JP4771657B2 - Proximity object detection system - Google Patents

Description

[0001]
Background of the Invention
In view of the dangers associated with driving a car, it is demanded today to raise the driver's awareness. One area where driver awareness can be raised is object detection around the vehicle. When a vehicle approaches an object (eg another car, pedestrian and obstacle), or when an object approaches a vehicle, the driver is always required to detect the object and avoid collision with this object Intervention is not always possible. For example, a vehicle driver may not be able to detect an object at the so-called “blind spot” of the vehicle.
[0002]
In order to raise the awareness of the situation of a truck, for example, a sensor system for detecting objects around the truck or a further simplified “sensor” has been proposed. Such sensors typically include optical detectors or infrared (IR) detectors that detect obstacles in the vehicle's path. In such an application, it is necessary to provide a sensor capable of detecting an object with high accuracy and reliability in a vehicle route.
[0003]
Radar is a suitable technique for realizing sensors used in vehicles such as automobiles and trucks. One type of radar suitable for this purpose is a frequency modulated continuous wave (FMCW) radar. In a typical FMCW radar, the frequency of the transmitted CW signal increases linearly from a first predetermined frequency to a second predetermined frequency. FMCW radar has the advantages of high sensitivity, relatively low transmitter power, and high range resolution.
[0004]
Characteristics that contribute to the accuracy and reliability of the sensor include sensitivity to noise and overall accuracy of processing received radio frequency signals (RF) to detect objects within the field of view of the sensor. Regarding sensitivity to noise, for example, the distance and position of an object may cause false detection, and a further danger is that it may be overlooked without detecting the object.
[0005]
A further important attribute of the sensor relates to its physical size and shape factor. Preferably, the sensor is housed in a relatively small enclosure or housing that can be mounted behind the surface of the vehicle. For accuracy and reliability, transmit and receive antennas and sensor circuits are not affected by vehicle attributes (eg vehicle grills, bumpers, etc.) and can be predicted and aligned to mount the sensor on the vehicle It is essential to do.
[0006]
Therefore, it would be desirable to provide a sensor system that can detect objects around the vehicle. It would also be desirable to be able to provide a system that can be adapted to provide detection zones around different sized vehicles. Furthermore, it would be desirable to be able to provide a system that can be reprogrammed at a remote location.
Summary of the Invention
In accordance with the present invention, a near object detection (NOD) system includes a plurality of radio frequency (RF) transmit / receive (TR) sensor modules (or simply “sensors” disposed around a vehicle (vehicle). ”), And one or more detection zones are deployed around the vehicle. In a preferred embodiment, when placing the sensors, each sensor detects an object in one or more coverage zones that substantially surround the vehicle. Of the plurality of sensors, the first sensor can be mounted on the rear and / or front bumper of the vehicle, while the second sensor can be mounted on the side panel of the vehicle. Each of the sensors includes a sensor antenna system, which includes a transmit antenna that emits or transmits an RF signal and one or more objects within the transmit RF signal within the field of view of the transmit antenna. It consists of a receiving antenna that receives the part that is blocked and reflected back. Alternatively, a monostatic antenna can be used. The transmit antenna can be composed of a planar array of antenna elements, while the receive antenna can be composed of a planar array of antenna elements or a single row of antenna elements. That is, the transmit and receive antennas can be provided with different numbers and different types of antenna elements. The NOD system further includes a receiving circuit coupled to the receiving antenna and receiving a signal from the receiving antenna to detect a path or path of one or more objects.
[0007]
With this specific configuration, an NOD system that detects an object in any region around the vehicle is provided. If one of the sensors determines that the vehicle is approaching the object or determines that the object is approaching the vehicle, the sensor executes according to a set of detection rules (rules) To start.
[0008]
In one embodiment, the aforementioned system is provided as a distributed processor system, where each of the sensors includes a processor. Each sensor is coupled to each other so that information can be shared between the sensors. In another embodiment, each of the sensors is coupled to a central sensor processor that receives information from each of the sensors and processes this information as appropriate.
Detailed Description of the Invention
The foregoing features of the invention, as well as the invention itself, may be better understood from the following description in conjunction with the drawings.
[0009]
Before describing the NOD system, some preliminary concepts and terminology are explained. As used herein, the term “sensor system” is located on a vehicle (transportation means) and can detect and indicate such an object, such as another vehicle or a stationary object. Refers to a system that has a corresponding output capable. The term “sensor” is also used herein to describe a sensor system. A sensor system or sensor receives data from various sensor systems and distinguishes it from a proximity object detection (NOD) system that combines and processes data from various sensor systems.
[0010]
Referring now to FIG. 1, a proximity object detection (NOD) system 10 is disposed on a vehicle (vehicle) 11. The vehicle 11 can be provided as, for example, an automobile such as an automobile, a motorcycle, or a truck, a marine transportation means such as a boat or an underwater transportation means, or an agricultural vehicle such as a mower. In this particular embodiment, the proximity object detection system 10 was assigned to the assignee of the present invention on July 27, 1999, entitled “Automotive Forward Looking Sensor Application”. Forward-looking sensor (FLS) system 12, which can be of the type described in issued US Pat. No. 5,929,802, electro-optics which can be an infrared (IR) sensor Electro-optic system (EOS) sensor 14, a co-pending application filed August 16, 2001 entitled "Radar Transmitter Circuitry and Techniques" assigned to the assignee of the present invention. A plurality of side-looking sensors (SLS) that can be of the type described in US patent application Ser. No. 09 / 931,636. ) Systems 12-22 (also referred to as side object detection (SOD) systems 16-22), and a plurality of rear-looking sensor (RLS) systems 24,26. Sensors 16-28 are co-pending US patents entitled "System and Technique for Mounting a Radar System on a Vehicle" assigned to the assignee of the present invention. Various techniques can be used to couple to the vehicle, including the techniques described in application 09 / 930,868. The system 10 can also include a stop and go (SNG) sensor 27. It should be understood that the processing performed by the stop and start sensor 27 and the detection zone provided by the sensor 27 can also be provided by the FLS 12, and therefore the sensor 27 can be omitted. When determining whether to obtain the stop and start processing function from the FLS 12 or a separate sensor (for example, the SNG sensor 27), a trade-off is necessary. As an example, trade-off considerations include the desired shortest and longest detection distance, tolerance around the zone and reaction time.
[0011]
The FLS, EOS, SLS, RLS and SNG (if included) systems 12-27 are each coupled to a bus 28 that provides a communication bus between each of the sensors 12-27. The bus 28 can be provided as a local area network (LAN) 28, for example. In some embodiments, it may be desirable to provide the LAN 28 as a wireless LAN.
[0012]
As used herein, “target route (tracking route) data”, “track (tracking route) data”, “target data” or equivalent “track (tracking route) information” refers to “path ( This is the data in the "tracking path) file" and describes the path of the target in a certain coordinate system. The object is also referred to herein as a “target” and includes another vehicle or a stationary object. The target route data is predicted by the target when the old target route data corresponding to the location where the target was, the new target route data corresponding to the location where the target is present when updating the current data, and the current and / or future target route data updating. Predicted target route data corresponding to the location to be performed can be included.
[0013]
It will be appreciated that the system 10 is a real time system and therefore information should be exchanged / transferred between each of the sensors 12-27 and the processor 30 as quickly as possible. That is, the bus 28 must be able to handle a relatively high data transfer rate.
[0014]
For example, it may be desirable for bus 28 to have an average bus bandwidth that takes into account additional protocol overhead at 157 kbits per second. This bandwidth is such that the transmit and receive antennas each have 7 antenna beams, each of the 7 antenna beams has an average of 2 target tracking paths, 14 Hz per tracking path (minimum), 100 bytes Calculations were made assuming that each track was reported (7 × 2 × 14 × 100 × 8 = 157 kbytes average bus bandwidth). In other words, the sensor can communicate through a conventional bus as currently possible on a vehicle (e.g., car area network (CAN)), but not more than the average bus bandwidth noted above. In some cases, it may be desirable to provide the bus 28 as a dedicated bus having at least such a bandwidth.
[0015]
Bus latency, as used herein, refers to the time difference between the detection of an object by a sensor and the detection report on the bus 28. The time delay introduced by bus latency should be relatively small, for example, on the order of a time delay corresponding to less than 0.5 meters of relative car movement. Relative movement, as used herein, refers to relative movement in units of distance either between an automobile and a stationary object or between a vehicle and a moving object, eg, another moving automobile. As used herein, relative speed is the speed of relative movement. The bus latency delay corresponding to 0.5 meters of relative vehicle movement can be determined by selecting the maximum relative vehicle speed, eg, 200 km / hr = 125 mph = 55.6 m / s. That is, dividing the car's relative travel of 0.5 meters by 55.6 m / s gives approximately 9 ms, which is the corresponding maximum bus latency time delay. Assuming the bus clock frequency is 33 kHz, 9 ms is equivalent to about 300 clock cycles. In summary, if the maximum relative vehicle speed is selected to be about 200 km / hr, a relative vehicle movement of 0.5 meters corresponds to about 9 ms. That is, it corresponds to about 300 clock cycles at a clock frequency of 300 KHz.
[0016]
While the previous description has described specific parameters including specific relative vehicle movement, specific selected maximum relative vehicle speed, and specific clock frequency, it will be appreciated that other parameters may be used in the present invention. I will be. However, the parameters describing the bus time latency must be selected according to various factors. The factors include, but are not limited to, the total system response time that allows the system to act on the vehicle by alarms, braking (braking), airbag activation, etc. It needs quickness. Other factors to consider can include fault tolerance, resistance to interference, reliability, and cost.
[0017]
The sensor is also coupled through bus 28 to a central tracker / data fusion (CT / DF) processor 30. The CT / DF processor 30 will be described below with reference to FIGS. 4, 6 and 7. It will suffice here to mention that the CT / DF processor 30 receives information provided from each of the sensors 12-27 and provides the information to each of the sensors 12-27. The sensors 12-27 use information provided by the CT / DF processor 30 to improve the overall performance of the system 10. This will eventually become clear.
[0018]
A human interface 32 is also coupled to the CT / DF processor 30 through the bus 28. The purpose of the interface 32 is to display or otherwise communicate the information collected by the sensors 12-28 to the driver of the vehicle 11 or other occupant (eg, by an audio signal or other signal). The interface 32 can also be provided as a head-up display, for example.
[0019]
In this particular embodiment, the CT / DF processor 30 is shown as a single processor provided as part of the sensor 16, whereas the FLS, EOS, SLS, through the bus 28 or other means. Each of RLS and SNG sensors 12-27 is coupled. Note that in another embodiment, one or more of the FLS, EOS, SLS, RLS and SNG sensors 12-27 include its own CT / DF processor, and directly communicate with another of the sensors 12-27 (eg, It will be appreciated that the processing necessary to share the transmission and reception information) may be performed. When it is desirable to provide redundancy in the CT / DF processing function, it may be desirable to provide the CT / DF processor 30 in two of the sensors 12 to 27. If each of the sensors 12-27 includes its own CT / DF system, the proximity object detection system can be provided as a distributed processing system. Factors to consider when choosing between a distributed and a single master processor include, but are not limited to, reliability, bus bandwidth, processing latency, and cost. Absent.
[0020]
In one embodiment, the CT / DF processor 30 provides specific information to a specific sensor among the sensors 12-27, and in another embodiment, the CT / DF processor 20 provides any information of the sensors 12-27. Provide to each.
[0021]
As shown in FIG. 1, at least one sensor 12-27 includes a central tracker data fusion (CT / DF) processor 30, each of the sensors 12-27 passing data to a CT / DF processor 30 via a bus 28. send. Regardless of whether the proximity object detection system includes a single or multiple CT / DF processors 30, the information collected by each of the sensors 12-27 is shared and the CT / DF processor 30 (or in the case of a distributed system). A plurality of processors) implement a decision tree or rule tree. For example, the CT / DF processor 30 may be coupled to one or more vehicle (vehicle) safety systems, such as a vehicle airbag system. In response to signals from one or more of the FLS, EOS, SLS, and RLS systems, the sensor processor may determine that it is suitable to “prepare” the vehicle airbag. Other examples include the activation of braking and steering systems, transmission controls, alarms, horns and / or automatic flashers.
[0022]
The NOD system 10 can therefore be coupled to a number of vehicle safety systems that are further described below. The CT / DF processor 30 receives any information provided and optimizes the processing power of the NODS system for the entire vehicle.
[0023]
The pair of RLS systems 24, 26 can detect objects at the rear of the vehicle using a triangulation method. The position (distance and direction) of the object may be determined from the distance or range (distance range) reading from each one of the pair of RLS systems 24, 26, and individually from each of the two sensors 24, 26. There is no need to elucidate the direction. Triangulation can be done by drawing two range circles. Each distance circle corresponds to a range (distance range) provided by each one of the pair of RLS systems 24, 26, and each distance circle has a radius equal to the range. The two distance circles thus given can intersect at two distances. Since one of the intersection distance points is located inside the host 11, it corresponds to an impossible distance. Select the other distance point and describe the position by distance and direction.
[0024]
To perform the triangulation described above, the distance between the sensors 24, 26 is known and large enough to allow a predetermined maximum triangulation error with respect to the distance measurement accuracy obtained by each of the sensors 24, 26. I must. It will be appreciated that the separation of the RLS systems 24, 26 may vary with various vehicle types, and the vehicle 11 is just one example, and some distance calibration is required. However, the calibration can be predetermined based on a known amount of separation.
[0025]
It will be appreciated that one or more of the sensors 12-27 may be detachably mounted on the vehicle 11. That is, in some embodiments, SLS, RLS, and FLS sensors may be located outside the vehicle body (i.e., located on the exposed surface of the vehicle body), while in other systems, one of sensors 12-27. One or more may be embedded in the bumper or other part of the vehicle (eg, door, panel, quarter panel, vehicle front end, and vehicle rear end). It is also possible to provide a system that is mounted inside the vehicle (for example, inside a bumper or other place) and that is detachable.
[0026]
Referring now to FIG. 2, like elements in FIG. 1 are given like reference numerals. Shown is a vehicle 11 in which a NOD system is located and is surrounded by a plurality of detection zones 32-40 forming a radar cocoon around the vehicle. It will be appreciated that if the size of the sensors 12-27 (FIG. 1) is different, the detection zones 32-40 are also different. That is, sensors 12 and 14 provide an adaptive cruise control and night vision zone 34, sensor 16 provides a lane keeping zone 36b, sensor 18 provides a road departure zone 36a, and sensor 20 , 22 are provided with lateral (side) object detection zones 38a, 38b, sensors 24, 26 are provided with a reverse and parking assist zone 40, and sensor 27 is provided with a stop and departure zone 42. In one example embodiment, the adaptive cruise control / night vision zone 34 has a limited angular range and is characterized by a long range (> 50 m) in order to operate at high relative speeds. The road departure and lane keeping zones 36a, 36b each have a short distance (short range), a wide angular range, and operate at a medium relative speed. The stop and start and reverse / park assist zones 42, 40 have a wide angular range but a very short distance (short range) and operate in a small relative speed range. The reverse / parking assistance zone 40 can also provide rear collision warning information in normal driving conditions. The lateral object detection zones 38a, 38b have a wide angular range and a relatively short distance and operate over a wide range of relative speeds.
[0027]
It will be appreciated that the size, shape and other characteristics of each sensor zone can be changed statically. The sensor zone can be statically modified to have a predetermined zone shape determined by detection characteristics and radar beam angles associated with sensors 12-27 (FIG. 1). There are many reasons for desiring to change the characteristics of one or more detection zones, including but not limited to the choice of the size of the vehicle or the operator's peripheral vision. Other reasons why it is desirable to change the detection zone size may also include trailer traction, lane size changes, traffic density and personal preference.
[0028]
The sensor zone can also be changed dynamically. Dynamic control can include, but is not limited to, a dwell on a radar beam. This will be described below in connection with FIG. Tracking handoffs can avoid or reduce capture verification steps (processes) and speed up sensor response to cue data or improve reliability. Dynamic changes are further described below in connection with FIG.
[0029]
By changing the characteristics of a single sensor, the sensor can obtain detection functions in coverage zones of different sizes and shapes, so that the sensor can be used in a vehicle larger or smaller than the vehicle 11. In this way, changing the coverage zone provided by an individual sensor can be done by programming the sensor.
[0030]
In one embodiment, the coverage zone is filed Aug. 16, 2001 and is a co-pending US patent entitled “Technique for Changing a Range Gate and Radar Coverage”. It can be changed by adjusting the distance gate of the sensor, as described in application 09 / 930,867. This application is assigned to the assignee of the present invention, the contents of which are hereby incorporated by reference. In another embodiment, a reconfigurable antenna is used to change the coverage zone. In yet another embodiment, a reconfigurable antenna is provided by using a microelectromechanical (MEM) device, which is used to change the beam shape and thus beam coverage. MEMS can change the shape of the aperture, and thus the beam shape.
[0031]
It should be noted that in the specific configuration of the sensor shown in FIG. 1, seven coverage zones 32-40 are provided. In one particular embodiment, a radar sensor system is utilized in each of the coverage zones. This radar sensor system is also referred to herein as a sensor and an RF sensor. Radar sensors can utilize a beamforming system that provides multiple transmit beams and multiple receive beams in each of the antenna and coverage zone. In this way, it is possible to find a direction in which another object or target approaches the vehicle or vice versa. In one particular embodiment, FLS sensor 12 (FIG. 1) may utilize an antenna system that includes eight separate antenna beams. The RF sensor system can operate as described in US Pat. No. 5,929,802, cited above. Similarly, sensors 16-27 can utilize an antenna system that includes seven separate transmit and receive antenna beams. Sensors 16-27 (FIG. 1) operate as described in previously cited US patent application Ser. No. 09 / 931,636 entitled “Radar Transmitter Circuit and Technique”. Can do. However, it will be appreciated that the radar sensor system that can be used with the present invention can have any number of transmit and receive beams.
[0032]
Next, referring to FIG. 3, the vehicle 11 in which the NOD system is arranged is traveling on a road 41 having three lanes 41 a, 41 b, 41 c. The vehicle 11 is in the lane 41 b and the first vehicle 50 is in front of the vehicle 11 and is visible in the detection zone 34. The second vehicle 52 is in the lane 41a on the right side of the vehicle 11 and is visible in the detection zone 38a. The third vehicle 54 is behind the vehicle 11 in the lane 41 b and is visible in the detection zone 40. The fourth vehicle 56 is in the left rear lane 41 c of the vehicle 11. Since the vehicle 56 is relatively far from the vehicle 11, the vehicle 56 is not visible in any detection zone and is therefore not detected by the NOD system located on the first vehicle 11.
[0033]
As shown in FIG. 3, the NOD system identifies three vehicles or targets 50, 52, 54 that are close to the vehicle 11. The NOD system maintains information for each target 50-54 and provides such information to the user (e.g., through the display 32 of FIG. 1) or a required function (e.g., a vehicle airbag system). Prepare).
[0034]
Furthermore, since the sensors 12 to 27 (FIG. 1) are in communication with the CT / DF processor 30 (FIG. 1) and in communication with each other, the sensors can share information about the target. For example, it is assumed that the sensor 18 mounted on the vehicle 11 detects the second vehicle 52 and starts tracking the second vehicle 52. After a certain period of time, the second vehicle 52 may start accelerating and overtake the vehicle 11. If the sensor 18 can detect that the second vehicle 52 overtakes the vehicle 11 from the right side, the sensor 18 can provide this information to the FLS 12. This information can be in the form of a public track file or a similar data set that indicates the target or second vehicle 52 in the global coordinate system of the vehicle 11. Such a path file allows the FLS 12 to have current target position information and target position prediction information and to actually observe / detect the target, ie the second vehicle 52.
[0035]
Thus, before the FLS 12 itself actually detects, captures, confirms, and tracks the target, the FLS 12 is given pre-route information about the identified target (ie, the “real” target). This pre-route information is also referred to herein as “queue data”. Queue data will be discussed further below in connection with FIG. Target detection, as used herein, refers to the process of distinguishing target signals that exceed an interference level based on a certain threshold. In this case, the target signal corresponds to backscattered RF energy from the target, and the interference corresponds to noise and / or clutter. Target acquisition, as used herein, refers to the process of associating (coordinating) a new target detection and associated target position with an existing target path corresponding to the “path file”. Target verification, as used herein, applies a set of rules, such as successive updates of the same tracking file or repeated target coordination for adjacent beams, to verify that the detected target is true. Say about the process. Target tracking, as used here, refers to the position and velocity of each update by associating the new target direction with the previous target direction and predicting and averaging the position state vector corresponding to the target position. Refers to the process of maintaining information about specific targets. These processes will be better understood when described in connection with FIG.
[0036]
By providing the FLS 12 (FIG. 1) with prior information, ie cue data (eg, information that the identified target is entering its field of view from the right side of the vehicle 11), the FLS 12 first detects the target. It is possible to proceed to the target tracking process without performing a target acquisition or target verification process, i.e., with a minimal amount of processing required to perform such a process. The FLS 12 checks the target and target tracking path with information from the sensor 18 (FIG. 1) without spending processing time to confirm that the vehicle 52 is actually a real target that is entering the field of view of the FLS 12. As such, the FLS 12 can perform more processing functions such as tracking multiple targets and other functions described below. In this way, by providing advance information to the FLS, the FLS 12 can track the target more quickly, and in particular, the FLS 12 can use a so-called “cut-in” target (ie, suddenly in front of the vehicle 11 lane 41b It is possible to detect and track the incoming vehicle) more quickly.
[0037]
Perhaps more importantly, it is advantageous for the FLS 12 to gain such prior knowledge by providing information about the path of the target 52 to the FLS 12 before the target 52 enters the FLS 12 detection zone. This is because the processing related to the operation of the defensive treatment can be started or executed in some cases. Defensive actions include, but are not limited to, air bag preparation, automatic adjustment of an automatic cruise control (ACC) system, and preparation of a braking (braking) system. In this way, the FLS 12 can also perform other functions related to driving the vehicle.
[0038]
It will be appreciated that the CT / FS processor 30 (FIG. 1) is a “target tracker” that performs the tracking function and also a “data fusion unit” that performs the fusion function. The central tracking function of the CT / DF processor 30 receives and maintains every tracking path from various sensors (eg, sensors 12-27 in FIG. 1) within the system 10 (FIG. 1), and another sensor. Is to assist the operation as described above.
[0039]
Referring now to FIGS. 4-4D, in FIG. 4, each of the radial sections 57a-57g corresponds to each of the seven beams provided by the sensor 18, and each of the radial sections 58a-58g is provided by the sensor 20. Corresponds to each of the seven given beams.
[0040]
FIGS. 4A, 4C, 4B, and 4D represent detections performed by the sensors 18 and 20 in a rectangular graph, with rows corresponding to the seven beams of each of the sensors 18 and 20, and columns corresponding to the range ( Corresponds to the (distance) cell. The dots in FIGS. 4A and 4B represent target detection in the seven beams of each sensor 18, 20. The dots in FIGS. 4C and 4D represent the fusion target detection associated with the seven beams of each sensor 18,20. Therefore, the dots 59a and 59b having the first parallel line pattern (cross hatch) correspond to detection of the target 52 in the beam of the sensor 18 corresponding to the radial sections 57a and 57b, respectively. The dots 59c and 59d having the second parallel line pattern correspond to detection of the target 54 in the beam of the sensor 18 corresponding to the radial sections 57f and 59g, respectively. The dots 60a-60c having the third parallel line pattern correspond to detection of the target 54 in the beam of the sensor 20 corresponding to the radial sections 58a-58c, respectively. The dots 60d and 60e having the fourth parallel line pattern correspond to detection of the target 56 in the beam of the sensor 20 corresponding to the radial sections 58f and 58g, respectively. The dots 61a and 61b having the first parallel line pattern correspond to fused detection of the target 52 in the beam of the sensor 18 corresponding to the radial sections 57a and 57b, respectively. The filled dots 61c and 61d correspond to the fusion detection of the target 54 in the beam of the sensor 18 corresponding to the radial sections 57f and 57g, respectively, and the filled dots 62a to 62c respectively correspond to the radial sections 58a to 58c. Corresponding to fusion detection of the target 54 in the beam of the corresponding sensor 20, the dots 62d and 62e having the third parallel line pattern respectively detect fusion of the target 56 in the beam of the sensor 20 corresponding to the radial sections 58f and 58g. Correspond.
[0041]
As previously described, FIGS. 4A and 4B correspond to unfused target data from sensors 18 and 20, respectively. 4C and 4D correspond to fusion target data from sensors 18 and 20, respectively, provided by the fusion function of CT / DF processor 30 (FIG. 1). Each of the seven rows shown in FIGS. 4A and 4C corresponds to a respective one of the seven beams associated with sensor 18. Similarly, each of the seven rows shown in FIGS. 4B and 4D corresponds to a respective one of the seven beams associated with the sensor 20. The fusion function corresponds to the cooperation of the individual target detection performed by the sensor 18 with that performed by the sensor 20. Therefore, the data corresponding to the dots 61c and 61d are linked or fused with the data corresponding to the dots 62a to 62c. The illustrated dots 61c-61d, 62a-62c are all filled, indicating that they are linked to the same target 54, and that the corresponding data from each of the two sensors 18, 20 are fused. Show. The dots 61a, 61b and 62d, 62e shown in the figure have the same parallel line pattern as the corresponding non-fusion dots 59a, 59b and 60d, 60e, respectively, and the target cooperation is obtained between the respective sensors 18, 20 by the fusion. Indicates no. 4C and 4D show the local (local) coordinate systems of the respective sensors 18 and 20, but the data corresponding to the dots 61a to 61d and 62a to 62e can be displayed also in the global coordinate system. It will be clear from the discussion. It should be noted that using such fused data, target detection and target tracking performed by two or more sensors are detections that have a higher accuracy probability than detection and target tracking performed by one sensor. It will be appreciated.
[0042]
In operation, many of the sensors 12-27 (FIG. 1) can track the same target. As shown in FIG. 4, for example, the target 54 appears in the field of view of the sensor 18 so that the sensor 18 can detect and track the target 54. Similarly, target 54 is detected and tracked by sensor 20 (FIG. 4). Thus, both sensors 18 and 20 can detect and track the target 54. The data provided by the sensors 18 and 20 corresponding to the target 54 can be merged. Such fused data can increase detection and tracking reliability over data from one of the sensors 18,20.
[0043]
Since the sensors 18 and 20 are located at different points on the vehicle 11, the sensors 18 and 20 track the target from two different aspect angles. Each of the sensors 18, 20 has its own unique local coordinate system. Because it has two different local coordinate systems, the sensors 18, 20 cannot determine that they are each tracking the same target. In order to adjust the detection and tracking data from each sensor 18, 20, each sensor 18, 20 provides its tracking information to the CT / DF processor 30 as a path file corresponding to the sensor 18, 20.
[0044]
The CT / DF processor 30 is provided with information specifying the physical position of each of the sensors 12 to 27 on the vehicle 11. Since the relative position of the sensor on a specific vehicle is fixed, the CT / DF processor 30 can convert the target route data given by each sensor in the respective local coordinate system into the global coordinate system of the vehicle. it can.
[0045]
In addition, the CT / DF processor 30 can supply target path data converted to the local coordinate system of any individual sensor. In this way, the CT / DF processor 30 can convert the path data in the local coordinate system given from each of the sensors 18 and 20 into its own global coordinate system. Thus, the CT / DF processor 30 observes the position of each target detected by each of the sensors 18 and 20 (more generally, any of the sensors 12 to 27) in a single coordinate system. It will be appreciated that the radiating segments 57a-57g correspond to the local coordinate system associated with the sensor 18, and the radiating segments 58a-58g correspond to the local coordinate system associated with the sensor 20.
[0046]
Since all target information can be seen in a single control system, the CT / DF processor 30 (FIG. 1) is more reliable than if each of a number of sensors is used to detect the target and determine the corresponding target path. Can be generated. The NOD system and the associated CT / DF processor 30 can fuse the data from each target tracking performed by each sensor (eg, sensors 18, 20) into a common file or (determined by tracking noise etc.) ) Only the highest quality data can be selected, assisting individual sensors and thereby increasing the processing power obtained.
[0047]
The process performed by the CT / DF processor 30 can include the fusion of target data provided by multiple sensors. Target data fusion may include transforming sensor target path data provided by a plurality of sensors in their respective local coordinate systems to a global coordinate system. This can be done by one or more coordinate transformations. The CT / DF processor 30 then links the track data provided by each sensor with the previously merged tracking track to obtain new merged track data.
[0048]
The process executed by the CT / DF processor 30 can also include “data linkage”, and when used here, position data having a “new” path or first assumed quality (prediction error statistic) 2 Refers to the process of comparing with existing route data having assumed quality. New tracking data that is likely to match (correlate with) a tracked path, i.e., when the position difference of the new track data is small when compared to the tracked track of an existing target, It is said to cooperate. Assume that this new position data was obtained from the same physical target as the tracking path. It is said that there is no association (cooperation) when new route data that is unlikely to match a certain tracking route, that is, when the position difference of the new data is large when compared with the tracking route.
[0049]
Further, the process performed by the CT / DF processor 30 may also include “recursive update” of the position path. In one such embodiment, the recursive update of the position path is performed by a Kalman filter. The Kalman filter is recognized as a filter that provides a position state vector that describes a target position, and can be applied to an existing target path in combination with new path data. The Kalman filter can reduce tracking errors by averaging the associated state vector data with each update. It will be appreciated that state vector filters other than Kalman filters can also be used in the present invention.
[0050]
The process executed by the CT / DF processor 30 can further include “tracking path start”, and when used here, a path file for new uncoordinated path data that is not linked to any of the existing path data. Say to create a new. Assume that uncoordinated route data corresponds to a new target that has never been tracked before. This process initializes any detection that does not work with an existing tracking path and creates a new path file representing that detection. Track new targets in subsequent data updates. Similarly, the CT / DF processor 30 can drop, or delete, a path that is out of the field of view of a particular sensor. Any target that has an existing path file that is not linked to new location data at the time of a data update is considered out of view, and the path file is deleted and not processed in future updates. The processing executed by the CT / DF processor 30 will be further described with reference to FIG.
[0051]
Referring now to FIG. 5, a radar system 66 includes an antenna section 67 having transmit and receive antennas 68, 69, a microwave section 70 having both a transmitter 72 and a receiver 74, and a digital signal processor (DSP). 80, a power supply 82, a control circuit 84, and a digital interface unit (DIU) 86. The transmitter 72 includes a digital ramp (slope) signal generator that generates a control signal to a voltage controlled oscillator (VCO). VCOs are described, for example, in co-pending US patent application Ser. No. 09 / 931,636, filed Aug. 16, 2001, entitled “Radar Transmitter Circuitry And Techniques”. Can be provided in the form. This application is assigned to the assignee of the present invention.
[0052]
The radar sensor 66 detects one or more objects, that is, targets, within the field of view of the sensor 66. In the exemplary embodiment, radar sensor 66 may be a proximity object detection system, such as NOD system 10 described above in connection with FIG. That is, the radar sensor 66 is suitable for use as a sensor such as one of the lateral (side) object detection (SOD) modules or sensors 16-27 described above in connection with FIG. As mentioned above, such sensors detect objects including, but not limited to, other vehicles, trees, signs, pedestrians, and other objects that can be detected in close proximity to the path in which the vehicle is located. For purposes, it is configured to be mounted on an automobile or other transport means 96. As will be apparent to those skilled in the art, radar sensor 66 is also suitable for use in many different types of applications. Such applications include, but are not limited to, marine applications, where radar system 60 can be mounted on a boat, ship or other vessel.
[0053]
In one example embodiment, the transmitter 72 operates as a frequency modulated continuous wave (FMCW) radar and the frequency of the transmitted signal increases linearly from a first predetermined frequency to a second predetermined frequency. FMCW radar has the advantages of high sensitivity, relatively low transmission power, and high distance resolution. However, it will be appreciated that other types of transmitters may be used.
[0054]
A control signal is supplied from the vehicle 96 to the radar system 60 through the control signal bus 92. The control signals can include a yaw rate signal corresponding to a yaw rate for the vehicle 96 and a speed signal corresponding to the speed of the vehicle. A digital signal processor (DSP) 80 processes these control signals and radar reflection signals received by the radar system 66 to detect objects in the field of view of the radar system 66.
[0055]
Radar system 66 further includes a CT / DF processor 88. This may be in the form of the CT / DF processor 30 described in FIG. The DSP 80 is coupled to a digital interface unit (DIU) 86 via a CT / DF processor 88. In another embodiment of the radar system 60, the CT / DF processor 88 may be omitted, in which case the DSP 80 is directly coupled to the digital interface unit 86. The format of the CT / DF processor 88 can be the format described above in connection with FIGS. 1-3 and described below. That is, the CT / DF processor 88 receives a signal from the DSP 80 and further receives information from the other radar system 66 arranged around the vehicle 96 via the DIU 86. The data provided to the CT / DF processor 88 by each of the radar sensors, eg, sensors 12-27 (FIG. 1), may be in the form of a path file or raw detection data in the sensor's coordinate system. The CT / DF processor 88 can also supply cue data (preceding data) to the sensor, in which case the cue data is obtained from detection of targets by each of the sensors 12 to 27 (FIG. 1). can get. The cue data can provide a target position that is not yet within the field of view of the sensor but is expected to move into the field of view.
[0056]
Radar system 66 provides one or more output signals characterizing objects within its field of view to vehicle 96 through output signal bus 94 to the vehicle. These output signals may include path data having a distance signal indicative of a distance associated with the target, a distance rate signal indicative of a distance rate associated with the target, and an orientation signal indicative of an orientation relative to the vehicle 96 relative to the target. Coupling the output signal to a control unit (not shown) of the vehicle 96 and further coupling this control unit to the safety system of the vehicle 96 for various uses, for example providing an intelligent cruise control system or collision avoidance system Can do.
[0057]
The antenna assembly 67 includes a receiving antenna 68 that receives an RF signal and a transmitting antenna 69 that transmits the RF signal. In this particular example, radar sensor 66 corresponds to a bistatic radar system. This is because it includes separate transmit and receive antennas located close to each other. Antennas 68 and 69 provide multiple beams to control the steering angle in parallel to aim the transmit and receive beams in the same direction. There are various suitable circuits for selecting the angle of each antenna 68, 69, including multiple position switches. A suitable antenna system is, for example, a co-pending US patent entitled “Switched Beam Antenna Architecture” filed on August 16, 2001 and assigned to the assignee of the present invention. It can be provided as the format described in application 09 / 932,574.
[0058]
Referring now to FIG. 6, an example application of the radar system 10 of FIG. 1 is shown in the form of an automotive proximity object detection (NOD) system 100. The NOD system 100 is disposed on the vehicle 120. The vehicle 120 can be provided as, for example, an automobile such as an automobile, motorcycle or truck, an ocean or underwater vehicle such as a boat, or an agricultural vehicle such as a mower. In this particular embodiment, the NOD system 100 includes a forward monitoring sensor (FLS) system 122, a photoelectric sensor (EOS) system 124 capable of obtaining image data, a plurality of side monitoring sensor (SLS) systems 128 or equivalent. A lateral object detection (SOD) system 128 and a plurality of rear monitoring sensor (RLS) systems 130. In the exemplary embodiment, radar system 10 of FIG. 1 is SOD system 128 and is shown in more detail in FIG.
[0059]
Each of the FLS, EOS, SLS, and RSL systems is coupled to a sensor processor 134. In this particular embodiment, sensor processor 134 is shown as a central processor, and each of the FLS, EOS, SLS, and RLS systems is coupled to it through a bus or other means. In another embodiment, one or more of the FLS, EOS, SLS, and RLS systems includes its own processor, such as the CT / DF processor 88 of FIG. 5, and performs the processing described below. It will be appreciated that this may be done. In this case, the NOD system 100 is provided as a distributed processor system.
[0060]
Regardless of whether the NOD system 100 includes one or many processors, the information collected by each of the sensor systems 122, 124, 128, 130 is shared and the sensor processor 134 (or multiple in the case of a distributed system). The processor) forms a decision tree or rule tree. The NOD system 100 can be used for a number of functions, including but not limited to blind spot detection, lane change detection, and vehicle air bag activation preparation, to perform a lane stop (sustain) function. For example, the sensor processor 134 can be coupled to the airbag system of the vehicle 132. In response to signals from one or more of the FLS, EOS, SLS, and RLS systems, the sensor processor may determine that it is suitable to “prepare” the vehicle airbag. Other examples are possible.
[0061]
The EOS system 124 includes an optical or IR sensor, or any other sensor that provides a relatively high resolution in the sensor's orientation plane. The pair of RLS systems 130 can detect an object at the rear of the vehicle using a triangulation method. An example of FLS system 122 is described in the aforementioned US Pat. No. 5,929,802. It will be appreciated that each of the SLS and RLS sensors may be provided with the same antenna system.
[0062]
Each of the sensors is arranged on the vehicle 120 such that a plurality of coverage zones exist around the vehicle. Thus, the vehicle is surrounded by a net or shroud like a sensor zone cage. In the specific configuration shown in FIG. 6, four coverage zones 68a-68d are provided. Each of the coverage zones 68a-68d utilizes one or more RF detection systems. The RF detection system utilizes an antenna system that provides multiple beams in each of the coverage zones 68a-68d. In this way, a specific direction in which another object approaches the vehicle or a direction in which the vehicle approaches another object can be found. One specific antenna that can be used is US patent application Ser. No. 09/931, filed Aug. 16, 2001, entitled “Slot Antenna Element for an Array Antenna”. 633, and US patent application Ser. No. 09 / 932,574 filed Aug. 16, 2001, entitled “Switched Beam Antenna Architecture”. Each of these is assigned to the assignee of the present invention.
[0063]
It will be appreciated that the SLR, RSL, and FLS systems can be removably deployed on the vehicle. That is, in some embodiments, the SLS, RLS, and FLS sensors can be located outside the vehicle body (ie, on the exposed surface of the vehicle body), while in other systems, the SLS, RLS, and FLS sensors are It may be embedded in bumpers and other vehicle parts (eg, doors, panels, quarter panels, the front end of the vehicle, and the rear end of the vehicle). It is also possible to provide a system that can be mounted inside a vehicle (for example, in a bumper or other place) and can be attached and detached. The mounting system is entitled “System and Technique for Mounting a Radar System on a Vehicle” filed on August 16, 2001 and assigned to the assignee of the present invention. It can be in the form as described in US patent application Ser. No. 09 / 930,868. The contents of these applications are incorporated herein by this reference.
[0064]
Referring now to FIG. 7, which may be provided by a CT / DF processor such as CT / DF processor 30 (FIG. 1), CT / DF processor 88 (FIG. 4), or sensor processor 134 (FIG. 5). An example of a set of elements 148 is shown and includes sensor measurement data at block 150. Sensor measurement data 150 includes imaging measurement data from infrared (IR) sensors and radar data from radar sensors provided by sensors such as sensors 12-27 described above in connection with FIG. The sensor data is given in each local coordinate system as described above in connection with FIG. Next, the sensor data is supplied to a multiple hypothesis tracker (MHT) 152 to perform data linkage of new tracking route data provided by the sensor 170 in each local coordinate system related to each tracking. . If a new target is detected but no path file exists for this new target, the MHT starts a new trace for each new target. The start of the tracking route and data linkage have been described above in connection with FIG.
[0065]
It is recognized that MHT 152 reduces the probability of false tracking path determination. The MHT 152 considers a number of possible guesses for data alignment of new track path data provided by various sensors, eg, sensors 12-27 (FIG. 1), based on a large number of measurements. The MHT 152 selects the cooperation with the highest probability among the predictions with the highest probability, that is, the cooperation with each existing tracking route of the new tracking route data.
[0066]
The relevance guess generator 154 generates guesses about data association, resolution, and data quality. In addition, the estimation of cooperation is performed individually. The purpose of the individual estimation process is to reduce the total number of estimations for computational efficiency. Individual estimation can include, but is not limited to, removal of low-probability guesses and a combination of correlated guesses.
[0067]
The tracking path data is received by either the Kalman filter 156 or a similar state prediction filter. Given the above, tracking path data may be existing, new, or future predicted tracking path data, and the output obtained by the Kalman filter works with each target tracking path. This is a state vector prediction that gives a target tracking route prediction that describes future tracking route data that is likely to be. State vector prediction can include various factors, including but not limited to target position and target velocity. The state vector prediction obtained by the Kalman filter 156 is then returned to the MHT 152 and used as one of many guesses. That is, it provides a guess related to the new track path data point obtained by the sensor. State vector prediction is used for both the filter averaging performed in the Kalman filter 156 and the linkage of the new tracking path data with the existing tracking path in the MHT 152, and increases the probability that target tracking is possible with each update. Because.
[0068]
The process then proceeds to the shared tracking path generator 160 to form a “public tracking path” or “shared tracking path file”. A shared tracking path refers to the tracking path data provided by any of the sensors 170, eg, sensors 12-27 (FIG. 1), for tracking tracking by the MHT 152, and for the tracking by the relevance guess generator 154. This is a tracking path generated for improvement. Forming a shared tracking route includes transforming the tracking route data provided in the local coordinate system by the relevance inference generator 154 and tracking the data in the vehicle's global coordinate system as described above. Data from the shared tracking path can ultimately provide information for sensor operation / resource scheduling performed by the sensor scheduler 158. The shared tracking path generator 160 performs target tracking associated with one or more sensors, eg, sensors 12-27 of FIG. 1, in the global coordinate system of the vehicle for each target.
[0069]
The shared tracking route data is given to the data fusion unit 162. The data fusion unit 162 fuses the shared tracking path by coordinating tracking path files provided by multiple sensors, eg, sensors 12-27 (FIG. 1), for current updates and from previous updates. Data fusion has been described in detail above in connection with FIGS. 4-4D.
[0070]
The fused shared tracking path file is then provided to the tracking path quality generator 164. Compare available shared track files to determine the best track track file based on factors including minimum track data distribution, track file creation time (age), and missed detection or federation history. However, it is not limited to these.
[0071]
The tracking path file supplied by the tracking path quality generator 164 is received by the determination unit 166. The discriminator 166 evaluates the road scene, i.e., all detected tracking paths, by interpreting the data output from the tracking path quality generator 164. The discriminator 166 identifies target sizes, identifies large and long targets such as trailers that generate a large number of tracking path files, identifies potential dangers such as blind spot detection, and sensors Perform processes including, but not limited to, determining whether sensor cueing is applicable. The queue data was explained earlier.
[0072]
Further, the determination unit 166 receives the cooperative shared tracking route sent from the shared tracking route generator 160, determines whether any change is necessary for radar scheduling, and sends scheduling information and queuing information to the sensor scheduler 158. provide. Scheduling information can include various factors, and if a target is detected in a radar beam and the target is considered to represent a very dangerous condition, the dwell in the radar beam Contains factors to be performed by the sensor. The scheduling information provided by the discriminator 166 can cause a sensor to process data from a radar beam when cue data from another sensor is linked to a particular radar beam. Information can also be included. The queuing information can adapt a particular sensor and direct the aiming of the radar beam primarily in the direction that the target is expected to appear by another sensor.
[0073]
The sensor scheduler 158 receives information from the discriminating unit 166, notifies the MHT 150 when the tracking path data should be updated, and notifies various sensors of the beam dwell to be performed. In addition, the sensor is notified of any appropriate queue data.
[0074]
The data tracking path from the tracking path quality generator 164 is received by the vehicle control collision management operator (operation unit) 168. Based on the road condition assessment performed by the discriminator 166 and the highest quality tracked path determined by the tracked path quality generator 164, the vehicle control collision management operating unit 168 is a safety coupled to the vehicle as described above. Various output functions associated with the system can be performed.
[0075]
The specific embodiment has been described above by exemplifying the fusion of data given by various sensors arranged on the vehicle. However, another embodiment of merging sensor data is also possible in the present invention. Needless to say. Other embodiments include, but are not limited to, filters other than the Kalman filter 156 and other sequences of blocks 150-170.
[0076]
While preferred embodiments of the present invention have been described above, it will now be apparent to those skilled in the art that other embodiments incorporating the concepts may be used. Thus, it is contemplated that these embodiments should not be limited to the disclosed embodiments, but only by the spirit and scope of the claims.
[0077]
All publications and documents cited herein are hereby expressly incorporated by reference in their entirety.
[Brief description of the drawings]
FIG. 1 is a block diagram of a proximity object detection (NOD) system located on a vehicle.
FIG. 2 is a diagram of a vehicle surrounded by a sensor zone cage provided by a NOD system of the type shown in FIG.
FIG. 3 is a diagram of a vehicle traveling along a road with another vehicle in the vicinity surrounded by a sensor zone fence provided by a NOD system of the type shown in FIG. 1;
FIG. 4 is a diagram of a vehicle when surrounded by a plurality of targets, where one target appears in the sensor zone of two different sensors.
FIG. 4A is a plot corresponding to radar reports in one local coordinate system of two different sensors.
FIG. 4B is a plot corresponding to radar reporting in the other local coordinate system of two different sensors.
FIG. 4C shows a plot corresponding to the fusion radar report from the sensor in FIG. 4A in a coordinate system corresponding to the local coordinate system of FIG. 4A.
FIG. 4D is a coordinate system corresponding to the local coordinate system of FIG. 4B and shows a plot corresponding to the fusion radar report from the sensor in FIG. 4B.
FIG. 5 is a block diagram of a proximity object detection (NOD) system having a central tracker / data fusion (CT / DF) processor.
FIG. 6 is a block diagram of a proximity object detection (NOD) system mounted on a vehicle having a single sensor processing system.
FIG. 7 is a block diagram of an example set of processing elements that a CT / DF processor can provide.

Claims (26)

A plurality of target sensors coupled to the vehicle, each providing range cell data in local coordinates associated with the respective target sensor;
A processor coupled to receive range cell data using the local coordinates from the plurality of target sensors and configured to process the range cell data in a global coordinate system of the vehicle; A processor configured to generate a control signal for controlling one or more vehicle safety systems, the processor comprising:
Coupled to receive range cell data from the plurality of target sensors to generate target path data and associated target paths from the range cell data, and for the target path of new target path data; A multiple guess tracker (MHT) configured to generate a linkage of
A relevance inference generator coupled to receive the target path from the MHT and configured to generate an inference for at least one of path data association, resolution, or target path quality;
A state variable filter coupled to the relevance inference generator and further coupled to the MHT, the state variable filter configured to perform target path prediction associated with the target path;
A shared path generator coupled to the relevance guess generator, the shared path generator configured to convert the local coordinates of the target path to the global coordinate system of the vehicle;
A data fusion device coupled to the shared path generator, the data fusion apparatus configured to combine target paths associated with the plurality of target sensors to provide a fused shared path;
A path quality generator coupled to the data fusion device, the path quality generator configured to determine a data quality value associated with the fused shared path;
A discriminator coupled to the shared path generator and the path quality generator and configured to perform target sensor scheduling;
A vehicle collision management actuator coupled to the path quality generator and the discriminator, the vehicle being configured to provide the control signal for controlling the one or more vehicle safety systems A collision management actuator ,
Proximity object detection system. The system of claim 1, wherein the processor is
A system comprising a combiner for combining a path file associated with each of the plurality of target sensors, the path file corresponding to a respective target position path of one or more targets.   The system of claim 1, wherein the state variable filter comprises a Kalman filter. The system of claim 1, further comprising:
A sensor scheduler coupled to the discriminator and the plurality of target sensors, the sensor scheduler providing an updated schedule associated with range cell data updates performed by the plurality of target sensors A range cell associated with a target detected by the selected target sensor at a higher rate than the other target sensors. A system that allows you to provide data.   5. The system of claim 4, wherein the sensor scheduler is further configured to provide a beam dwell associated with at least one of the plurality of target sensors, the beam dwell comprising the plurality of targets. A system in which each of the sensors determines one or more directional beams that concentrate processing time. A plurality of target sensors coupled to the vehicle and each configured to provide target data, the plurality of target sensors comprising at least one of an infrared (IR) sensor or a radar sensor; ,
A processor that receives the target data, processes the data, and provides control signals to control one or more vehicle safety systems;
A proximity object detection system comprising:
Combined to receive target data from the plurality of target sensors, generates target route data and associated target routes from the target data, and generates a linkage of new target route data with the target route. A multiple guess tracker (MHT) configured as follows:
A relevance inference generator coupled to receive the target path from the MHT and configured to generate an inference for at least one of path data association, resolution, or target path quality;
A state variable filter coupled to the relevance inference generator and further coupled to the MHT, the state variable filter configured to perform target path prediction associated with the target path;
A shared path generator coupled to the relevance guess generator, the shared path generator configured to convert local coordinates of a target path associated with the target to a global coordinate system of a vehicle;
A data fusion device coupled to the shared path generator, the data fusion apparatus configured to combine target paths associated with the plurality of target sensors to provide a fused shared path;
A path quality generator coupled to the data fusion device, the path quality generator configured to determine a data quality value associated with the fused shared path;
A discriminator coupled to the shared path generator and the path quality generator and configured to perform target sensor scheduling;
A vehicle collision management actuator coupled to the path quality generator and the discriminator, the vehicle being configured to provide the control signal for controlling the one or more vehicle safety systems A collision management actuator,
Proximity object detection system.   7. The system of claim 6, wherein the state variable filter includes a Kalman filter. The system of claim 6, further comprising:
A sensor scheduler coupled to the discriminator and the plurality of target sensors, the sensor scheduler providing an updated schedule associated with target data updates performed by the plurality of target sensors; And the selected one of the plurality of target sensors has a range cell data associated with the target detected by the selected target sensor at a higher rate than the other target sensors. A system that makes it possible to provide.   9. The system of claim 8, wherein the sensor scheduler is further configured to provide a beam dwell associated with the radar sensor, the beam dwell being processed by each of the plurality of target sensors. A system for determining one or more directional beams to concentrate time. A proximity object detection method,
Performing target tracking using a plurality of target sensors coupled to the vehicle, providing range cell data in local coordinates associated with each of the target sensors;
Sharing the range cell data provided by each of the plurality of target sensors in a processor coupled to receive and process the range cell data in a global coordinate system of a vehicle;
Generating a control signal to control one or more vehicle safety systems in response to the sharing;
The sharing of the target data is as follows:
Generating target path data and associated target paths from the range cell data using a multiple guess tracker (MHT) that receives range cell data provided by the plurality of target sensors; Generate a linkage of the target route data with the target route;
Examining a route guess using a relevance guess generator coupled to the MHT, wherein the route guess includes a guess for at least one of route data association, resolution, or target route quality;
Filtering using a state variable filter coupled to the relevance guess generator and further coupled to the MHT, the filtering includes generating a target path prediction associated with the target path;
Generating a shared path using a shared path generator coupled to the relevance guess generator, wherein the shared path generator converts the local coordinates of the target path associated with the target into a global coordinate system of the vehicle; Composed of
Fusing data using a data fusion device coupled to the shared path generator, wherein the data fusion apparatus combines target paths associated with the plurality of target sensors to provide a fused shared path. Configured,
Generating a route quality value using a route quality generator coupled to the data fusion device, wherein the route quality generator is configured to determine a data quality value associated with the fused shared route;
Making a determination using a classifier coupled to the shared path generator and the path quality generator, the classifier configured to perform target sensor scheduling;
Generating a control signal for controlling the one or more vehicle safety systems using a vehicle collision management actuator coupled to the path quality generator and the discriminator;
A method involving that. The method of claim 10, wherein the sharing of the target data is:
Combining a path file associated with each of the plurality of target sensors, wherein the path file corresponds to a respective target position path of one or more targets.   The method of claim 10, wherein the state variable filter comprises a Kalman filter. The method of claim 10, further comprising:
Scheduling the target sensor using a sensor scheduler coupled to the discriminator and the plurality of target sensors, the sensor scheduler updating range cell data performed by the plurality of target sensors A selected one of the plurality of target sensors is detected by the selected target sensor at a higher rate than the other target sensors. Enabling providing range cell data associated with the target. The method of claim 13, further comprising:
The sensor scheduler is used to generate a beam dwell associated with at least one of the plurality of target sensors, the beam dwell being one in which each of the plurality of target sensors concentrates processing time. Or a method of determining a plurality of directional beams. A proximity object detection method,
Target tracking is performed using a plurality of target sensors coupled to the vehicle, the target sensor configured to detect one or more targets within a predetermined coverage zone, and the plurality of targets The sensors are configured to generate respective target data, the target tracking comprising at least one of imaging by an infrared (IR) sensor or radar detection by a radar sensor;
The target data provided by each of the plurality of target sensors is shared in the processor to generate a control signal that controls one or more vehicle safety systems;
The sharing of the target data is as follows:
A multiple guess tracker (MHT) that receives target data provided by the plurality of target sensors is used to generate target path data and associated target paths from the target data and to generate new target path data Generate a linkage with the target route;
Examining a route guess using a relevance guess generator coupled to the MHT, wherein the route guess includes a guess for at least one of route data association, resolution, or target route quality;
Filtering using a state variable filter coupled to the relevance guess generator and further coupled to the MHT, the filtering includes generating a target path prediction associated with the target path;
Generating a shared path using a shared path generator coupled to the relevance guess generator, wherein the shared path generator converts the local coordinates of the target path associated with the target into a global coordinate system of the vehicle; Composed of
Data fusion is performed using a data fusion device coupled to the shared path generator, and the data fusion apparatus combines target data associated with each of the plurality of target sensors to provide a fused shared path. Configured to
Generating a route quality value using a route quality generator coupled to the data fusion device, wherein the route quality generator is configured to determine a data quality value associated with the fused shared route;
Performing discrimination using a classifier coupled to the shared path generator and the path quality generator, the classifier configured to perform sensor scheduling;
Generating a control signal for controlling the one or more vehicle safety systems using a vehicle collision management actuator coupled to the path quality generator and the discriminator;
A method involving that.   The method of claim 15, wherein the state variable filter comprises a Kalman filter. The method of claim 15, further comprising:
Scheduling the target sensor using a sensor scheduler coupled to the discriminator and the plurality of target sensors, wherein the sensor scheduler is associated with target data updates performed by the plurality of target sensors. A target configured to provide an updated schedule to be detected, wherein a selected one of the plurality of target sensors is detected by the selected target sensor at a higher rate than the other target sensors. Enabling to provide range cell data associated with the. The method of claim 17, further comprising:
The sensor scheduler is used to generate a beam dwell associated with at least one of the plurality of target sensors, the beam dwell being one in which each of the plurality of target sensors concentrates processing time. Or a method of determining a plurality of directional beams.   The system of claim 1, wherein the plurality of target sensors comprises a vehicle forward monitoring sensor and a vehicle side monitoring sensor.   2. The system of claim 1, wherein the processor moves the target sensor from a field of view of a first sensor of the plurality of target sensors to a field of view of a second sensor of the plurality of target sensors. A system configured to deliver detection of a target associated with the first sensor to the second sensor.   7. The system of claim 6, wherein the plurality of target sensors comprises a vehicle forward monitoring sensor and a vehicle side monitoring sensor.   7. The system of claim 6, wherein the processor moves the target sensor from a field of view of a first sensor of the plurality of target sensors to a field of view of a second sensor of the plurality of target sensors. A system configured to deliver detection of a target associated with the first sensor to the second sensor.   12. The method of claim 10, wherein the plurality of target sensors comprises a vehicle forward monitoring sensor and a vehicle side monitoring sensor.   11. The method of claim 10, further comprising: moving a target from a first sensor field of the plurality of target sensors to a second sensor field of the plurality of target sensors. Passing the detection of a target associated with one sensor to the second sensor.   16. The method of claim 15, wherein the plurality of target sensors comprises a vehicle forward monitoring sensor and a vehicle side monitoring sensor.   16. The method of claim 15, further comprising: moving a target from a first sensor field of the plurality of target sensors to a second sensor field of the plurality of target sensors. Passing the detection of a target associated with one sensor to the second sensor.

Images