System, method, and apparatus for classifying passengers in vehicle seats
The passenger classification system uses two seat weight sensors and vehicle motion data to enhance accuracy and reliability in passenger classification, addressing errors from vehicle and passenger behavior, thereby improving safety device operations.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- ADVANCED MANUFACTURING ZF AUTOMOTIVE TECHNOLOGY (GUANGZHOU) CO LTD
- Filing Date
- 2020-07-15
- Publication Date
- 2026-06-04
Smart Images
Figure 0007870138000002 
Figure 0007870138000003 
Figure 0007870138000004
Abstract
Description
Technical Field
[0001]
[0001] The present invention generally relates to vehicle safety systems. More particularly, the present invention relates to a vehicle safety system that implements a passenger classification system for detecting and classifying passengers in a vehicle seat so that a safety system can operate a vehicle safety device such as an airbag (particularly, a front airbag) according to the detected passenger classification (e.g., size, weight, presence, etc.).
Background Art
[0002]
[0002] Systems, methods, and devices for detecting and classifying passengers in a vehicle seat are known. Referring to FIG. 1, a safety system 10 includes a controller 12 operably connected to four seat weight sensors (SWS1, SWS2, SWS3, SWS4) that measure the weight of a vehicle seat 14. Typically, as shown in FIG. 1, the seat weight sensors are disposed at four corner positions on a seat frame / rail 16 where the seat 14 is supported by the vehicle. These corner positions are Front Left (Fr-L), Rear Left (Rr-L), Front Right (Fr-R), and Rear Right (Rr-R).
[0003]
[0003] Conventional weight-based passenger classification systems use the sensed passenger weight as the only parameter when determining passenger classification and thus rely on inputs from all four corner positions. As shown in FIG. 1, SWS1 is disposed at the Fr-L position of the vehicle seat 14. SWS2 is disposed at the Rr-L of the vehicle seat 14. SWS3 is disposed at the Fr-R position of the vehicle seat 14. SWS4 is disposed at the Rr-R position of the vehicle seat 14. Together, the seat weight sensors can measure the weight of a passenger on the seat as the sum of SWS1 + SWS2 + SWS3 + SWS4.
[0004]
[0004] Naturally, the seat weight sensor responds to the inclination of the passenger load resulting from the behavior of both the vehicle and the passenger. Examples of changes in vehicle behavior include changes due to high acceleration or deceleration, cornering or turning operations, and rough roads, while examples of changes in passenger behavior include changes in the passenger's position.
[0005]
[0005] The seat weight sensor response is a form of increase and / or decrease in weight sensed at various corner positions of the vehicle seat, which can lead to erroneous changes in passenger classification. Conventional vehicle safety systems with seat weight sensors at four corner positions are better at handling these tilting passenger loads, and when such tilting loads is detected, changes in passenger classification can be prohibited. [Overview of the project] [Means for solving the problem]
[0006]
[0006] The system, method, and apparatus for classifying passengers in vehicle seats relies solely on two seat weight sensor inputs while being robust to inclinations of passenger loads resulting from vehicle and / or passenger behavior. Changes in vehicle and / or passenger behavior can dramatically affect the loads sensed by the two-sensor passenger classification system due to the influence of unsensed load paths, but the system utilizes vehicle motion parameters such as wheel speed, lateral acceleration, and longitudinal acceleration, as well as its own filtering scheme for determining passenger classification.
[0007]
[0007] Passenger classification is used to operate or adjust the deployment of the front airbags. Therefore, it is important for the vehicle safety system to accurately and reliably classify the passenger in the front passenger seat.
[0008]
[0008] According to one embodiment, a method for determining the passenger class of a vehicle seat includes obtaining first and second seat weight indicators of the vehicle seat via first and second seat weight sensors. The first seat weight sensor is located on the side of the vehicle seat at a forward position of the vehicle seat. The second seat weight sensor is located on the side of the vehicle seat at a rear position of the vehicle seat. The method also includes obtaining a vehicle acceleration value from a vehicle acceleration sensor. The method also includes determining the raw weight of the vehicle seat as twice the sum of the first and second seat weight indicators, and determining a filtered weight based on the raw weight. The method further includes determining the passenger class based on the filtered weight, depending on whether the vehicle acceleration value is less than a predetermined value.
[0009]
[0009] In another embodiment, the method may include suppressing classification changes in response to the vehicle acceleration value being greater than a predetermined threshold.
[0010]
[0010] In another embodiment, suppressing classification changes, either alone or in combination with any other embodiment, may include suppressing classification for time delay.
[0011]
[0011] According to another embodiment, the vehicle acceleration value may include vehicle lateral acceleration and / or vehicle longitudinal acceleration, either alone or in combination with any other embodiment.
[0012]
[0012] In another embodiment, either alone or in combination with any other embodiment, the passenger class can be selected from one of the following classes: namely, the no-passenger class, the child seat class, the small adult class, and the large adult class. In this embodiment, the no-passenger class can be associated with a measured seat weight up to a first weight, the child seat class can be associated with a measured seat weight from a first weight up to a second weight greater than the first weight, the small adult class can be associated with a measured seat weight from a second weight up to a third weight greater than the second weight, and the large adult class can be associated with a measured seat weight of the third weight or greater. For example, the first weight may be about 10.8 kg, the second weight may be about 29.4 kg, and the third weight may be about 54.8 kg.
[0013]
[0013] In another embodiment, determining the filtered weight, either alone or in combination with any other embodiment, may include selecting one of the unfiltered weight, short-filtered weight, and long-filtered weight. The short-filtered weight is determined using a low-pass filter with a relatively short time constant, and the long-filtered weight is determined using a low-pass filter with a relatively long time constant. For example, the relatively short time constant may be about 1 to 2 seconds, and the relatively long time constant may be about 5 to 20 seconds.
[0014]
[0014] In another embodiment, selecting one of the unfiltered weight, short-filtered weight, and long-filtered weight, either alone or in combination with any other embodiment, may include selecting the long-filtered weight in response to a determination that the vehicle is moving and the seats are occupied, and selecting the short-filtered weight in response to a determination that the vehicle is not moving and the seats are not occupied.
[0015]
[0015] In another embodiment, selecting one of the unfiltered weight, short-filtered weight, and long-filtered weight, either alone or in combination with any other embodiment, may include selecting the unfiltered weight for a predetermined start time, selecting the short-filtered weight after the start time has elapsed, selecting the long-filtered weight in response to a determination that the vehicle is moving and the seats are occupied, and selecting the short-filtered weight in response to a determination that the vehicle is not moving and the seats are not occupied.
[0016]
[0016] In another embodiment, selecting one of the unfiltered weight, short-filtered weight, and long-filtered weight, either alone or in combination with any other embodiment, may include selecting the unfiltered weight during the initial start-up time, selecting the short-filtered weight until the vehicle is moving and the seats are occupied, and then selecting the long-filtered weight until the vehicle comes to a stop.
[0017]
[0017] In another embodiment, the method may, either alone or in combination with any other embodiment, include disabling a determined passenger class in response to determining that a seat belt buckle associated with a vehicle seat is released.
[0018]
[0018] In another embodiment, determining a passenger class based on filtered weight, either alone or in combination with any other embodiment, may include implementing a hysteresis logic function that assigns a passenger class based on filtered weight and prevents changes in the assigned passenger class due to variations in filtered weight due to seat load in response to vehicle operation and / or changes in the passenger's position on the seat. The hysteresis logic function implements superimposing weight ranges for each passenger class, with each weight range consisting of a high threshold and a low threshold. The hysteresis logic function can assign the next highest passenger class depending on whether the filtered weight exceeds the high threshold. The hysteresis logic function can assign the next lowest passenger class depending on whether the filtered weight falls below the low threshold. Each weight range may also include a fitting threshold. The fitting threshold can be the nominal weight value of the corresponding passenger class and can be used to determine the initial passenger class.
[0019]
[0019] In another embodiment, determining a passenger class based on filtered weight, either alone or in combination with any other embodiment, may include assigning a passenger class based on filtered weight and implementing a hysteresis logic function to prevent changes in the assigned passenger class due to fluctuations in filtered weight caused by seat loads in response to vehicle operation, and / or changes in the position of passengers on the seats.
[0020]
[0020] In another embodiment, determining the filtered weight based on raw weight, either alone or in combination with any other embodiment, may include implementing a freeze filter function that freezes the filtered weight value in response to the vehicle lateral acceleration exceeding a predetermined threshold.
[0021]
[0021] According to another aspect, alone or in combination with any other aspect, the first seat weight sensor measures the seat weight at a front inner mounting position of the seat, and the second seat weight sensor measures the seat weight at a rear inner mounting position of the seat. Alternatively, the first seat weight sensor measures the seat weight at a front outer mounting position of the seat, and the second seat weight sensor measures the seat weight at a rear outer mounting position of the seat.
[0022]
[0022] According to another aspect, alone or in combination with any other aspect, a passenger classification system for determining a passenger class associated with a vehicle seat includes a controller operable to implement a passenger classification algorithm according to a method for determining the passenger class of a vehicle seat.
[0023]
[0023] According to another aspect, alone or in combination with any other aspect, a vehicle safety system can include a passenger classification system, at least one vehicle passenger safety device, and an airbag ECU operable to control the operation of the at least one vehicle passenger safety device. The airbag ECU is operably connected to the controller of the passenger classification system. The airbag ECU is configured to control the operation of the at least one vehicle passenger safety device according to the passenger classification determined via the passenger classification algorithm implemented in the controller.
Brief Description of the Drawings
[0024] [Figure 1]
[0024] It is a schematic diagram of a vehicle seat and a vehicle safety system. [Figure 2]
[0025] It is a block diagram showing a passenger classification system implemented by a vehicle safety system. [Figure 3]
[0026] It is a block diagram showing a passenger classification algorithm implemented by a passenger classification system. [Figure 4]
[0026] It is a block diagram showing a passenger classification algorithm implemented by a passenger classification system. [Figure 5]
[0026] This is a block diagram showing a passenger classification algorithm implemented by a passenger classification system. [Figure 6]
[0026] This is a block diagram showing a passenger classification algorithm implemented by a passenger classification system. [Figure 7]
[0026] This is a block diagram showing a passenger classification algorithm implemented by a passenger classification system. [Figure 8]
[0026] This is a block diagram showing a passenger classification algorithm implemented by a passenger classification system. [Figure 9]
[0026] This is a block diagram showing a passenger classification algorithm implemented by a passenger classification system. [Figure 10]
[0026] This is a block diagram showing a passenger classification algorithm implemented by a passenger classification system.
Mode for Carrying Out the Invention
[0025]
[0027] Referring to FIG. 1, the safety system 10 includes a controller 12 operatively connected to four seat weight sensors (SWS1, SWS2, SWS3, SWS4) that measure the weight of the vehicle seat 14 on the passenger side. The seat weight sensors are at four corner positions of the seat frame / rail 16, where the seat 14 is supported on the passenger side of the vehicle. These corner positions are front left (Fr-L), rear left (Rr-L), front right (Fr-R), and rear right (Rr-R).
[0026]
[0028] Referring to FIG. 2, the vehicle safety system 10 includes / implements a passenger classification system 40 for determining whether a passenger is seated in the passenger seat and, if so, classifying the passenger so seated. The classification is known and specified in the field of vehicle safety systems. For example, the classification includes seats that are empty, occupied by a child or a child seat, occupied by a small adult (e.g., 5% of female passengers), or occupied by a large adult (e.g., 50% of male passengers).
[0027]
[0029] The passenger classification determined by the passenger classification system 40 may be used by the airbag controller or ECU 16 to control the operation of safety devices 18, such as airbags (e.g., front airbags) and load limiters, when a vehicle collision is detected. For example, if the passenger classification system 40 indicates that a seat is empty or occupied by a child and / or child safety seat, airbag deployment may be suppressed. If the passenger classification system 40 indicates that a seat is occupied by a small adult, the airbag may be deployed at a first intensity (timing, inflation fluid volume / flow rate, etc.). If the passenger classification system 40 indicates that a seat is occupied by a large adult, the airbag may be deployed at a second intensity greater than the first intensity, for example.
[0028]
[0030] The passenger classification system 40 implements the passenger classification algorithm 20 in the controller 12. The passenger classification algorithm 20 operates to determine the final passenger classification 30 based on the data input to the controller 12. The data can be obtained, for example, directly from sensors via a wired or wireless connection to the controller 12, or indirectly from other vehicle systems via a data / communication bus connection.
[0029]
[0031] The inputs to controller 12 include the following: • Front left seat weight sensor SWS1 • Rear left seat weight sensor SWS2 • Filtered vehicle wheel speed 22 • Filtered vehicle longitudinal acceleration 24 • Filtered vehicle lateral acceleration 26 • Passenger buckle condition 28
[0030]
[0032] It is important to note that only two seat weight sensors are used by the passenger classification algorithm 20 to determine the final passenger classification 30. As shown in Figure 2, these two sensors, SWS1 and SWS2, are located on the front left and rear left, respectively. Since the passenger classification system 40 is for the passenger side of the vehicle, these SWS1 and SWS2 are seat weight sensors located on the inside. However, the passenger classification system 40 can utilize SWS3 and SWS4, which are weight sensors located on the right / outer front and rear. Either pair, namely SWS1 and SWS2, or SWS3 and SWS4, can be implemented in the passenger classification system 40.
[0031]
[0033] The passenger classification system 40 classifies passengers in the vehicle seats 14 using only two seat weight sensors, SWS1 and SWS2, along with vehicle motion data. In the example configuration shown in the figure, the motion data is determined from filtered wheel speed 22, filtered vehicle longitudinal acceleration 24, and filtered vehicle lateral acceleration 26. Essentially, the passenger classification system 40 determines passenger weight as twice the weight measured by the two seat weight sensors SWS1 and SWS2. Recognizing that this measurement only works when no forces other than gravity are acting on the passenger, the passenger classification system 40 monitors the vehicle motion data to determine whether the vehicle longitudinal acceleration or vehicle lateral acceleration acting on the passenger would inaccurate the weight measured by SWS1 and SWS2.
[0032]
[0034] For example, when a vehicle makes a right turn, i.e., turns a corner on the right, the passengers lean inward due to inertia / centrifugal force. When this occurs, SWS1 and SWS2 perceive an increased weight due to the passenger's lean. Similarly, when a vehicle makes a left turn, i.e., turns a corner on the left, the passengers lean outward due to inertia / centrifugal force. When this occurs, SWS1 and SWS2 perceive a decreased weight due to the passenger's lean. To counteract this phenomenon, the passenger classification system 40 monitors the vehicle's motion data to determine when passenger leaning occurs or is likely to occur, and takes this into account in the determined passenger classification, for example, by prohibiting changes to the passenger classification.
[0033]
[0035] Figure 3 shows an overview of the passenger classification algorithm 20 implemented by the passenger classification system 40. The passenger classification algorithm 20 includes a vehicle speed determination function 50 that receives filtered vehicle wheel speeds 22 and generates a Boolean indicator, a vehicle movement flag output (VehicleMoving), i.e., VehicleMoving = YES or NO.
[0034]
[0036] The total weight calculation function 150 receives seat weight signals sensed from seat weight sensors SWS1 and SWS2. Function 150 calculates the total weight as twice the sum of the seat weight signals SWS1 and SWS2, i.e., TotalWeight = 2 × (SWS1 + SWS2).
[0035]
[0037] The classification change suppression function 100 receives filtered vehicle longitudinal acceleration 24, filtered vehicle lateral acceleration 26, and a vehicle movement flag (VehicleMoving). The classification change suppression function 100 generates a LongAccelHigh flag and a LatAccelHigh flag, which are Boolean indicators, i.e., LongAccelHigh = YES or NO, LatAccelHigh = YES or NO. The classification change suppression function 100 operates to determine whether the vehicle acceleration may affect the weight of the vehicle seats, which is determined via SWS1 and SWS2. In response to the determination that the seat weight may be affected, the classification change suppression function 100 operates to prohibit or suppress changes in the passenger classification determined by the passenger classification algorithm 20.
[0036]
[0038] The signal processing function 200 receives TotalWeight from the total weight function 150 and LongAccelHigh and LatAccelHigh flags from the classification change suppression function 100. The signal processing function 200 implements a freeze filter algorithm, a two-stage weight filter algorithm, and a filter switch control algorithm. The signal processing function 200 generates a filtered total weight signal (FilteredTotalWeight) and a filter selection flag (FilterSelect). The signal processing function 200 determines the filtered total weight, taking into account the vehicle's kinematics and any potential impact this may have on the measured seat weight.
[0037]
[0039] The signal processing function 200 includes hysteresis logic 400 that generates an InstantClass. The InstantClass is the determined classification of the passengers in the vehicle seats 14 based on the FilteredTotalWeight determined by the signal processing function 200. For example, the classifications shown in the table below are possible. [Table 1]
[0038]
[0040] Figure 4 is a functional block diagram showing the vehicle speed determination function 50 of the passenger classification algorithm 20 (see Figure 3) implemented by the passenger classification system 40. Function 50 begins in step 52 and proceeds to step 54, where the default state is established as VehicleMoving=OFF (or Boolean zero). Function 50 then proceeds to step 56, where it is determined whether the speed of at least one wheel is greater than a predetermined speed threshold for a period of time longer than a predetermined time threshold. The predetermined speed threshold in step 56 can be any value greater than zero.
[0039]
[0041] In step 56, if the wheel speed is greater than the speed threshold for a given time threshold, function 50 proceeds to step 58, where the vehicle movement flag is set to VehicleMoving=ON (or Boolean value 1). Next, process 50 proceeds to step 60, where a determination is made as to whether all wheel speeds have been less than or equal to a predetermined speed for a given time longer than a predetermined threshold. If the determination in step 60 is true (YES), the process returns to step 54, and the vehicle movement flag is reset to its default state, i.e., VehicleMoving=OFF. The predetermined speed in step 60 could be, for example, zero.
[0040]
[0042] In step 56, if the wheel speed is not greater than the speed threshold during the time threshold, function 50 returns to step 54. The vehicle movement flag is kept in its default state, i.e., VehicleMoving=OFF, and the process continues as described above.
[0041]
[0043] Figure 5 is a functional block diagram showing the classification change suppression function 100 of the passenger classification algorithm 20 (see Figure 3) implemented by the passenger classification system 40. Function 100 begins at step 102, where the count is initialized to be equal to zero. Function 100 proceeds to step 104, where the vehicle movement flag (see Figure 4) is queried. If the vehicle is not moving, function 100 proceeds to step 110, where a change in passenger classification is permitted.
[0042]
[0044] In step 104, if the vehicle is moving (VehicleMoving=ON), function 100 proceeds to step 106, where it is determined whether the lateral acceleration (LAT_ACCEL) is greater than the lateral acceleration threshold, or whether the longitudinal acceleration (LONG_ACCEL) is greater than the longitudinal acceleration threshold. If either of these conditions is true, function 100 proceeds to step 108 to prohibit / suppress the change in passenger classification. The count is set to equal the acceleration latch time (ACCEL_LATCH), which is an integer equal to the number of samples the system has acquired. Based on the system's sample rate, the count can be seen as acting as a time lag or delay and is used to prohibit the change in classification of several samples indicated by the count.
[0043]
[0045] If, in step 106, it is determined that neither condition is true, function 100 proceeds to step 112, where it is determined whether the count is greater than zero. If the count is greater than zero, function 100 proceeds to step 114, where it prohibits / suppresses the change in passenger classification and decrements the count by 1. This periodic countdown of the counter to zero functions as the aforementioned time delay. If, in step 112, it is determined that the count is less than or equal to zero, function 100 proceeds to step 110, where the change in passenger classification is permitted.
[0044]
[0046] From the above, it can be seen that the classification change suppression function 100 functions to suppress changes in passenger classification when the vehicle is subjected to longitudinal or lateral acceleration greater than a predetermined threshold, because these accelerations can affect the seat weight sensed via sensors SWS1 and SWS2. These acceleration thresholds can vary, and the lateral acceleration threshold is lower than the longitudinal acceleration threshold. For example, the lateral acceleration threshold is 0.18G (gravity) and the longitudinal acceleration threshold is 1.0G.
[0045]
[0047] Figure 6 is a block diagram showing the signal processing function 200 of the passenger classification algorithm 20 (see Figure 3) implemented by the passenger classification system 40. The signal processing function 200 includes a freeze filter 250 that receives a lateral acceleration high (Lat_Accel_High) flag 202 from a classification change suppression function (see Figure 5). The freeze filter 250 also receives a total weight display (TotalWeight) 204 from a total weight calculation function 150 (see Figure 3). The freeze filter 250 also receives a long-filtered weight (LongFilteredWeight) from a unit delay function 208. The freeze filter 250 also receives a boolean freeze filter enable flag (FreezeOK) from a filter switch control function 350 (see Figure 9). Using these inputs, the freeze filter 250 performs a function (see Figure 7) that generates a weight output (WeightOut) and a boolean freeze weight flag (WeightFrozen), where ON = Freeze and OFF = Unfreeze.
[0046]
[0048] Figure 7 is a block diagram showing the freeze filter function 250 of the signal processing function 200 of the passenger classification algorithm 20 (see Figure 3) implemented by the passenger classification system 40. The freeze filter function 250 starts at 252 and proceeds to step 254, where the default state is established. In the default state, the WeightOut of the freeze filter function 250 is set to equal to the TotalWeight, and the WeightFrozen flag is set to OFF. The freeze filter function 250 proceeds to step 256, where a determination is made as to whether the FreezeOK flag is ON or OFF (see Figure 9, filter switch control function 350).
[0047]
[0049] If the freeze filter enable flag is OFF, the freeze filter function 250 returns to step 254 and proceeds as described above, and the function operates with a loop delay until the freeze filter enable flag is switched ON. When this occurs (FreezeOK=ON), the function 250 proceeds to step 258, where it is determined whether the vehicle lateral acceleration (LAT_ACCEL) is greater than or equal to the vehicle lateral acceleration threshold (e.g., 0.18G) for the minimum time, whether the weight history is available, and whether the freeze filter enable flag (FreezeOK) is ON. If this determination is NO, the freeze filter function 250 returns to step 254 and proceeds as described above, and the function operates with a loop delay until the freeze filter enable flag is switched ON.
[0048]
[0050] If the determination in step 258 of the freeze filter function 250 is YES, the function proceeds to step 260 and enters a frozen state, where WeightOut is set to equal to the oldest historical weight sample and the WeightFrozen flag is set to ON. Next, the freeze filter function 250 proceeds to step 262, where it is determined whether the vehicle lateral acceleration (LAT_ACCEL) is less than the vehicle lateral acceleration threshold (e.g., 0.18G) for the minimum amount of time. If this determination is YES, the freeze filter function 250 returns to step 254 and proceeds as described above, and the function operates with a loop delay until the freeze filter enable flag is switched to ON. If the determination in step 262 is NO, the freeze filter function 250 returns to step 260 and proceeds as described above, and the function operates with a loop delay until the determination in step 262 is switched to YES.
[0049]
[0051] Referring again to Figure 6, the signal processing function 200 also includes a two-stage weight filter 300 that receives the weight output from the freeze filter 250. The two-stage weight filter 300 performs the function of generating raw weight, short-filtered weight, and long-filtered weight (see Figure 8).
[0050]
[0052] Figure 8 is a block diagram showing the two-stage weight filter 300 of the signal processing function 200 of the passenger classification algorithm 20 (see Figure 3) implemented by the passenger classification system 40. The two-stage weight filter receives the weight output (WeightOut) from the freeze filter function 250 (see Figures 6 and 7), as shown in 302. The WeightOut passes through the two-stage weight filter function 300 and is output as the raw weight, as shown in 304. The short filter 306 receives the WeightOut, filters it, and outputs the short-filtered weight, as shown in 308. The short filter is a moving average FIR (finite impulse response) filter with a small time constant. Typical time constant values are within 1 to 2 seconds. The long filter 310 receives the ShortFilteredWeight, filters it, and outputs the long-filtered weight, as shown in 312. A long filter is an IIR (Infinite Impulse Response) filter with a large time constant. Typical time constant values are within 5 to 20 seconds.
[0051]
[0053] Referring again to Figure 6, the signal processing function 200 also includes a filter switch control function 350 that receives a vehicle movement flag (VehicleMoving) from the vehicle speed determination function 50 (see Figures 3 and 4). Filter switch control function 350 This function generates a freeze filter enable flag (FreezeOK) and a filter selector (SelectFilter) output (see Figure 9).
[0052]
[0054] Figure 9 is a block diagram showing the filter switch control function 350 of the signal processing function 200 of the passenger classification algorithm 20 (see Figure 3) implemented by the passenger classification system 40. The filter switch control function 350 starts at 352 and proceeds to step 354, entering the startup state. In the startup state, the filter selector is initialized and uses the raw weight (SelectFilter=UseRawWeight) determined by the two-stage weight filter 350 (see Figure 8) for a fixed time. Next, the filter switch control function 350 proceeds to step 356, where, after the fixed time of step 354 has elapsed, the short filter 306 (see Figure 8) is preset to the raw weight (RawWeight), and the filter switch control function 350 transitions to the short filter state and proceeds to step 358. In the short filter state, the filter selector is set to use the short filtered weight (SelectFilter=UseShortFilteredWeight).
[0053]
[0055] The fixed time value implemented in step 356 depends on the time required for the short filter output to stabilize within approximately ±5% of a certain final value, given a constant input. Ideally, the fixed time value is less than 1 second. If it is guaranteed that the short filter output is stable when the passenger classification algorithm is first invoked by the airbag ECU, the fixed time can be shortened, and may even be set to 0 seconds.
[0054]
[0056] The filter switch control function 350 proceeds to step 362, where it is determined whether the vehicle is moving or not (VehicleMoving=YES, see Figure 4), and it is determined that a seat is occupied. The seat occupancy determination can be determined to be YES, for example, if the measured seat weight (e.g., FilteredTotalWeight) is greater than a predetermined threshold, typically 10KG. If the determination in step 362 is NO, the filter switch control function 350 returns to step 358 and proceeds as described above, and the function operates with a loop delay until the determination in step 362 switches to YES.
[0055]
[0057] If the determination in step 362 is Yes, the filter switch control function 350 proceeds to step 366, where the long filter (see Figure 8) is preset to ShortFilteredWeight, transitioning to the long filter state and proceeding to step 364. In the long filter state, the filter selector is set to use LongFilteredWeight (SelectFilter=UseLongFilteredWeight). The filter switch control function 350 proceeds to step 360, where it is determined whether the vehicle is stationary (VehicleMoving=NO, see Figure 4) and the seats are not occupied, for example, FilteredTotalWeight < 10KG. If the determination in step 360 is NO, the filter switch control function 350 returns to step 364 and proceeds as described above, and the function operates with a loop delay until the determination in step 360 switches to YES. If the determination in step 360 is Yes, the filter switch control function 350 returns to step 358 and enters the short filter state.
[0056]
[0058] From the above, it will be understood that the filter switch control function 350 operates to determine which of the measured seat weight values will be used to determine passenger classification. The raw weight is used during the initial startup time, and then the short-filtered weight is used until it is determined that the vehicle is moving and the seats are occupied. Once this is determined, the long-filtered weight is used.
[0057]
[0059] Referring again to Figure 6, the signal processing function 200 also includes a first weight selector 210 that operates to select from the weights provided by the two-stage weight filter function 300 (see Figure 8) based on a filter selection value (SelectFilter) from the filter switch control function 350 (see Figure 9). The first weight selector 210 provides the selected filtered weight output (SelectedFilteredWeight) to a second weight selector 212. The second weight filter 212 selects either the selected filtered weight output (SelectedFilteredWeight) or the weight output (WeightOut) from the freeze filter function 250 (see Figure 7) based on the weight frozen flag (WeightFrozen) from the freeze filter function. If the weight is frozen (WeightFrozen=YES), the freeze filtered weight output (WeightOut) is selected. If the weight is not frozen (WeightFrozen=NO), the selected filtered weight is selected. The weight value selected by the second weight selector 212 is output as the Filtered Total Weight.
[0058]
[0060] Referring again to Figure 3, the Filtered Total Weight is provided to the hysteresis logic function 400 shown in Figure 10. The hysteresis logic function 400 operates to provide hysteresis to the passenger class determination. Referring to Figure 10, the four passenger classifications (Class 0 to Class 3) are associated with four corresponding weight ranges (WR0 to WR3). Since the Filtered Total Weight is shown on the Y-axis, we can see that the classes and their respective weight ranges increase from Class 0 / WR0 to Class 3 / WR3. The classes can correspond to, for example, the classes shown in Table 1. The weight ranges include the respective weight values shown in Table 1 and overlap each other as shown in Figure 10.
[0059]
[0061] Each region of overlapping class or weight range includes three thresholds: a low weight / class threshold T(n)L, a fitting weight / class threshold T(n)C, and a high weight / class threshold T(n)H. The fitting weight / class threshold T(n)C represents the nominal weight value of the class, i.e., the weight specified in Table 1. The fitting weight / class threshold T(n)C is used to obtain the passenger classification during the time between the first algorithm call and InitialClassDelay (i.e., the debounce time of the first classification). These are also used when the passenger classification is level 0. The high weight / class threshold T(n)H and low weight / class threshold T(n)L define the upper and lower limits of the deadband for transitions between passenger classes.
[0060]
[0062] During the operation of the passenger classification algorithm 20 (see Figure 3) implemented by the passenger classification system 40, the filtered total weight may increase and / or decrease. The hysteresis logic function 400 determines the passenger class when these changes occur. Deadbands defined by the high weight / class threshold T(n)H and the low weight / class threshold T(n)L prevent switching between classes when the weight is at the boundary between the two.
[0061]
[0063] As the Filtered Total Weight increases, the class will not transition to Class 1 until the T1H high weight / class threshold is reached. The class will not revert to Class 0 until the T1L low weight / class threshold is reached. Similarly, the class will not transition from Class 1 to Class 2 until the T2H high weight / class threshold is reached. The class will not transition from Class 2 to Class 1 until the T2L low weight / class threshold is reached. Finally, the class will not transition from Class 2 to Class 3 until the T3H high weight / class threshold is reached. The class will not transition from Class 3 to Class 2 until the T3L low weight / class threshold is reached.
[0062]
[0064] Referring to Figure 3, the class selected by the hysteresis logic function 400 is output as the InstantClass. The InstantClass is provided to the classification debounce function 450, which helps reduce chattering in the InstantClass that may occur due to rough roads or passenger tilt. The classification debounce function 450 prohibits classification changes if either the LongAccelHigh flag or the LatAccelHigh flag is set to YES (classification change suppression function - see Figure 5). The SelectFilter flag is used to determine the debounce time to reduce chattering. When a short filter is used, the debounce uses a shorter time to reduce chattering. When a long filter is used, the debounce uses a longer time to reduce chattering.
[0063]
[0065] The classification debounce function 450 provides the passenger class (OccupantClass) to the classification deactivation function 500 of the passenger classification algorithm 20 (see Figure 3), which outputs the final classification (FinalClass). The classification deactivation function forces the final classification (FinalClass) to be class 1 if the passenger classification (OccupantClass) is 0 and the passenger buckle state 28 indicates BUCKLED. <Note> [Form 1] A method for determining the passenger class of a vehicle seat, A step of obtaining a first seat weight display of the vehicle seat via a first seat weight sensor, wherein the first seat weight sensor is positioned on the side of the vehicle seat at a front position of the vehicle seat, and A step of obtaining a second seat weight indication of the vehicle seat via a second seat weight sensor, wherein the second seat weight sensor is positioned on the side of the vehicle seat at a rear position of the vehicle seat, and The steps include obtaining vehicle acceleration values via a vehicle acceleration sensor, The steps include determining the raw weight of the vehicle seat as twice the sum of the first and second seat weight indicators, The steps include determining the filtered weight based on the raw weight, A method comprising the step of determining the passenger class based on the filtered weight, depending on whether the vehicle acceleration value is less than a predetermined value. [Form 2] A method according to Embodiment 1, further comprising the step of suppressing a classification change in response to the vehicle acceleration value being greater than a predetermined threshold. [Form 3] A method according to Embodiment 2, wherein the step of suppressing classification changes includes a step of suppressing the classification due to a time delay. [Form 4] A method according to Embodiment 2, wherein the vehicle acceleration value includes vehicle lateral acceleration and / or vehicle longitudinal acceleration. [Form 5] The method according to Embodiment 1, wherein the passenger class is selected from one of the following classes: namely, the no-passenger class, the child seat class, the small adult class, and the large adult class. [Form 6] The method according to Embodiment 5, wherein the unoccupied class is associated with a measured seat weight up to a first weight, the child seat class is associated with a measured seat weight from the first weight up to a second weight greater than the first weight, the small adult class is associated with a measured seat weight from the second weight up to a third weight greater than the second weight, and the large adult class is associated with a measured seat weight of the third weight or greater. [Form 7] The method according to Embodiment 6, wherein the weight of the first is approximately 10.8 kg, the weight of the second is approximately 29.4 kg, and the weight of the third is approximately 54.8 kg. [Form 8] The method according to Embodiment 1, wherein the step of determining the filtered weight includes the step of selecting one of the unfiltered weight, the short-filtered weight, and the long-filtered weight, wherein the short-filtered weight is determined using a low-pass filter with a relatively short time constant, and the long-filtered weight is determined using a low-pass filter with a relatively long time constant. [Form 9] A method according to Embodiment 8, wherein the relatively short time constant is about 1 to 2 seconds, and the relatively long time constant is about 5 to 20 seconds. [Form 10] In the method described in Embodiment 8, the step of selecting one of the unfiltered weight, the short-filtered weight, and the long-filtered weight is: The steps include selecting the long-filtered weight in response to the determination that the vehicle is moving and the seat is occupied, A method comprising the step of selecting the short-filtered weight in response to a determination that the vehicle is not moving and the seats are not occupied. [Form 11] In the method described in Embodiment 8, the step of selecting one of the unfiltered weight, the short-filtered weight, and the long-filtered weight is: The steps include selecting the unfiltered weight during a predetermined startup time, After the aforementioned startup time has elapsed, the step of selecting the short-filtered weight, The steps include selecting the long-filtered weight in response to the determination that the vehicle is moving and the seat is occupied, A method comprising the step of selecting the short-filtered weight in response to a determination that the vehicle is not moving and the seats are not occupied. [Form 12] In the method described in Embodiment 8, the step of selecting one of the unfiltered weight, the short-filtered weight, and the long-filtered weight is: During the initial startup time, the steps include selecting the unfiltered weight, The steps include selecting the short-filtered weight while the vehicle is moving and the seat is occupied, A method comprising the step of selecting the long filtered weight until the vehicle comes to a stop. [Form 13] A method according to Embodiment 1, further comprising the step of invalidating the determined passenger class in response to determining that the buckle of a seat belt associated with the vehicle seat is released. [Form 14] A method according to Embodiment 1, wherein the step of determining the passenger class based on the filtered weight includes the step of assigning the passenger class based on the filtered weight and performing a hysteresis logic function to prevent changes in the assigned passenger class due to fluctuations in the filtered weight due to seat load in response to vehicle operation and / or changes in the position of the passenger on the seat. [Form 15] A method according to Embodiment 14, wherein the hysteresis logic function superimposes weight ranges for each passenger class, each weight range comprising a high threshold and a low threshold, the hysteresis logic function assigns the next highest passenger class in the event that the filtered weight exceeds the high threshold, and the hysteresis logic function assigns the next lowest passenger class in the event that the filtered weight falls below the low threshold. [Form 16] The method according to Embodiment 15, wherein each weight range further comprises a fitting threshold, the fitting threshold being the nominal weight value of the corresponding passenger class, and the fitting threshold is used to determine the initial passenger class. [Form 17] A method according to Embodiment 1, wherein the step of determining the passenger class based on the filtered weight includes the step of assigning the passenger class based on the filtered weight and performing a hysteresis logic function to prevent changes in the assigned passenger class due to fluctuations in the filtered weight due to seat load in response to vehicle operation and / or changes in the position of the passenger on the seat. [Form 18] A method according to Embodiment 1, wherein the step of determining the filtered weight based on the raw weight includes the step of performing a freeze filter function that freezes the filtered weight value in response to the vehicle lateral acceleration exceeding a predetermined threshold. [Form 19] In the method described in Embodiment 1, The first seat weight sensor measures the seat weight at the front inner mounting position of the seat, and the second seat weight sensor measures the seat weight at the rear inner mounting position of the seat, or, A method comprising: the first seat weight sensor measuring the seat weight at a front outer mounting position of the seat; and the second seat weight sensor measuring the seat weight at a rear outer mounting position of the seat. [Form 20] A passenger classification system for determining a passenger class associated with a vehicle seat, comprising a controller operable to implement a passenger classification algorithm according to the method described in Embodiment 1. [Form 21] A vehicle safety system comprising the passenger classification system according to Embodiment 20, at least one vehicle passenger safety device, and an airbag ECU operable to control the operation of the at least one vehicle passenger safety device, wherein the airbag ECU is operably connected to the controller of the passenger classification system and configured to control the operation of the at least one vehicle passenger safety device in accordance with the passenger classification determined via the passenger classification algorithm performed in the controller. [Explanation of Symbols]
[0064] 10. Vehicle Safety Systems 12 controllers 14 seats in the vehicle 16 rails 18 Safety Devices 20 Passenger Classification Algorithms 22 Filtered Vehicle Wheel Speed 24 Filtered Vehicle Longitudinal Acceleration 26 Filtered Vehicle Lateral Acceleration 28. Passenger buckle status 30. Final passenger classification 40 Passenger Classification System 50. Vehicle speed detection function 100 Classification change suppression function 150 Total weight calculation function 200 signal processing functions 202 High lateral acceleration flag 204 Total weight display 206 Vehicle movement 208 Unit Delay Function 210 First weight selector 212 Second weight selector 214 Filtered Gross Weight 250 Freeze Filter Functions 300 Two-stage weight filter function 302 weight output 304 Raw weight 306 Short Filter 308 Short Filtered Weight 310 Long Filter 312 Long Filtered Weight 350 Filter switch control function 400 Hysteresis Logical Functions 450 Classification Debounce Function 500 Classification Disable Function
Claims
1. A method for determining the passenger class of a vehicle seat, A step of obtaining a first seat weight indication of the vehicle seat via a first seat weight sensor, wherein the first seat weight sensor is positioned in front of the vehicle seat and located on one end of the vehicle seat in the width direction, A step of obtaining a second seat weight indication of the vehicle seat via a second seat weight sensor, wherein the second seat weight sensor is located at the rear position of the vehicle seat and on one end side in the width direction of the vehicle seat, The steps include obtaining vehicle acceleration values via a vehicle acceleration sensor, The steps include: calculating the raw weight of the vehicle seat as twice the sum of the first and second seat weight indications; A step of determining the filtered weight based on the raw weight, comprising the step of selecting one of the unfiltered weight, the short-filtered weight, and the long-filtered weight, wherein the short-filtered weight is determined using a low-pass filter with a relatively short time constant, and the long-filtered weight is determined using a low-pass filter with a relatively long time constant. The step of determining the passenger class based on the filtered weight, depending on whether the vehicle acceleration value is less than a predetermined value, The method further includes a step of suppressing the change in the passenger class in response to the vehicle acceleration value being greater than a predetermined threshold, A method for suppressing changes in passenger class, comprising determining that the vehicle acceleration value is greater than a predetermined threshold, and then suppressing changes in passenger class during a time delay.
2. A method for determining the passenger class of a vehicle seat, A step of obtaining a first seat weight indication of the vehicle seat via a first seat weight sensor, wherein the first seat weight sensor is positioned in front of the vehicle seat and located on one end of the vehicle seat in the width direction, A step of obtaining a second seat weight indication of the vehicle seat via a second seat weight sensor, wherein the second seat weight sensor is located at the rear position of the vehicle seat and on one end side in the width direction of the vehicle seat, The steps include obtaining vehicle acceleration values via a vehicle acceleration sensor, The steps include: calculating the raw weight of the vehicle seat as twice the sum of the first and second seat weight indications; A step of determining the filtered weight based on the raw weight, comprising the step of selecting one of the unfiltered weight, the short-filtered weight, and the long-filtered weight, wherein the short-filtered weight is determined using a low-pass filter with a relatively short time constant, and the long-filtered weight is determined using a low-pass filter with a relatively long time constant. The step of determining the passenger class based on the filtered weight, depending on whether the vehicle acceleration value is less than a predetermined value, A method further comprising the step of invalidating a determined passenger class in response to determining that the buckle of a seat belt associated with the vehicle seat is released.
3. The method according to claim 1, wherein the vehicle acceleration value includes vehicle lateral acceleration and / or vehicle longitudinal acceleration.
4. The method according to claim 1 or 2, wherein the passenger class is selected from one of the following classes: no passenger class, child seat class, small adult class, and large adult class.
5. The method according to claim 4, wherein the unoccupied class is associated with a measured seat weight up to a first weight, the child seat class is associated with a measured seat weight from the first weight up to a second weight greater than the first weight, the small adult class is associated with a measured seat weight from the second weight up to a third weight greater than the second weight, and the large adult class is associated with a measured seat weight of the third weight or greater.
6. The method according to claim 5, wherein the first weight is 10.8 kg, the second weight is 29.4 kg, and the third weight is 54.8 kg.
7. A method according to claim 1 or 2, wherein the relatively short time constant is 1 to 2 seconds, and the relatively long time constant is 5 to 20 seconds.
8. In the method according to claim 1 or 2, The first seat weight sensor measures the seat weight at a mounting position on the front right side of the seat, and the second seat weight sensor measures the seat weight at a mounting position on the rear right side of the seat, or, A method comprising: the first seat weight sensor measuring the seat weight at a mounting position on the front left side of the seat; and the second seat weight sensor measuring the seat weight at a mounting position on the rear left side of the seat.
9. A passenger classification system for determining a passenger class associated with a vehicle seat, comprising a controller operable to implement a passenger classification algorithm according to the method described in claim 1 or 2.
10. A vehicle safety system comprising the passenger classification system according to claim 9, at least one vehicle passenger safety device, and an airbag ECU operable to control the operation of the at least one vehicle passenger safety device, wherein the airbag ECU is operably connected to the controller of the passenger classification system and is configured to control the operation of the at least one vehicle passenger safety device in accordance with the passenger classification determined via the passenger classification algorithm performed in the controller.