Method and apparatus for measuring the relative and absolute instantaneous velocity vector of objects moving relative to each other by means of a measuring device permanently installed on one of the objects

Abstract

A device and its use for measuring the instantaneous velocity vector of objects (1), (3) and (4) moving relative to each other are described, wherein the device is permanently installed on the object (1), referred to as the measuring platform, and, for example, road traffic moves along with the flow of traffic. By means of the presented device several measuring barriers (7) send beams orthogonal to the movement of the measuring platform (1), each beam being designed as a sequence of temporally separated electromagnetic pulses. If the beams of the measuring barrier (7) are reflected by the stationary objects in the environment (29), the electromagnetic pulses propagated back are registered in the detector part of the distance measuring sensors (6) of the device and the lateral distance (11) of the objects to the direction of movement x of the measuring platform (1) is calculated from the transit time of the pulses. The lateral measuring range, defined by the measuring barriers (7), is divided into sensitive distance ranges A (8) and disjoint, non-sensitive distance ranges A (9) by the selection of the areas to be monitored. If an object (3) passes through the sensitive distance area A (8) and a sequence of valid distance measurements is taken in A (8), the computer evaluates the presence of an object in A (8) as, for example, a motor vehicle, and the time of the beginning and end of the passage through A (8) is determined using the current system time of the computer. From the known distances {x i | i ∈ {1,n}} of the distance measuring sensors {S i | i={1,n}} (6) among themselves and parallel to the direction of movement of the measuring platform (1) and from the associated measuring times t j of passing and the associated distance y i (t j) the relative velocity vector of the passing object (3), e.g. a motor vehicle, is calculated. The speed of the measuring platform (1) in the direction of movement x is calculated from the ongoing distance data to the lateral environment (29). The absolute velocity of the object (3) in the x-direction is calculated as the sum of the speed of the measuring platform (1) itself and the x-component of the relative velocity vector of the object (3) to the measuring platform (1). Continuously measured rotation rates and accelerations of the measuring platform (1) using an inertial measurement unit enable support for the calculation and correction of the velocities as well as real-time quality control of the measurement data.

Description

[0001] The measurement methods predominantly used today for the precise, external determination of the speed of moving objects, for example in public road traffic, are stationary in the sense that either the measuring device is installed at a fixed location to be monitored, or the monitoring location can be changed, but the measuring device itself remains stationary during the measurement.

[0002] Although this leads to a significant simplification of the measurement because the velocity vector of the measuring device itself can be set as the zero vector with respect to the earth-based reference system, the method loses the degree of freedom of the measuring device's own mobility during the measurement and thus a significant application potential.

[0003] The term speed is generally understood to mean the magnitude of the component of the speed vector in the direction of the roadway.

[0004] Especially in public road traffic, several methods are known for measuring speeds in moving traffic, also known as video tracking methods. A video tracking method is based on a speed measuring device carried on board the measuring vehicle. This means that the device is permanently installed in the moving measuring vehicle and the trajectories of both vehicles are assumed to be sufficiently similar. The instantaneous speed measurements from the measuring vehicle are transferred to the video data stream of the recorded traffic situation and stored.

[0005] What the latter methods have in common is that the measuring vehicle follows the vehicle to be measured in flowing traffic over a distance necessary for the measurement. After the measuring vehicle has adjusted its speed to that of the vehicle in front or behind it, a simultaneous distance and time measurement is started and the measuring vehicle follows the vehicle for several hundred meters. The distance is usually a measured section of road, or the distance traveled is measured using a calibrated roller tachometer function of the measuring vehicle. To achieve a permissible measurement, it is prescribed that the distance to the monitored vehicle remains constant within a required accuracy at the beginning and end of the measurement and that an identical trajectory is followed during the measurement.The average speed of the measuring vehicle determined at the end of the measurement can be used as a minimum speed for the following vehicle, which may result in further action.

[0006] All of these measurement methods have in common an inherent complexity of the measurement process.

[0007] When using a mobile measuring system in a stationary location, the measuring devices must be transported, set up, and aligned at a suitable location. The measurement locations must be selected and, if necessary, prepared for safe use.

[0008] With the video tracking system, it may be necessary to actively drive at high speed over a longer distance in order to maintain the required distance to the measured vehicle.

[0009] Based on the prior art, the present invention counteracts these disadvantages by generally having the measuring device on the measuring platform (1) move passively in the path plane (2) and only be passed by other objects (3), (4), Fig. 1.

[0010] The aim of the new method is: a) to continuously record the own velocity vector of the measuring platform (1) with measurement M, b) to detect the passing of the objects (3) or (4) in the measurement data from M, and c) to determine the relative, instantaneous velocity vector of the object (3) or (4) during the short time interval of passing the measuring platform (1) from the measurement data from M. The result of the velocity measurement is the mean value of several relative velocity vectors calculated from the measurement data M during the passing of the object (3) or (4) through the monitoring area (5) of the measuring platform (1).

[0011] The speed is thus derived from the distance-time relationship.

[0012] The monitoring area (5) is protected by measuring barriers (7) of several, laterally measuring, distance measuring sensors {S i | i ∈ {1,n}} (6). The distance measuring sensors S i (6) are identical transmitter and receiver devices that operate according to the time-of-flight principle of electromagnetic waves. The collimated and pulsed beams of the measuring barriers (7) of each distance measuring sensor \S i (6) run at a height of approximately 0.5 m parallel to the track plane (2) and each define a distance-sensitive, horizontal measuring barrier {Li | i ∈ {1,n} (7), Fig. 1 and Fig. 2.

[0013] The beams of the measuring barriers (7) of the distance measuring sensors S i (6) run parallel to each other in horizontal and vertical planes, Fig. 1 and Fig. 2.

[0014] The lateral distance between the distance measuring sensors {Si | i ∈ {1,n}} (6) is precisely measured and forms the basis {x i | i ∈ {1,n}} of the monitoring area (5), where x1= 0 and x n = B (10) is defined, Fig. 6. The width B (10) of the monitoring area (5) is typically one meter.

[0015] The design of the measuring arrangement is such that during operation at least two, generally several distance measuring sensors S i (6) measure the distance (11) to the lateral objects in the environment (29) at the specified measuring frequency, Fig. 1.

[0016] The lateral extent of the monitoring area (5) is divided into mutually disjoint distance areas A (8) and A (9), Fig. 1 and Fig. 2.

[0017] A (8) defines distances and thus ranges within which the monitored objects may move during the measurement. Ā (9) defines the remaining ranges. This allows algorithmic differentiation between the searched objects (3) or (4) and the remaining objects in the environment (29).

[0018] If there is no reflection or the distance measuring sensor (6) does not provide a valid measurement, the distance of this measurement is set to infinity.

[0019] In road traffic, areas A (8) are generally equated with the roadways. Areas A (9) are equated with either all areas where the objects to be detected cannot normally be located, or those potential areas A (8) that should be excluded from the current measurement M.

[0020] During the measurement M, the pulses of the measuring barriers (7) are continuously scattered and reflected either on the surfaces of the objects changing laterally to the direction of movement or no measurable reflection occurs.

[0021] From the measured runtime of the pulses, the respective distance measuring sensor S i (6) the distance y i calculated, and the digitized value is output to the arithmetic unit (25). In the arithmetic unit (25), the data is continuously buffered in a ring buffer in a "common-stop" mode for approximately 4 seconds, analyzed, and then overwritten. Fig. 6.

[0022] The current measured value of the distance measuring sensor S i (6) at time t represents the lateral position coordinate y i(t) between the x-axis of the coordinate system on the measuring platform (1) and the respective object at which the reflection takes place at time t. From the temporal sequence of the measurements and the movement of the measuring platform (1), a time series of the lateral distance profile y is obtained. i (t) (12) of the distance measuring sensor S i (6), Fig. 3.

[0023] In the ideal case of a rectilinear, unaccelerated movement of the measuring platform (1) relative to the stationary objects in the environment (29), the distance profiles y i (t), y j (t) are essentially identical on a coarser scale and by a time difference ΔT jı (t) shifted from each other. The time difference ΔT ji (t) is determined using the usual, well-known correlation methods. From the fixed distance (x i -x j ) of the distance sensors S i and S j (6) in the device and the time Δ T ji(t) can be the own speed VE x,ji (t) (23) in the x-direction as the quotient (x j -x i ) / ΔT jı (t). To support the calculation, the rotation and acceleration rates of the measuring vehicle are continuously measured with an inertial measuring unit (26) and processed in the arithmetic unit (25).

[0024] The sampling frequency of the pulse sequence of the distance measuring sensors (6), and thus the temporal quantization, is selected such that a specified accuracy of the resulting speed measurement is mathematically achieved. The frequency can range from one to several kilohertz.

[0025] The speeds of the measuring platform (1) and the passing object (3) or (4) are negligible compared to the propagation speed of the electromagnetic radiation of the distance measuring methods presented.

[0026] The distance profiles yi (t) (12) are evaluated in real time in the arithmetic unit (25) of the device and compared with preset parameters of the distance ranges A (8).

[0027] The presence of a searched object (3) or (4) is detected if in the distance profiles selected for triggering y i (t) (13) and y j (t) (14) of the distance measuring sensors S i and S j (6) a distance in the monitoring area A (8) is measured over a predetermined number of consecutive measurements in coincidence K (15), Fig. 4.

[0028] The trigger condition T can be continuously adjusted during operation in the arithmetic unit (25). For example, external influences can be taken into account, or different regions in the orbital plane (2) can be selected in different directions.

[0029] If the trigger condition T is fulfilled, the program branches after clearly defined jumps in the distance profiles y i (t) (13) and y j (t) (14) between the areas A (8) and A (9).

[0030] The Fig. 4 shows these jumps as an example as points in time tiH (18) and tjH (19) to which a specific part of the object, e.g. a profile edge H (28), the specified measuring limits L i and L j (7) happened.

[0031] In an analogous manner, the times before the trigger condition T can be determined in the buffered data tiV (20) and tjV (21) are searched for at which another profile edge V has passed the measuring barriers.

[0032] From the period ΔtjiH=tjH−tiH (16) or ΔtjiV=tjV−tiV (17) and the known horizontal distance (x i -x j ) of the associated distance measuring sensors S i and S j the relative velocity components VRx,jiH and VRx,jiV of the passing object (3) or (4) to the measuring platform (1) in the x-direction is calculated: VRx,jiH=(xj−xi) / ΔtjiH VRx,jiV=(xj−xi) / ΔtjiV

[0033] In an analogous manner, the lateral y-values ​​are determined at the times tjH,tiH,tjV,tiV evaluated. The result is the relative velocity components VRy,jiH and VRy,jiV of the passing object (3) or (4) to the measuring platform (1) in the y-direction is calculated: VRy,jiH=(yj(tjH)−yi(tiH)) / ΔtjiH VRy,jiV=(yj(tjV)−yi(tiV)) / ΔtjiV

[0034] The number of installed distance sensors defines the set of possible pairings for the difference calculation. For example, with four distance sensors S1, S2, S3, S4, there are six possible pairings (S1, S2), (S1, S3), (S1, S4), (S2, S3), (S2, S4), (S3, S4) for calculating 12 independent, relative velocities.

[0035] The overall result of the measurement is calculated by averaging.

[0036] In the example Fig. 5, the device consists of two distance measuring sensors S1 and S2, whose horizontal distance (x2-x1) simultaneously represents the entire width B (10) of the monitoring area (5).

[0037] The equations reduce to: VRx,21H=(x2−x1) / Δt21H=B / Δt21H VRx,21V=(x2−x1) / Δt21V=B / Δt21V VRx=(VRx,21H+VRx,21V) / 2 VRy,21H=(y2(t2H)−y1(t1H)) / Δt21H VRy,21V=(y2(t2V)−y1(t1V)) / Δt21V VRy=(VRy,21H+VRy,21V) / 2

[0038] In a typical case, e.g., in road traffic, the measuring platform (1) and the object (3) or (4) pass each other in the trajectory plane (2) during a short measurement period of Δt ≈ 0.2 s or less due to the compactness of the measuring device. During this short period, the trajectories can be considered to be almost straight and unaccelerated, so that the velocity vector VR determined in this way represents the instantaneous relative trajectory. The relative velocity component VR x corresponds to the commonly used term of relative vehicle speed.

[0039] In general, the absolute velocity vector VO of the passing object (3) or (4) is formed by the vector sum of VE and VR.

[0040] In the case of road traffic, the measuring platform (1) generally moves on the roadway, so that the absolute speed VO (24) is formed as the sum of the x-components of VE (23) and VR (22). VO=VRx+VEx

[0041] An inertial measurement unit (26) continuously measures the rotation and acceleration rates of the measuring platform as a unit. From this data, the correction of the velocity measurement is calculated based on the accelerations of the measuring platform, in particular the rotational movement during the measurement period Δt, and the value of VO (24) is corrected if necessary.

[0042] On the other hand, the measured values ​​from the inertial measurement unit (26) serve as quality control for the velocity measurements of VR and VE. Only data recorded during a measurement phase that does not exceed certain limits for acceleration and angular rates are accepted as usable measured values.

[0043] They show: Fig. 1: Schematic representation of the measurement situation in the orbital plane during the movement of the objects Fig. 2: Schematic representation of the path of the measuring beams over the orbit plane. Fig. 3: Time series of the distance profile of a distance measuring sensor. Fig. 4: Time-delayed distance profiles of two distance measuring sensors and their coincidence. Fig. 5: Schematic representation of the passing movement. Fig. 6: Schematic of the device with components.

Claims

[1] Method for detecting and discriminating a moving object when moving on a track plane (2), in particular on public roadways, and measuring the relative and absolute instantaneous speed vector of the object by means of a likewise moving measuring platform (1) with the following provisions: a. Providing a device installed on the measuring platform (1) and moving with it, which consists of several distance measuring sensors (6) which emit the beams of the measuring barriers (7) and register the reflected beams, digitize them and feed them to the computing unit (25). b. Providing a sensory inertial measuring unit (26) which moves along the measuring platform (1) and which registers the dynamic data on accelerations and rotation rates, digitizes them and feeds them to the arithmetic unit (25). c. Designing the arrangement of the distance measuring sensors (6) in such a way that the beams of the measuring barriers (7) form a spatially defined, lateral monitoring area (5). d. Designing the distance measuring sensors (6) in such a way that several beams of the measuring barriers (7) of independent distance measuring sensors (6) delimit the monitoring area (5) or define sub-areas therein. e. Designing the distance measuring sensors (6) in such a way that a time- and distance-resolved digitized measurement sequence generated with a sufficient temporal repetition frequency and sufficient spatial accuracy is fed to the arithmetic unit (25). f. Designing the measuring method in such a way that different measuring methods according to the LIDAR, LEDAR or RADAR principle are used in the distance measuring sensors (6) and are combined with imaging and distance-measuring 3D pixel detectors (27). g. Designing the measuring method in such a way that a continuous evaluation of the distance data takes place in real time in the arithmetic unit (25). h. Designing the calculation algorithm in such a way that a machine detection of an event E, namely the presence of an object (3) or (4) in the surveillance area (5) by means of a coincidence condition K in the distance profiles y i (t) (12) of several distance measuring sensors (6) with a very high correct acceptance rate and with a very low false acceptance rate in real time. i. Calculating the relative velocity vector VR of the object (3) or (4) from the combination of the data of different distance profiles y i (t) of the distance measuring sensors (6) in event E. j. Calculate the airspeed VE x of the measuring platform (1) in the direction of movement x in the computing unit (25) from the continuously measured distance profiles y i(t) to the objects in the surveillance area (5) by means of time correlation. [2] Method according to claim 1, characterized by that the measuring platform (1) can move at any permissible speed. [3] Method according to claim 1 and 2, characterized by that the measuring platform (1) can be at a standstill, ie with its own speed of 0 km / h. [4] Method according to claim 1 and 2, characterized by that the measuring platform (1) can move in any direction relative to the object (3) or (4) to be measured, but still at permissible speed. [5] Method according to claims 1, 2, 3 and 4 characterized by that the monitoring area (5) laterally covers several separate areas and thus several areas can be monitored simultaneously from the measuring platform (1). [6] Method according to claims 1, 2, 3, 4 and 5 characterized bythat the monitoring parameters for defining the condition of the event E in the arithmetic unit (25) are updated and adapted from the current data. [7] Method according to claims 1, 2, 3, 4, 5 and 6 characterized by that in the arithmetic unit (25) the real-time measurement data of the lateral distances to the stationary objects of the environment (29) are corrected in real time for the effects of the own acceleration and rotational movement of the measuring platform (1). [8] Method according to claims 1, 2, 3, 4, 5 and 6 characterized by that in the arithmetic unit (25) the real-time measurement data of the lateral distances to the moving object (3) or (4) are corrected in real time for the effects of the own acceleration and rotational movement of the measuring platform (1). [9] Method according to claims 1, 2, 3, 4 and 5 characterized by that the corrected relative velocity component VR x (22) to the value of the airspeed VE X(23) of the measuring platform and thus the absolute speed VO x (24) of the object to be measured (3) or (4) is determined in the earth-bound reference system in the orbital plane.

Images