the thermal event probability P1 indicates the probability that the root cause of the accident is smoke or flames in the vehicle 10, the rollover probability P2 indicates the probability that the root cause of the accident is because the vehicle 10 rolls over, the failure probability P3 indicates the probability that the root cause of the accident is because one or more system failures occur in the vehicle 10, and the high impact collision probability P5 indicates the probability that the root cause of the accident is a high impact collision of the vehicle 10. The water event probability P0, the thermal event probability P1, the rollover probability P2, the failure probability P3, and the high impact collision probability P5 are combined in the noise or model 104. In one embodiment, the root cause probability index (E|π) is determined based on Equation 2 as:

[0060]

[0061] Among them, S represents the set of parents that are true (on), and π i Include a range containing all parents of the desired value E, E1+E0=1, where n=5, in this example, for example, P1, P2, P3, P4, P5, represents the number of potential root causes.

[0062] In a non-limiting embodiment, the root cause of the accident may be one or more of the following: a rollover event, a failure event in which one or more vehicle systems fail to operate normally, and a high impact collision. In a non-limiting embodiment, when the root cause probability index (E|π) is equal to or greater than 0.5, the one or more controllers 20 determine that the root cause is true. As an example, when the severity A of the accident that occurs to the vehicle 10 is SL When the thermal event indicator S0 is true, the water event indicator S1 is false, and the root cause probability indicator P(E|π) indicates that none of the root causes is true, one or more controllers 20 select to contact emergency personnel, send an SOS signal to one or more vehicles within a predetermined radius of the vehicle 10, and contact non-emergency personnel as remedial actions.

[0063] Referring generally to the accompanying drawings, the disclosed accident severity estimation system has various technical effects and benefits. Specifically, the accident severity estimation system employs a multimodal approach to estimate the severity of an accident. The severity may be used to select one or more remedial actions to help an occupant after the accident occurs. In other words, the disclosed accident severity estimation system may select remedial actions based on the severity of the accident, which may allow more appropriate or useful actions to be taken to help an occupant after the accident occurs.

[0064] The controller may refer to an electronic circuit, a combinational logic circuit, a field programmable gate array (FPGA), a processor (shared, dedicated or grouped) that executes code, such as in a system on a chip, or a combination of some or all of them, or a part of an electronic circuit, a combinational logic circuit, a field programmable gate array (FPGA), a processor that executes code. In addition, the controller may be based on a microprocessor, such as a computer having at least one processor, a memory (RAM and / or ROM), and associated input and output buses. The processor may run under the control of an operating system residing in the memory. The operating system may manage computer resources, and the computer program code embodied as one or more computer software applications (e.g., applications residing in the memory) may have instructions executed by the processor. In an alternative embodiment, the processor may execute the application directly, in which case the operating system may be omitted.

[0065] The description of the present disclosure is merely exemplary in nature, and variations that do not depart from the gist of the present disclosure are intended to fall within the scope of the present disclosure. Such variations should not be regarded as departing from the spirit and scope of the present disclosure.

Claims

1. An accident severity estimation system for estimating the severity of a vehicle accident, the accident severity estimation system comprising: one or more microphones that capture a plurality of audio-based inputs indicative of verbal and non-verbal sounds uttered by one or more occupants of the vehicle; a vision system that captures a plurality of vision-based inputs representing image data indicative of the occupant; a motion-based input system that collects a plurality of motion-based inputs indicative of motion of said vehicle during said incident; a thermal event system that collects a plurality of thermal inputs indicative of thermal events within the vehicle; a propulsion system providing state-based inputs for a propulsion system of the vehicle; as well as one or more controllers in electronic communication with the one or more microphones, the vision system, the motion-based input system, the thermal event system, and the propulsion system, the one or more controllers executing instructions to: combining the plurality of audio-based inputs based on at least a two-valued logic system to determine an audio-based indicator, combining the plurality of visual-based inputs based on at least a two-valued logic system to determine a visual-based indicator, and combining the plurality of thermal inputs based on at least a two-valued logic system to determine a thermal-based indicator; combining the plurality of motion-based inputs from the motion-based input system based on a weighted sum model to determine a motion-based metric; determining a propulsion-based indicator based on the state-based input; as well as A severity of the accident is estimated based on the audio-based indicator, the visual-based indicator, the thermal-based indicator, the motion-based indicator, and the propulsion-based indicator.

2. The accident severity estimation system according to claim 1, wherein: The one or more controllers execute instructions to: A severity of the accident occurring with the vehicle is estimated based on comparison values ​​of the audio-based indicator, the vision-based indicator, the motion-based indicator, the thermal-based indicator, and the propulsion-based indicator stored in a lookup table.

3. The accident severity estimation system according to claim 1, wherein: The one or more controllers execute instructions to estimate a severity of the accident based on a fuzzy logic severity estimation technique.

4. The accident severity estimation system according to claim 3, wherein: The fuzzy logic severity estimation technique includes: Assign a corresponding fuzzy value to each indicator based on the membership function; estimating a corresponding fuzzy value of each indicator based on one or more fuzzy logic rules to determine a fuzzified output value, wherein the fuzzified output value indicates a severity of the accident and a true value corresponding to the severity; and The fuzzified output value is converted into the severity of the accident based on the true value.

5. The accident severity estimation system according to claim 1, wherein: The one or more controllers execute instructions to select one or more remedial actions based on a severity of the incident, a thermal event indicator, a water event indicator, and a root cause probability indicator.

6. The accident severity estimation system according to claim 5, wherein: The one or more remedial actions include one or more of: contacting emergency personnel, unlocking a door of the vehicle, lowering a window of the vehicle, sending an SOS signal to one or more vehicles within a predetermined radius of the vehicle, and contacting non-emergency personnel.

7. The accident severity estimation system according to claim 1, wherein: The propulsion system includes a battery pack that provides power to one or more electric motors.

8. The accident severity estimation system according to claim 7, wherein: The status-based input is a battery-based input indicative of a health of the battery pack.

9. The accident severity estimation system according to claim 1, wherein: The propulsion system includes an internal combustion engine.

10. The accident severity estimation system according to claim 9, wherein: The state-based inputs include an engine- and position-based state input indicative of a state of the internal combustion engine and a position input indicative of a position of the vehicle.

Images