EP1148460B1 - Warning system for protection of a temporary road construction site - Google Patents

Abstract

When a lane is blocked during temporary road works a blocking sign (2) is mounted on the rear of a vehicle. In addition a sensor (3), e.g. a laser sensor, is used to determine the approach of an oncoming vehicle (5). The vehicle acceleration and distance of approach are determined and fuzzy logic used to determine the probability that the vehicle will impact the road works. If necessary a warning (6) is sounded.

Description

The invention relates to a warning system for securing a temporarily arranged on a trunk road construction site according to the preamble of claim 1.

For temporary construction sites along major roads, the guidelines for "construction sites of shorter duration" apply. Thereafter, the lane occupied by the construction site of the multi-lane trunk road for the incoming traffic is secured by a construction vehicle with a shut-off board, which prompts the incoming traffic for lane change. Due to over-fatigue or lack of attention of motorists, despite effective visual warning signals, there are always collisions with the construction vehicle securing the construction site. Naturally driving up trucks are particularly dangerous. The accident statistics show that even speed limits established in the construction site area can not adequately mitigate this dangerous situation. Likewise, all attempts to present the site situation for the incoming traffic optically even more effective, although the risk of accidents reduced, but not yet reduced to a degree that would already be described as satisfactory.

From the published patent application DE 198 44 286 A1 is a warning system for jobs of shorter duration in the street. A detection unit, which is designed as a laser detector operating according to the pulsed laser measuring method, is integrated into a blocking panel for securing the workstation against inbound traffic. A sensor system in the form of a transmitting / receiving device continuously detects the approach speed of the vehicles as a function of the distance to the blocking board by means of a large number of individual measurements. In an evaluation unit it becomes the speed change calculated per unit of time and compared by means of evaluation algorithms, the associated brake distances and delay times at relevant thresholds. If the threshold is exceeded, an audible warning signal is emitted in the direction of the working area.

The invention is therefore an object of the invention to provide for a warning system of the type mentioned a further embodiment, which provides the opportunity to defuse the danger situation at a temporary construction site crucial, there to take into operation without further ado and especially one high operational safety, especially with regard to serious accidents.

This object is achieved in a warning system of the type mentioned by the features described in the characterizing part of claim 1.

All fuses for temporary road construction sites are based on passive solutions. All efforts are aimed at identifying the danger zone by means of optical signals, thus directing the attention of the approaching traffic to the particular traffic situation. However, the solution according to the invention is based on the consideration that all such measures probably can not prevent inattentive or fatigued motorists misjudge the dangerous situation or in the worst case, not realize at all. Rear-end collisions that endanger workers on construction sites can not therefore be ruled out. The invention therefore resolves itself from the consideration of making all efforts in particular only to warn the incoming traffic. Instead, it opens the way to additionally observe this incoming traffic, thereby detecting critical situations and then involving those involved, i. especially those working in the field of construction, to warn as early as possible. Even if a warning can only be given in the short term before an accident, then at risk persons are prepared and can try to reach safety.

For this purpose, individual vehicles that approach this construction site on a lane occupied by the construction site are detected. In short time intervals their distance and speed are measured and from this their acceleration or deceleration is determined. These measured variables are evaluated continuously in order to predict a particularly critical situation with some certainty in the form of a value of the collision probability.

Particularly advantageous is a further development of the invention, according to which the evaluation of these measured variables takes place in a multistage fuzzy decision logic, which in a first stage first assigns the respective measured values for speed, distance and acceleration of the vehicle located in the monitored section to ranges of values, e.g. "low", "medium", "high" assigns to a corresponding linguistic variable "speed", "distance" or "acceleration". In this fuzzy decision logic, affiliations for a linguistic output variable "danger" are determined in a second step in connection of these affiliations with the aid of a rule base that interprets the distance accordingly. In a third stage of the decision logic, the determined affiliations of the linguistic output variable "danger" are respectively converted into sharp output variables for a collision probability using an averaging function, the warning device being triggered at a specific value for the collision probability.

This development of the invention uses in a particularly advantageous manner the possibilities of fuzzy logic for the given application. Their characteristic of working with uncertainty relations allows a decision process regarding a possibly occurring event based on current data. Since the possibilities of fuzzy logic, their properties and their practical application to existing in a particular case criteria in traffic engineering are already well known as well as the technical implementation options are readily available the means available, the warning system according to the invention also with justifiable technical Effort to realize.

Another advantage is that fuzzy logic - as opposed to classical decision and control methods based on the evaluation of absolute values - is the result of the decision process Provides probability values that can form the basis for forming the warning system in multiple stages. Thus, for example, there is the possibility of first trying again in a first stage to attract the attention of the careless driver, who is probably not properly assessing the critical danger situation. Only then, if this driver still does not react and thus in all likelihood a collision with the vehicle shutting off the construction site seems unavoidable, then in a second stage, a warning signal is given to those working in the construction site area.

Furthermore, the properties of the fuzzy logic make it possible to adjust the warning system according to the invention with relatively simple means to the local environment conditions at a temporary construction site. The fuzzy logic makes it possible to adapt the membership functions of the linguistic variables "distance", "speed" and "acceleration" respectively to the geographical and also weather-related conditions at a construction site by simply changing parameters. The warning system according to the invention is therefore to be set up in a very flexible manner such that the decision process always proceeds based on the local conditions based on the measured data and thus leads to optimized results.

FIG 1

schematisch eine Baustellensituation, bei der die linke Randspur einer Fernverkehrsstraße durch eine Baustelle belegt ist, wobei von dem die Baustelle sichernden Baufahrzeug aus auf der belegten Fahrspur näher kommende Fahrzeuge überwacht sowie deren sich ändernder Abstand und Geschwindigkeit gemessen und anschließend ausgewertet wird,

FIG 2

ein Flussdiagramm einer Auswerteeinrichtung für diese Messdaten, die auf einer Fuzzy-Entscheidungslogik beruht,

FIG 3 - 5

jeweils eine schematische Darstellung der Zugehörigkeitsfunktionen für linguistische Variablen "Geschwindigkeit", "Abstand" bzw. "Beschleunigung" dieser Fuzzy-Entscheidungslogik und

FIG 6

als Ergebnis des Entscheidungsprozesses in einem Diagramm die Zugehörigkeitsfunktion für eine Ausgangsvariable "Auffahrwahrscheinlichkeit", ausgedrückt in prozentualen Werten.

An embodiment of the invention will be described below with reference to the drawing. It shows

FIG. 1

schematically a construction site situation in which the left edge of a highway intersection is occupied by a construction site, monitored by the construction site securing construction vehicle on the occupied lane coming closer vehicles and their changing distance and speed is measured and then evaluated,

FIG. 2

a flow chart of an evaluation device for this measurement data, which is based on a fuzzy decision logic,

3 - 5

each a schematic representation of the membership functions for linguistic variables "speed", "distance" or "acceleration" of this fuzzy decision logic and

FIG. 6

as a result of the decision process in a diagram, the membership function for an output variable "Auffahrwahrscheinlichkeit", expressed in percentage values.

FIG. 1 schematically shows a traffic situation which is typical for temporary construction sites, in particular on multi-lane highways. In the drawing, this construction site is indicated schematically by parked on the left lane construction vehicles 1 and 1 '. Seen in the direction of traffic, carries the first, set before the actual construction site construction vehicle 1 arranged on its rear wall a widely recognizable locking panel 2, which prompts the oncoming traffic for lane change in the adjacent inner lane. In particular, at short notice and also for a short time set up construction sites, so-called. Daily construction sites, this construction site to the rear completing construction vehicle 1 next to warning lights on the construction vehicles the most essential safeguard for the workers in the construction site.

Such construction sites of shorter duration, which are set up in the interest of traffic flow without elaborate hedging elements for traffic management, but are associated with a non-negligible accident risk. Accidents in which a collision of an approaching vehicle with the construction site backwards securing construction vehicle 1 occurs, are mostly due to fatigue or carelessness of the approaching motorists. Particularly serious consequences are caused by accidents caused by trucks. It Thus, the problem arises, without sacrificing the flexibility in the establishment of such construction sites of shorter duration, to take further security measures that reduce the risk of accidents or at least give the workers on site the opportunity to bring themselves in time before a collision in safety.

FIG. 1 schematically shows a security system as a possibility for the solution of this problem. It includes a detector device 3 with, for example, a laser detector as a sensor, which is mounted on or on the construction site before the oncoming traffic shut off construction vehicle 1. As indicated in FIG. 1 by a bundle of rays 4 emitted schematically in the direction of the incoming traffic, the vehicles approaching the closed-off traffic lane of the construction site, for example a lorry 5, are detected. At short intervals, for example after every 100 ms, the distance d and the speed v of the truck 5 approaching in the example on the locked lane are measured. In addition, the acceleration a or deceleration of the approaching vehicle can be determined from the measurement results of successive intervals.

The technology for such measurements is well known in the traffic engineering industry from the majority of applications. However, in the present case of application, specific framework conditions must certainly be taken into account. This includes that the measured variables, ie the velocity v and the distance d are subject to errors. This is all the more true for a calculated acceleration a, which is based on an evaluation of successive intervals. If the detector device 3 works with a laser beam as a scanning beam, the reflection properties of the laser beam are also to be considered. In general, the laser beam is reflected relatively poorly on the window surfaces of motor vehicles. But also colored surfaces have different reflection properties, bright areas for scanning are cheaper than dark ones. With regard to these reflectivity characteristics, practical experiments have shown that in general, "flat" front lorries are detected at a lower error rate than passenger cars. It is essential that the laser beam hits an approaching vehicle frontally and not laterally, motion components transverse to the detection direction are not detected. Finally, it should be added that the detector device must be statically rigidly fixed, so that erroneous measurements are not carried out by wind or other shocks. The above-mentioned framework conditions prove that it is expedient to check all measurement results with regard to their plausibility in order to deduce no wrong actions from them.

If one now wants to build a safety system to warn of a probable collision of a construction site approaching vehicle on the basis of such distance and speed measurements, these measurements should be made at a sufficiently large distance from the construction site securing construction vehicle 1, which is several 100 m , For example, in a range of 250 to 300 m. Very significant for the present application of such speed or distance measurements are for the derived action further criteria. For example, from the fact that a vehicle is approaching the construction site at considerable speed v in the closed lane, a high probability of collision can not yet be inferred. This does not apply even if the vehicle in question also accelerates. In many cases, such a driving behavior is present when the driver completes an overtaking process in preparation for a lane change. The situation is different when a vehicle on the closed lane with a relatively low, almost constant speed approaches. This situation may seem uncritical at first glance, but it could also affect this driving style underlying that the respective driver has not yet realized the construction site in front of him as a danger area. These examples may prove that classical decision-making or control methods based on absolute values of the measured quantities do not necessarily lead to a reliable decision result.

FIG. 2 schematically shows in a flow chart an evaluation and decision procedure which is better adapted to the various parameters of the traffic situation to be assessed than classical decision-making or control methods based on digital statements. This diagram shows as input quantities 7 the speed and distance values v and d measured at predetermined intervals of, for example, 100 ms. In a calculation step 8, the corresponding acceleration a is determined from the velocity values vi and v _i-1 for two consecutive intervals i-1 or i according to the following relationship: $$\mathrm{a}=\{\begin{array}{ll}\frac{{\mathrm{v}}_{\mathrm{i}}-{\mathrm{v}}_{\mathrm{i}-1}}{\mathrm{\Delta }\mathrm{t}}\hfill & \mathrm{For}\mathrm{}{\mathrm{v}}_{\mathrm{i}}\cdot {\mathrm{v}}_{\mathrm{i}-1}>0\hfill \\ 0\hfill & \mathrm{For}{\mathrm{v}}_{\mathrm{i}}\le 0\hfill \\ 0\hfill & \mathrm{For}\mathrm{}{\mathrm{v}}_{\mathrm{i}-1}\le 0\hfill \end{array}$$

$${\tilde{\mathrm{a}}}_{\mathrm{F}\mathrm{z}}\left(t_i\right)={\beta }_a\cdot {\mathrm{a}}_{\mathrm{F}\mathrm{z}}\left({\mathrm{t}}_{\mathrm{i}}\right)+\left(1-{\beta }_a\right)\cdot {\tilde{\mathrm{a}}}_{\mathrm{F}\mathrm{z}}\left(t_{i-1}\right)\mathrm{\hspace{0.17em}\hspace{1em}}0\le {\beta }_{\mathrm{a}}\le 1$$

$${\tilde{\mathrm{d}}}_{\mathrm{F}\mathrm{z}}\left({\mathrm{t}}_{\mathrm{i}}\right)={\beta }_d\cdot {\mathrm{d}}_{\mathrm{F}\mathrm{z}}\left({\mathrm{t}}_{\mathrm{i}}\right)+\left(1-{\beta }_d\right)\cdot {\tilde{\mathrm{d}}}_{\mathrm{F}\mathrm{z}}\left(t_{i-1}\right)\mathrm{\hspace{1em}}\mathrm{\hspace{0.17em}}0\le {\beta }_d\le 1$$

In the next step, a control step 9, the thus determined measurement results for the current speed v, the instantaneous distance d or the current acceleration a of the approaching vehicle are compared with the corresponding measured values of the already existing measurement series at the past time intervals in order to place them on their Plausibility check. In many cases, it might be useful to smooth out quite plausible measured values even if these are too noisy. Among other things, noise components may be due to the different reflection properties of the scanning laser beam set forth above. This smoothing is performed in a computing process 10. In the process, the smoothed values which compensate for the noise component become the speed v, the acceleration a and the distance d of a single vehicle are determined, for example, on the basis of the following relationships: $${\tilde{\mathrm{v}}}_{\mathrm{F}\mathrm{z}}\left(t_i\right)={\beta }_{\mathrm{v}}\cdot {\mathrm{v}}_{\mathrm{F}\mathrm{z}}\left({\mathrm{t}}_{\mathrm{i}}\right)+\left(1-{\beta }_{\mathrm{v}}\right)\cdot {\tilde{\mathrm{v}}}_{\mathrm{F}\mathrm{z}}\left(t_{i-1}\right)\mathrm{}0\le {\beta }_{\mathrm{v}}\le 1$$

$${\tilde{\mathrm{a}}}_{\mathrm{F}\mathrm{z}}\left(t_i\right)={\beta }_a\cdot {\mathrm{a}}_{\mathrm{F}\mathrm{z}}\left({\mathrm{t}}_{\mathrm{i}}\right)+\left(1-{\beta }_a\right)\cdot {\tilde{\mathrm{a}}}_{\mathrm{F}\mathrm{z}}\left(t_{i-1}\right)\mathrm{}0\le {\beta }_{\mathrm{a}}\le 1$$

$${\tilde{\mathrm{d}}}_{\mathrm{F}\mathrm{z}}\left({\mathrm{t}}_{\mathrm{i}}\right)={\beta }_d\cdot {\mathrm{d}}_{\mathrm{F}\mathrm{z}}\left({\mathrm{t}}_{\mathrm{i}}\right)+\left(1-{\beta }_d\right)\cdot {\tilde{\mathrm{d}}}_{\mathrm{F}\mathrm{z}}\left(t_{i-1}\right)\mathrm{}\mathrm{}0\le {\beta }_d\le 1$$

The subsequent process 11 in the flowchart of FIG. 2 now shows schematically that an actual linking of the measurement results processed so far to a decision criterion takes place with the aid of a fuzzy logic. In contrast to a classical decision or control method, the fuzzy logic does not immediately evaluate its supplied quantified input variables in the decision process as absolute values. Rather, the fuzzy logic is based on describing a state not quantitatively, but qualitatively with so-called linguistic variables, by assigning it to specific membership domains. By a plurality of applications of fuzzy logic in the field of traffic engineering it has to be proven that the expert in this field is quite familiar with dealing with fuzzy logic. It is therefore not necessary here to explain fuzzy logic as such in detail.

For the necessary understanding, it is sufficient to explain how the different measured values for speed v, distance d or acceleration a of a vehicle approaching the construction site are to be linked with one another. In this case, it must be carried out that differently significant statements can be assigned to the different measured variables, each of which is included in an overall statement with individual weighting. As a result, the fuzzy logic in this case provides a probability statement about the presence of a danger of collision emanating from a vehicle approaching the construction site. Because of such a probability statement it is also possible to provide graduated warnings in their urgency.

In a first step of the fuzzyfication, all decision variables, in this case with respect to the velocity v, of the distance d or of the acceleration a, are converted into probable statements by means of a membership function. In other words, a "sharp", i. absolute value of one of the input variables assigned to the value ranges of the corresponding linguistic variables.

For the speed v, the membership function for this application is shown in FIG. The mode of representation is essentially self-explanatory for the person familiar with fuzzy logic, so that no detailed explanation is required here. There are three variables "low", "medium" and "high" for speed v. After that, a determined speed of the detected motor vehicle of, for example, 10 km / h is very likely to be classified as low, whereas a speed v of 50 km / h is just as likely to classify it as low as medium speed. These overlaps or blurring in the assignment of individual measured values to corresponding variables are typical for the system concept of fuzzy logic.

Similarly, the linguistic variable for the distance d and the acceleration a is defined in FIGS. 4 and 5, respectively.

After this assignment of the individual measured values for the speed v, the distance d or the acceleration a, the correlations determined in this way are linked to one another and, in so doing, affiliations for the linguistic output variable "danger" are determined. In the case of application, it is expedient, the affiliations for a To interpret linguistic output variable "danger" in different ways depending on the distance d of the incoming vehicle. This is defined in the following tables for the corresponding affiliations of the linguistic variable "distance".

In a further step, the memberships of the output variable "Danger" resulting from the rule base shown in Tables 1 to 4 are converted by means of an averaging function into a sharp output variable "Auffahrwahrscheinlichkeit". The principle used in this case, known per se, is referred to as max-min inference. shown is this conversion of the membership of the output variable "Danger" into the sharp, expressed in percent output variable "Auffahrwahrscheinlichkeit" schematically in Figure 6.

Although the approach probability determined in this way is interpreted as a sharp initial value in the sense of fuzzy logic, in absolute terms a single value does not always have to be correct, i. be faultless. If a warning were always output due to a single, ascertained high collision probability of an approaching vehicle, this could more frequently lead to false alarms. It would be particularly critical if, as a result, the workers at the construction site became accustomed to these false alarms and therefore could no longer assign a single warning to a truly critical situation. It therefore seems expedient, at least not to draw the attention of at least the workers in the construction site area, in particular by means of an acoustic warning signal due to a single value.

In order to eliminate or compensate for individual, erroneous statements regarding the percentage collision probability, it would be possible to form average values of the last n measurements or of the last n results, for example, where n ≥ 3 is selected. Another possibility would be to issue a warning only when a defined threshold regarding the percentage drift probability has been exceeded in at least n consecutive intervals. In both cases, this would result in a systematic delay, although a time delay for the warning to be issued, in view of the reliability of the warning system and thus its acceptance by those affected would be quite acceptable.

Finally, the statement of a collision probability in percentage values also allows the possibility of providing different warning levels. So could for example at a Auffahrwahrscheinlichkeit of 50% first, the driver of the motor vehicle, which approaches the site in a critical manner, be visually and / or acoustically warned, while a warning for the construction site area itself is issued only when the Auffahrwahrscheinlichkeit has reached a critical value of 80% ,

In FIG. 2, the above-described evaluations of the results for the collision probability are represented functionally schematically in the test step 12, followed by the alarm 13, if critical values for the collision probability, as described above, are exceeded.

So far, it has not been taken into account that weather and road conditions have a decisive influence on whether or not the established manner of approaching a motor vehicle to the construction site area is in fact critical. Essential ambient conditions are wet or sloping roads, as well as static speed restrictions at the construction site. These parameters influence the vehicle dynamics and the possible behavior of the driver in a decisive manner. In order to ensure a corresponding operational safety of the warning system, therefore, the operator must be given the opportunity to enter the respective operating conditions in the system, so that the decision logic takes into account these environmental conditions in the determination of the membership functions. In order to achieve reliable operation or commissioning of the warning system, it is expedient to predefine standardized parameter sets for the ambient conditions, which allow the operator to make the simplest and clearest possible selection. An example of how such parameter sets could be determined is shown in Table 5 below. Table 5 Operating parameter sets # Road condition road slope permissible km / h dry wet gradient level pitch 1 + + 2 + + 3 + + 4 + + 5 + + 6 + + 7 + + ≤ 80 8th + + ≤ 80 9 + + ≤ 80 10 + + ≤ 80 11 + + ≤ 80 12 + + ≤ 80

The table thus describes twelve different operating situations, from which the operator selects the appropriate one and enters it into the warning system by means of the corresponding number of the parameter. The variants thus given are probably sufficient to adapt the decision logic to the respective situation at the place of use, that is to say the membership functions of the described linguistic variables "speed", "distance" or "acceleration".

In an advantageous embodiment of the warning system according to the invention, a monitoring device is further provided for recording images of the predetermined section lying in front of the blocking panel, wherein the images recorded within a period of time are stored together with the measured values of the detector device recorded at the same time triggered by a predetermined event. This makes it possible to use the personal warning system additionally for accident monitoring. In the absence of the specified event, for example discarded by a video camera as well as the measured values recorded by the detector unit for the speed and the distance from the obstruction panel of an approaching vehicle after a predetermined period of time and replaced by currently recorded data. Upon occurrence of the predetermined event, such as an accident, which is detectable for example on the noise caused thereby, the recorded images and measured values of a lasting until the time of the predetermined event period are permanently stored. This could be checked after an accident, if the warning system has worked properly.

Claims (11)

1. Warning system for protecting roadworks which have been set up temporarily on a long-distance road, which roadworks occupy at least one lane of the long-distance road, wherein the closure of the occupied lane to the approaching traffic is indicated by means of a lane-closed sign, wherein in addition a visually active detector device which is spatially assigned to the lane-closed sign and has a transceiver unit is provided, said detector device scanning a predetermined section of road ahead of the lane-closed sign in the opposite direction to that of the traffic on the affected lane, measuring at predefined intervals (Δt) the distance (d) and the velocity (v) of a vehicle (5) located on it, wherein the detector device is assigned an evaluation device by means of which an acceleration (a) of the vehicle is determined at successive intervals from a change in the measured velocity, and which also determines, from a logic combination of the measured values, a probability value that the approaching vehicle will impact with the roadworks, and that a warning device is provided which is triggered by a critical value for the probability of a collision and outputs an alarm signal to the vicinity of the roadworks and/or to the approaching vehicle,
   characterized in that
   the evaluation device contains a multistage fuzzy decision logic which, in a first stage, firstly assigns memberships to value ranges, for example "low", "medium", "high" of a corresponding linguistic variable "velocity", "distance" or "acceleration", to the respective measured values for velocity (v), distance (d) and acceleration (a) of the vehicle (5) located on the monitored section of road, in a second stage said fuzzy decision logic then determines memberships for a linguistic output variable "danger" by combining these memberships using a rule base which correspondingly interprets the distance (d), and in a third stage said fuzzy decision logic converts the determined memberships of the linguistic output variables "danger" in each case into sharp output variables for a probability of a collision by applying an averaging function, wherein the warning device is triggered at a predetermined value for the probability of a collision.
2. Warning system according to Claim 1,
   characterized in that, in the rule base, on the one hand value ranges "medium", "very high" or also "very high" for the linguistic output variable "danger" are assigned to the memberships "low", "medium" and "high" of the linguistic variables "velocity" when the distances (d) are very small or small and there is no acceleration (a), and, on the other hand, when the distance ranges are the same but there is acceleration the value ranges "low", "medium" or "high" for the linguistic output variable "danger" are assigned to the aforesaid memberships of the linguistic variables "velocity".
3. Warning system according to Claim 2,
   characterized in that, in the rule base, on the one hand the value ranges "medium", and also "medium" or "high" for the linguistic output variable "danger" are additionally assigned to the memberships "low", "medium" and "high" of the linguistic variables "velocity" when the distances (d) are medium-sized and there is no acceleration (a), and, on the other hand, when the distance range is the same but there is acceleration the value ranges "low", "medium" or "high" for the linguistic output variable "danger" are assigned to the aforesaid memberships of the linguistic variables "velocity".
4. Warning system according to Claim 3,
   characterized in that, furthermore, in the rule base, on the one hand the value ranges "low", "medium" or "high" for the linguistic variable "danger" are assigned to the memberships "low", "medium" or "high" of the linguistic variables "velocity" when the distances (d) are large and there is no acceleration (a), and, on the other hand, the value ranges "low", also "low" or "medium" for the linguistic output variable "danger" are also assigned to the aforesaid memberships of the linguistic variables "velocity" when the distances are large but there is acceleration.
5. Warning system according to one of Claims 1 to 4,
   characterized in that in order to adapt it to local operating conditions, predetermined parameter sets for setting the decision logic of the evaluation device are provided, by means of which parameter sets in particular a state of the underlying surface, a longitudinal inclination of the underlying surface and/or a velocity limit in the vicinity of the roadworks are defined.
6. Warning system according to one of Claims 1 to 5,
   characterized by a plausibility check of the measured values for the distance (d) and the velocity (v) as well as the values determined therefrom for the acceleration (a) before their further processing.
7. Warning system according to Claim 6,
   characterized in that in order to compensate for individual measurement errors, the values which appear plausible for the distance (d), the velocity (v) or the acceleration (a) are smoothed by comparing the corresponding values of successive intervals.
8. Warning system according to one of Claims 1 to 7,
   characterized in that, in order to compensate for evaluation errors with respect to the driving behaviour of the vehicle (5) which is approaching on the closed lane, the output variables for the probability of a collision are averaged over a plurality of time intervals.
9. Warning system according to one of Claims 1 to 8,
   characterized in that the warning device is not triggered in order to compensate for evaluation errors with respect to the driving behaviour of the vehicle (5) which is approaching on the closed lane until the output variables for the probability of a collision have exceeded a predefined threshold value over a plurality of time intervals.
10. Warning system according to one of Claims 1 to 7,
    characterized in that the warning device is of multistage design, wherein an individual critical probability value for an imminent collision of the vehicle (5) which is approaching on the closed lane in the vicinity of the roadworks is assigned to each warning stage, and individual warning signals for the approaching vehicle and/or the people working in the vicinity of the roadworks are provided for the warning stages.
11. Warning system according to one of Claims 1 to 10,
    characterized in that a monitoring device for taking images of the predetermined section of road ahead of the lane-closed sign is also provided, wherein the images taken within a period of time are stored together with the measured values of the detector device which are recorded at the same time, triggered by a predefined event.

Images