CN101395647B - Object course prediction method, device, program, and automatic driving system - Google Patents

Abstract

An object course prediction method capable of ensuring safety under circumstances that may occur as reality, a device, a program, and an automatic driving system. For this purpose, a computer that is provided a storing means for storing the position of an object and internal states including the object's speed reads the position of the object and internal states from the storing means, creates, as a track in a time/space constituted of time and space, changes in position that the object can assume with elapse of time based on the read position of the object and internal states, and stochastically predicts the course of the object by using this created track.

Description

Object route Forecasting Methodology, device and program and automatic operation system

Technical field

The present invention relates to a kind of based on the position of object and object route Forecasting Methodology, device and program and a kind of automatic operation system of internal state prediction object route.

Background technology

In recent years, having carried out various trials realizes the automatic operation such as the movable body of four wheeler.For the automatic operation that realizes movable body, correctly detect such as vehicle, pedestrian and to appear at the object of the obstacle around the movable body and avert danger based on testing result when the operation be important.In these two factors, the known technology that will utilize the object detection technology of various sensors and radar as object around accurately detecting.

The automatic operative technique of movable body be a kind of only by the input destination make movable body move to the technology of destination automatically from starting point.When moving range is narrow, can also predict that in advance the influence of dynamic barrier realizes this technology with the route technology of looking for by the map of setting up moving range in advance.Yet when the moving range of movable body was wide, such as when movable body is automobile, operative technique can't realize with the route technology of looking for automatically.Here wide region is to avoid required time t of dynamic barrier and operation scope all mutually far short of what is expected apart from required time τ, for example is that wherein τ is several hours and t is the situation in several seconds.

If the moving range of movable body is wide, operative technique can not realize mainly containing two reasons by the route technology of looking for automatically.At first, first reason is as follows: for example consider when movable body after starting point through the situation the during time of about 10t.In this case, the influence of dynamic barrier is deployed on the whole road, and then can't define the route that can not bump.That is, if the moving range of movable body is wide, can't precompute route from starting point to the destination.

Secondly, second reason is as follows: if the moving range of movable body is wide, as previously discussed, operation is all long a lot of than t apart from required time τ.Therefore, the computing machine that is installed on the automobile can not be finished required calculating within the utility time of actual collision that can realize dodging.

In the automatic operative technique such as the movable body that in wide region, moves of automobile, as previously discussed, look for the technology except not considering influence of other dynamic barriers at least or the route that does not need in practice to calculate its influence, also need a kind of route calculation technology, finish in utility time by this technology and avoid colliding required calculating to calculate the route that when moving, averts danger with dynamic barrier.

The technology of avoiding threatening during for the operation of above-mentioned route calculation technology, known a kind of technology, by this technology, in the system that forms by a plurality of objects and main vehicle, by utilizing route that information relevant with speed with the position of main vehicle and the information relevant with speed with the position of a plurality of objects except that main vehicle generates each object that comprises main vehicle possibility (for example, referring to non-patent literature 1) with any two object collisions of predicting this system of formation.According to this technology, predict route that all objects that constitute this system are taked and with its output by means of the sequence of operation of the same framework that utilizes the probability notion.Then, predict the outcome, judge and output is used to the total system that comprises main vehicle to realize the route of safety case based on resulting.

Non-patent literature 1:A.Broadhurst, S.Baker and T.Kanade, " Monte Carlo road safety demonstration " (" Monte Carlo Road Safety Reasoning "), (the 4th of IEEE intelligent vehicle discussion (IEEE IntelligentVehicle Symposium), 2005), IEEE (in June, 2005).

Summary of the invention

The problem to be solved in the present invention

Yet because the technology that non-patent literature 1 discloses is conceived to make in prediction all safe route of all objects of construction system, and whether the uncertain route that obtains from this prediction can fully guarantee the safety of specified object (such as main vehicle).

This names a person for a particular job and more specifically describes.In the actual road conditions, the driver of another vehicle or pedestrian's possible errors ground identification road conditions cause under the situation that the party concerned does not recognize comprising the unfavorable behavior of object on every side of main vehicle.On the contrary, non-patent literature 1 supposes that as usual all objects all can show the preferential behavior of safety, therefore do not know in the situation that reality takes place, such as at some object when object has unfavorable behavior on every side, whether also can guarantee safety.

Even consider that above situation is made the present invention and the object of the invention provides object route Forecasting Methodology, device and program and the automatic operation system that also can guarantee safety in the situation that reality takes place.

The means of dealing with problems

In order to address the above problem and to achieve the goal, object route Forecasting Methodology according to the present invention is a kind of object route Forecasting Methodology that is used for by the route of computer forecast object, described computing machine has memory cell, described memory cell is stored the position of described object at least and is comprised the internal state of the speed of described object, and described method comprises: the time aerial track generation step that generates track that the variation in the position that can take according to the described object of passing in time based on the position of the described object that is read and internal state position of reading described object from described memory cell and internal state after is being made of time and space; And by utilizing the prediction steps that generates the route of the described object of track probability ground prediction that generates in the step at described track.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described track generates step and comprises: step is selected in the operation that is chosen in the operation that described object carries out from a plurality of operations; Make the operation of in described operation selection step, selecting carry out the object control step of a scheduled time slot; And judge whether the position of carrying out the described object in selected operation back in described object control step and internal state satisfy the determining step of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein, repeat one group of processing till reaching the track rise time that generates described track from described operation selection step to described determining step.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described operation selects step to select probability to select operation according to the operation that each operation in described a plurality of operations is authorized, if and judged result is that the position and the internal state of described object satisfies described controlled condition and described mobile condition in described determining step, then increases and be back to described operation after the described period and select step.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, select probability by utilizing random number to define described operation.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, preestablish and to generate the track number that generates in the step at described track.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, if judged result is for satisfying described controlled condition and described mobile condition in described determining step, but then carries out recursive call after the described period all selection operations are performed by increasing.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, the position and the internal state of a plurality of objects of described cell stores, and described track generates step aerial track that generates each object in described a plurality of objects when described.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described prediction steps is specified an object and is calculated the aerial probability that exists when described of object except that specified object from described a plurality of objects.

Further, according to object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, comprise that further output comprises the output step of the information that predicts the outcome in the described prediction steps.

Further, object route Forecasting Methodology according to the present invention is a kind of object route Forecasting Methodology that is used for by the route of a plurality of objects of computer forecast, described computing machine has memory cell, described memory cell is stored the position of described a plurality of objects at least and is comprised the internal state of the speed of each object, and described method comprises: the time aerial track generation step that generates track that the variation of the position that can take according to each object the described a plurality of object of passing in time based on the position of the described object that is read and internal state position of reading described a plurality of objects from described memory cell and internal state after is being made of time and space; By utilizing the prediction steps that generates the route of the described a plurality of objects of track probability ground prediction that generate in the step at described track; And calculating the degree of disturbance calculation procedure of degree of disturbance based on predicting the outcome in the described prediction steps, described degree of disturbance is represented the annoyance level between the route that route that specified object can be taked and other objects can take quantitatively.

Further, according to object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculation procedure is leaned on to such an extent that the value of the degree of disturbance between described specified object and each other object is increased or reduce specified amount than the nearer number of times of interference distance according to described specified object and each other object, and described interference distance is the space length that object interferes with each other.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, when described degree of disturbance calculation procedure is leaned on closelyer than described interference distance at described specified object and one of other objects, and the probability of the two articles that draws closer together aerial collision when described increases the value of the degree of disturbance between the described two articles pro rata.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, when described degree of disturbance calculation procedure is leaned on closelyer than described interference distance at described specified object and one of other objects, and the value that increases the degree of disturbance between the described two articles of the relative velocity in moment of drawing closer together of the two articles that draws closer together with being in proportion.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described memory cell will be used to estimate the size of the relative velocity when colliding between the damage grade evaluation value of the damage grade that collision causes or loss from spoilage amount that collision causes and the different object and store accordingly, and described degree of disturbance calculation procedure reads described damage grade evaluation value or described loss from spoilage amount in the size of the relative velocity in the moment that draws closer together from described memory cell according to two articles when described specified object and one of other objects lean on closelyer than described interference distance, and and described damage grade evaluation value or described loss from spoilage amount increase degree of disturbance between the described two articles pro rata.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculation procedure will be the value of described degree of disturbance from the initial position of each object from the required described time set of described initial position at one of described specified object and other objects when required time is less than the value of the degree of disturbance the described two articles when leaning on closelyer than described interference distance.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculation procedure is weighted the back addition to the value of each degree of disturbance between described specified object and other objects.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described track generates step and comprises: step is selected in the operation that is chosen in the operation that described object carries out from a plurality of operations; Make the operation of in described operation selection step, selecting carry out the object control step of a scheduled time slot; And judge whether the position of carrying out the described object in selected operation back in described object control step and internal state satisfy the determining step of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein, repeat one group of processing till reaching the track rise time that generates described track from described operation selection step to described determining step.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described operation selects step to select probability to select operation according to the operation that each operation in described a plurality of operations is authorized, if and judged result is that the position and the internal state of described object satisfies described controlled condition and described mobile condition in described determining step, then increases and be back to described operation after the described period and select step.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, select probability by utilizing random number to define described operation.

Further, according to object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, preestablish and to generate the track number that generates in the step at described track.

Further, according to object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, further comprise the output step of output corresponding to the information of the degree of disturbance that calculates in the described degree of disturbance calculation procedure.

Further, according to object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, further comprise the route selection step of selecting to be included in the route that the described specified object within described a plurality of object will take according to the degree of disturbance that in described degree of disturbance calculation procedure, calculates.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just more little, and described route selection step is selected the route of degree of disturbance minimum.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described route selection step is selected the route with predetermined additional choice criteria optimum matching from described a plurality of routes when the route of a plurality of degree of disturbance minimums is arranged.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just big more, and described route selection step is selected the route of degree of disturbance maximum.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described route selection step is selected the route with predetermined additional choice criteria optimum matching from described a plurality of routes when the route of a plurality of degree of disturbance maximums is arranged.

Further, according to object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, further be included in according to the record of the position of selected route in the described route selection step and the sequence of operation that is used to realize described route is transferred to exterior actuated signal transmitting step with the actuated signal that is produced after producing actuated signal.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculation procedure is leaned on to such an extent that the value of the degree of disturbance between described specified object and each other object is increased or reduce specified amount than the nearer number of times of interference distance according to described specified object and each other object, and described interference distance is the space length that object interferes with each other.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculation procedure is weighted the back addition to the value of each degree of disturbance between described specified object and other objects.

Further, in object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, described track generates step and comprises: step is selected in the operation that is chosen in the operation that described object carries out from a plurality of operations; Make the operation of in described operation selection step, selecting carry out the object control step of a scheduled time slot; And judge whether the position of carrying out the described object in selected operation back in described object control step and internal state satisfy the determining step of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein, repeat one group of processing till reaching the track rise time that generates described track from described operation selection step to described determining step.

Further, according to object route Forecasting Methodology of the present invention, in aforesaid an aspect of of the present present invention, further comprise the output step of the information that output is relevant with the route of selecting in described route selection step.

Object route prediction unit according to the present invention comprises: store the position of object and the memory cell of the internal state of the speed that comprises described object at least; The time aerial track generation unit that generates track that variation in the position that can take according to the described object of passing in time based on the position of the described object that is read and internal state after position of reading described object from described memory cell and the internal state is being made of time and space; And the predicting unit of predicting the route of described object by the track probability ground that utilizes described track generation unit to generate.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described track generation unit comprises: the operation selected cell that is chosen in the operation that described object carries out from a plurality of operations; The object operating unit of a scheduled time slot is carried out in the operation that described operation selected cell is selected; And judge in position that described object operating unit is carried out the described object in selected operation back the judging unit that whether satisfies controlled condition relevant and the mobile condition relevant with internal state with the moving area of described object with the control of described object, wherein, repeat one group of operation of carrying out from described operation selected cell and select to handle the processing of the judgment processing of carrying out to described judging unit till reaching the track rise time that generates described track.

Further, according to object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described operation selected cell selects probability to select operation according to the operation that each operation in described a plurality of operations is authorized, if and the judged result of described judging unit is that the position and the internal state of described object satisfies described controlled condition and mobile condition, then increases and be back to the operation that described operation selected cell carries out after the described period and select to handle.

Further, according to object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, select probability by utilizing random number to define described operation.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, preestablishing will be by the track number of described track generation unit generation.

Further, according to object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, if the result of the judgement that described judging unit is carried out is for satisfying described controlled condition and described mobile condition, but then carries out recursive call after the described period all selection operations are performed by increasing.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, the position and the internal state of a plurality of objects of described cell stores, and described track generation unit aerial track that generates each object in described a plurality of objects when described.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described predicting unit is specified an object and is calculated the aerial probability that exists when described of object except that the object of described appointment from described a plurality of objects.

Further, according to object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, comprise that further output comprises the output unit of the information that predicts the outcome of described predicting unit.

Object route prediction unit according to the present invention comprises: store the position of a plurality of objects and the memory cell of the internal state of the speed that comprises each object at least; The time aerial track generation unit that generates track that the variation of the position that can take according to each object the described a plurality of object of passing in time based on the position of the described object that is read and internal state after position of reading described a plurality of objects from described memory cell and internal state is being made of time and space; Predict the predicting unit of the route of described a plurality of objects by the track probability ground that utilizes described track generation unit to generate; And calculating the degree of disturbance calculating unit of degree of disturbance based on predicting the outcome of described predicting unit, described degree of disturbance is represented the annoyance level between the route that route that described specified object can be taked and other objects can take quantitatively.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculating unit leans on to such an extent that the value of the degree of disturbance between described specified object and each other object is increased or reduce specified amount than the nearer number of times of interference distance according to described specified object and each other object, and described interference distance is the space length that object interferes with each other.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, when described specified object and one of other objects lean on closelyer than described interference distance, and the collision probability of the two articles that draws closer together in the air when described increases the value of the degree of disturbance between the described two articles pro rata.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculating unit in described specified object with when moving to such an extent that becomes nearer than described interference distance with one of other objects with the value of the degree of disturbance of the two articles that draws closer together between the described two articles of increase of mobile the relative velocity that becomes nearer moment with being in proportion.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described memory cell will be used to estimate the size of the relative velocity when colliding between the damage grade evaluation value of the damage grade that collision causes or loss from spoilage amount that collision causes and the different object and store accordingly, and described degree of disturbance calculating unit reads described damage grade evaluation value or described loss from spoilage amount according to the size at the relative velocity in the moment that described two articles draws closer together from described memory cell when described specified object and one of other objects lean on closelyer than described interference distance, and and described damage grade evaluation value or described loss from spoilage amount increase degree of disturbance between the described two articles pro rata.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculating unit will be the value of described degree of disturbance from the initial position of each object from the required described time set of described initial position at one of described specified object and other objects when required time is less than the value of the degree of disturbance the described two articles when leaning on closelyer than described interference distance.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculating unit is weighted the back addition to the value of each degree of disturbance between described specified object and other objects.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described track generation unit comprises: the operation selected cell that is chosen in the operation that described object carries out from a plurality of operations; The object operating unit of a scheduled time slot is carried out in the operation that described operation selected cell is selected; And judge whether position that described object operating unit carries out the described object in selected operation back and internal state satisfy the judging unit of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein, repeat one group of operation of carrying out from described operation selected cell and select to handle the processing of the judgment processing of carrying out to described judging unit till reaching the track rise time that generates described track.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described operation selected cell selects probability to select operation according to the operation that each operation in described a plurality of operations is authorized, if and the result of the judgement carried out of described judging unit satisfies described controlled condition and described mobile condition for the position and the internal state of described object, then increase and be back to the operation that described operation selected cell carries out after the described period and select to handle.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, select probability by utilizing random number to define described operation.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, preestablishing will be by the track number of described track generation unit generation.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, further comprise the output unit of the information of the degree of disturbance that output calculates corresponding to described degree of disturbance calculating unit.

Further, according to object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, further comprise the route selection unit of selecting the route that described specified object will take according to the degree of disturbance that calculates by described degree of disturbance calculating unit.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just more little, and the route of degree of disturbance minimum is selected in described route selection unit.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, the route with predetermined additional choice criteria optimum matching is selected in described route selection unit from described a plurality of routes when the route of a plurality of degree of disturbance minimums is arranged.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just big more, and the route of degree of disturbance maximum is selected in described route selection unit.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, the route with predetermined additional choice criteria optimum matching is selected in described route selection unit from described a plurality of routes when the route of a plurality of degree of disturbance maximums is arranged.

Further, according to object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, the record and being used to that further is included in the position of the route of selecting according to described route selection unit realizes that the sequence of operation of described route is transferred to exterior actuated signal transmitting device with the actuated signal that is produced after producing actuated signal.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculating unit must increase the value of the degree of disturbance between described specified object and each other object or reduces specified amount according to described specified object and each other movement of objects than the nearer number of times of interference distance, and described interference distance is the space length that object interferes with each other.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described degree of disturbance calculating unit is weighted the back addition to the value of each degree of disturbance between described specified object and other objects.

Further, in object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, described track generation unit comprises: the operation selected cell that is chosen in the operation that described object carries out from a plurality of operations; The object operating unit of a scheduled time slot is carried out in the operation that described operation selected cell is selected; And judge whether position that described object operating unit carries out the described object in selected operation back and internal state satisfy the judging unit of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein, repeat one group of operation of carrying out from described operation selected cell and select to handle the processing of the judgment processing of carrying out to described judging unit till reaching the track rise time that generates described track.

Further, according to object route prediction unit of the present invention, in aforesaid an aspect of of the present present invention, further comprise the output unit of the information that output is relevant with the route of described route selection unit selection.

A kind of object route predictor according to the present invention makes described computing machine carry out the object route Forecasting Methodology of the above-mentioned either side according to the present invention.

A kind of automatic operation system according to the present invention is a kind of automatic operation system that is installed on the vehicle, is used for automatically operating described vehicle, and it comprises: according to the object route prediction unit of aforesaid either side of the present invention; And realize route of selecting by the route selection unit that is located in the described object route prediction unit and the actuating device of operating described vehicle according to actuated signal.Effect of the present invention

Object route Forecasting Methodology, device and program and automatic operation system in according to the present invention are even also can guarantee safety in the situation that reality can take place.

Description of drawings

Fig. 1 is the block diagram that illustrates according to the functional structure of the object route Forecasting Methodology of first embodiment of the invention;

Fig. 2 is the diagram of circuit that illustrates according to the overview of the object route Forecasting Methodology of first embodiment of the invention;

Fig. 3 illustrates the diagram of circuit that the track of carrying out according to the object route Forecasting Methodology of first embodiment of the invention generates the overview of handling;

Fig. 4 is the chart that is shown schematically in the track that generates in the three-dimensional space-time;

Fig. 5 is the chart that is shown schematically in the one group of track that generates in the described three-dimensional space-time;

Fig. 6 is schematically illustrated by the instruction diagram according to the structure of the formed space-time environment of object path Forecasting Methodology of first embodiment of the invention;

Fig. 7 is the diagram that illustrates according to the demonstration output example that predicts the outcome of the object route prediction unit of first embodiment of the invention;

Fig. 8 is the diagram that illustrates according to the demonstration output example that predicts the outcome (second example) of the object route prediction unit of first embodiment of the invention;

Fig. 9 is the chart of the structure of the space-time environment that forms when being shown schematically in the pattern that adopts the operation keep main vehicle;

Figure 10 illustrates the diagram of circuit that the track of carrying out according to the object route Forecasting Methodology of second embodiment of the invention generates the overview of handling;

Figure 11 be illustrate the track of carrying out according to the object route Forecasting Methodology of second embodiment of the invention generate handle the diagram of circuit of details;

Figure 12 is the block diagram that illustrates according to the functional structure of the object route prediction unit of third embodiment of the invention;

Figure 13 is the diagram of circuit that illustrates according to the overview of the object route Forecasting Methodology of third embodiment of the invention;

The problem of the schematically illustrated conventional route prediction and calculation of Figure 14;

Figure 15 is the diagram of the advantage of the route prediction and calculation carried out of schematically illustrated object route Forecasting Methodology according to third embodiment of the invention;

Figure 16 is the diagram of circuit that the details of the degree of disturbance computing of carrying out according to the object route Forecasting Methodology of third embodiment of the invention is shown;

Figure 17 is the chart of a THE RELATIONSHIP OF THE PATH of a track of aerial main vehicle when being shown schematically in and another vehicle;

Figure 18 is the chart that casehistory provides the function of the time dependence that disturbs between the object;

Figure 19 is the diagram of circuit that the details of the degree of disturbance computing of carrying out according to the object route Forecasting Methodology of four embodiment of the invention is shown;

Figure 20 is the diagram of circuit that the details of the degree of disturbance computing of carrying out according to the object route Forecasting Methodology of fifth embodiment of the invention is shown;

Figure 21 is the instruction diagram of the another kind of structure of schematically illustrated space-time environment;

Figure 22 illustrates the block diagram that comprises according to the functional structure of the automatic operation system of the object route prediction unit of sixth embodiment of the invention;

Figure 23 is the diagram of circuit that illustrates according to the overview of the object route Forecasting Methodology of sixth embodiment of the invention;

Figure 24 is the diagram of circuit that the details of the route selection processing of carrying out according to the object route Forecasting Methodology of sixth embodiment of the invention is shown;

Figure 25 illustrates the diagram of setting result's demonstration output example according to the route of the object route prediction unit of sixth embodiment of the invention;

Figure 26 illustrates the diagram of setting result's demonstration output example (second example) according to the route of the object route prediction unit of sixth embodiment of the invention;

Figure 27 is the diagram of circuit that the details of the degree of disturbance computing of carrying out according to the object route Forecasting Methodology of the 7th embodiment of the present invention is shown; And

Figure 28 is the diagram of circuit that the details of the route selection processing of carrying out according to the object route Forecasting Methodology of seventh embodiment of the invention is shown.

The explanation of numeral and letter

1,101,201 object route prediction units

2 input sections

3 sensor segments

4 tracks generate section (track generation unit)

5,105 prediction sections (predicting unit)

6,107,209 deferent segments (output unit)

7,108,210 memory paragraphs (memory cell)

41 operation selection portions (operation selected cell)

42 object operating portions (object operating unit)

43 judging parts (judging unit)

51 prediction and calculation portions

52 image production parts

61,171,291 display parts

106 degree of disturbance calculating parts (degree of disturbance calculating unit)

172,292 alarm tones send portion

207 route selection portions (route selection unit)

208 actuated signal transport parts (actuated signal transmission unit)

211 actuating devices

1000 automatic operation systems

B1, B2, B3 route

The CN read-out

D _a, D _bThe zone

Env (P1, P2), Env ' (P1, P2), Env (P ₁, P ₂, P ₃) space-time environment

The F Windshield

The H arrow

O ₁, O ₂, O ₃Object

R, Rd road

The ST bearing circle

The specific embodiment

Explanation is used to implement best mode of the present invention (hereinafter referred to as " embodiment ") below with reference to accompanying drawings.

(first embodiment)

Fig. 1 is the block diagram that illustrates according to the functional structure of the object route prediction unit of first embodiment of the invention.Object route prediction unit 1 shown in the figure is a kind of being installed in such as being present in the object in the preset range of main vehicle periphery with detection on the movable body of four wheeler and predicting the device of the route of detected object and main vehicle.

Object route prediction unit 1 comprises the input section 2 of importing various information from the outside, be used to detect the position that is present in the object in the preset range and the sensor segments 3 of internal state, the variation that is used for the position that can take according to the object of passing in time based on the testing result of sensor segments 3 generates section 4 at the time aerial track that generates track that is being made of time and space, be used to utilize track to generate section 4 tracks that generate carry out the probability prediction to the route of object a prediction section 5, be used to export the deferent segment 6 that comprises the prediction section 5 various information that predict the outcome of making at least and be used to store the memory paragraph 7 that comprises track aerial when track generation unit 4 is created on and prediction section 5 information that predict the outcome of making.

Input section 2 has the function of the set information of the various routes that are used to predict object of input etc. and by utilizing remote controller, keyboard (comprising the touch screen type that input operation can be carried out on screen), pointing device realizations such as (such as mouse and tracking plates).Microphone can be made as input section 2, form the inlet of acoustic information by microphone.If default various set informations, the memory paragraph 7 that then has the ROM (read-only memory (ROM)) of storage this type of information can replace the function of input section 2.

Sensor segments 3 is by utilizing realizations such as millimeter wave radar, laser radar or imageing sensor.Sensor segments 3 has various sensors, such as speed sensor, acceleration pick-up, rotation angle sensor and angular velocity sensor, and also can detect the mobile situation of main vehicle.The internal state of sensor segments 3 detected objects is a kind of can be used for the predicting useful state of object and physical quantitys preferably, such as speed (having speed and direction) and cireular frequency (having size and Orientation).Certainly, in the physics situation (object is in static state) of measuring 0 value is also included within.

Track generates section 4 predictions and also is created on the track that can follow through object before the schedule time, and have be used for from a plurality of operations select to be used to make object with the operation selection portion 41 of the virtual mobile operation of analog system, be used for the object operating portion 42 of the operation selected in a scheduled time slot executable operations selection portion 41 and be used to judge carrying out the position of described operation back object and the judging part 43 whether internal state satisfies predetermined condition at object operating portion 42.

Prediction section 5 has and is used for generating and will being shown and the image production part 52 of the image of output by deferent segment 6 by utilizing the track that generates each object of section 4 outputs from track to carry out the prediction and calculation portion 51 of probability prediction and calculation and being used for result according to the prediction and calculation of prediction and calculation portion 51.

Deferent segment 6 has the display part 61 that is used for showing and exporting the image that is generated by the image production part 52 of predicting section 5.Display part 61 is realized by utilizing telltales such as liquid crystal, plasma, electroluminescence.In first embodiment, projector is arranged on the top of operator's saddle back as display part 61.Projector has the function of permission by showing on the Windshield that is superimposed upon four wheeler.Export acoustic information to exterior loud speaker and can be set to deferent segment 6.

Except that the testing result of sensor segments 3, memory paragraph 7 also storage track generates the predicting the outcome of tracks that section 4 generates, prediction section 5, track and generates operation that the operation selection portion 41 in the section 4 selects etc.Memory paragraph 7 realizes by utilizing ROM and RAM (random access memory), stores the program that is used to start predetermined OS (operating system) among the ROM in advance, according to the object route predictor of first embodiment etc., stores operating parameter, data etc. among the RAM.Memory paragraph 7 also can be by being provided with interface and outfit and the cooresponding recording medium realization of described interface that can equip computer readable recording medium storing program for performing for object route prediction unit 1.

Object route prediction unit 1 with above functional structure is the electronics package (computing machine) of a kind of CPU of being provided with (central processing unit), and described CPU has operation and controllable function.The CPU that is provided with object route prediction unit 1 carries out the processing that is used for according to the object route Forecasting Methodology of first embodiment by read information that is stored in the memory paragraph 7 and the various programs that comprise object route predictor from memory paragraph 7.

Can be according to the object route predictor of first embodiment by being recorded in such as in the computer readable recording medium storing program for performing of hard disk, floppy disk, CD-ROM, DVD-ROM, flash memory, MO dish etc. and extensively issue.

Next, with the object route Forecasting Methodology of explanation according to first embodiment of the invention.Fig. 2 is the diagram of circuit that illustrates according to the processing overview of the object route Forecasting Methodology of first embodiment.Below in the explanation, suppose that all objects that will predict move on two dimensional surface.

At first, sensor segments 3 detects with respect to the position of the object in the preset range of main vehicle and internal state and with detected information storage (step S1) in memory paragraph 7.After this, the position of supposing object is represented by the central value of object and the internal state of object is specified by speed (speed v, direction θ).At step S1, main vehicle inside state is also detected naturally and be stored in the memory paragraph 7.

Next, track generation unit 4 generates track (step S2) by the testing result of utilizing sensor segments 3 inputs for each object.Fig. 3 illustrates the diagram of circuit that track that track generation unit 4 carries out generates the details of handling.Among this figure, the sum of supposing sensor segments 3 detected objects (comprising main vehicle) is K and is an object O _kThe operation that (1≤k≤K, k are natural numbers) generates track is performed N _kInferior (from this angle, K and N _kAll be natural number).Suppose that it is T (〉 0 that track generates the time (track rise time) in period).

In first embodiment, can be by suitably setting track rise time T (and operating time Δ t described later) predict the external world in the computing time of practicality variation---such as the variation of the route of other vehicles.This also is applicable to other embodiments of the present invention.

At first, be 1 with the value initialization of the counting machine k of recognition object and also expression be used for same object O _kTrack generate the counting machine n of number of times _kBe initialized as 1 (step 201).Hereinafter, will be by with reference to common objects O _kProcessing is described.

Next, track generates section 4 and is set at initial condition (step S202) from the result of memory paragraph 7 read sensor sections 3 detections and with the testing result that reads.More specifically, time t is set at 0 and respectively from sensor segments 3 with initial position (x _k(0), y _kAnd initial internal state (v (0)) _k(0), θ _k(0)) is set at input information (x _k0, y _k0) and (v _kO, θ _k0).

Subsequently, but operation selection portion 41 selects probability to select the operation u that will carry out from a plurality of selection operations in ensuing time Δ t according to the operation that is additional to each operation in advance _k(t) (step S203).For example, by will be as u _k(t) optional operational group { u _KcIn element and predetermined random number correspondence define and select operation u _KcOperation select probability.On this meaning, can give each operation u _KcProbability p (u is selected in different operations _Kc) maybe can give operational group { u _KcIn the probability that equates of all elements.In the latter event, p (u _Kc)=1/ (all optional function digits) is fixing.Also operation can be selected Probability p (u _Kc) be defined as the position of depending on main vehicle and the function of internal state and surrounding road condition.

Operation u _KcUsually by depending on object o _kBut a plurality of elements of selection operation and the content of type constitute.For example, if object o _kBe four wheeler, then the acceleration/accel of this four wheeler or cireular frequency by how steering wheel rotation or how to trample the accelerator decision.In view of this, at object o _k(it is a four wheeler) goes up the operation u that carries out _KcBy the element decision that comprises acceleration/accel and cireular frequency.On the contrary, if object o _kBe the people, then operate u _KcCan specify by speed.

To provide operation u _KcSetting example more specifically.If object o _kBe automobile, degree of will speed up is set in the scope of-10 to+30 (km/h/sec) and deflection angle is set in the scope of-7 to+7 (deg/sec) (in two kinds of situations, direction is all specified by symbol).If object O _kBe the people, with rate setting in the scope of 0 to+36 (km/h) and with direction setting in the scope of 0 to 360 (deg).Amount described herein all is a continuous quantity.In this situation, be limited to constitute the operational group { u of each operation by carrying out the suitable discrete number of elements that makes each operation _Kc.

Then, object operating portion 42 makes the operation u that selects in step 203 _KcExecution time Δ t (step S204).Time Δ t is preferably little aspect precision, but can be the value of about 0.1 to 0.5 (sec) in practice.Below in the explanation, suppose that track rise time T is the integral multiple of Δ t, yet the value of T can be according to object O _kRate variation and can not be the integral multiple of Δ t.

Subsequently, judging part 43 judgment object O _kInternal state make the operation u _KcAfter carrying out, step 204 whether satisfies predetermined control condition (step S205) and also judgment object O _kThe position the operation u _KcAfter the execution whether in moving area (step S206).According to object O _kType judge and to be used for the controlled condition judged at step S205, and if object O for example _kBe four wheeler, then by judging at the range of speed after the operation of step S204, the vehicle G etc. of peak acceleration after the operation of step S204.On the other hand, the moving area of judging at step S206 is meant the zone of road (comprising moving traffic lane and footway) etc.When object is positioned at moving area, below will use the phraseology of " satisfying mobile condition ".

If the judged result of judging part 43 is not for satisfying any described condition (be "No" or be "No" at step S206 at step S205), described processing is back to step S202.On the contrary, if the judged result of judging part 43 is object O _kThe position and internal state at the operation u of step S204 _KcAfter satisfy all conditions (be "Yes" and be "Yes") at step S206 at step S205, the time is pushed away Δ t (t ← t+ Δ t) forward and will be (x in the set positions after the operation of step S204 _k(t), y _k(t)) and with internal state be set at (v _k(t), θ _k(t)) (step S207).

Repeat above-mentioned processing at step S202 to S207 up to through till the track rise time T.That is, if do not reach T (is "No" at step S208), then by being back to step S203 reprocessing at the newly-generated time t of step S207.On the other hand, if reach T (is "Yes" at step S208), will be used for the track output of object Ok and be stored in memory paragraph 7 (step S209) in the newly-generated time of step S207.

Fig. 4 is schematically illustrated by repeating the processing of a class range from step S203 to step S207 at time t=0, Δ t, and 2 Δ t ..., the object O that T generated _kThe chart of track.Track P shown in this figure _k(m) (1≤m≤N _k, m is a natural number) pass space two-dimensional (x, y) and the three-dimensional space-time that constitutes of time one dimension (t) (x, y, t).By with P _k(m) be projected on the x-y plane, can obtain object O _kAt two-dimensional space (x, y) the prediction route in.

If counting machine n behind step S209 _kValue do not reach N _k(is "No" at step S210) is then with counting machine n _kValue add 1 (step S211) and repeat processing from step S202 to S207 till reaching track rise time T by being back to step S202.

If counting machine n _kS210 reaches N in step _k(is "Yes" at step S210) then finishes at object O _kThe generation of whole tracks.Fig. 5 is schematically illustrated by N _kIndividual in described three-dimensional space-time at an object O _kThe track P that generates _k(1), P _k(2) ..., P _k(N _k) trajectory set { P that formed _k(n _k) instruction diagram.Constitute trajectory set { P _k(n _k) the starting point of each track of element---that is initial position (x _K0, y _K0, t)---identical (referring to step S202).Have, Fig. 5 is a kind of scheme drawing and N on stricti jurise again _kValue desirable such as the value of hundreds of to several ten thousand.

If counting machine n _kS210 reaches N in step _kAnd the counting machine k that is used for object identification does not reach the total K of object (is "No" at step S212), then the value of counting machine k is added 1 and track is generated the counting machine n of number of times before being back to step S202 _kValue initialization be that 1 (step S213) is to repeat described processing.On the contrary, if the counting machine k of described object reaches K (is "Yes" at step S212), then generate and finished and before the track of step S2 generates the step S3 that handles below proceeding to, stop at the track of all objects.

As previously discussed, by with the pre-determined number being sensor segments 3 detected all objects execution tracks generation processing, the space-time environment that the trajectory set that formation is followed by a plurality of objects in the preset range that is present in three-dimensional space-time is formed.Fig. 6 is the instruction diagram of the topology example of schematically illustrated space-time environment.Space-time environment Env (P shown in the figure ₁, P ₂) by object O ₁Trajectory set { P ₁(n ₁) (in Fig. 6, representing with solid line) and object O ₂Trajectory set { P ₂(n ₂) (in Fig. 6, representing with solid line) composition.More specifically, (P1, P2) representative is as two articles O for space-time environment Env ₁And O ₂At space-time environment such as the straight road R upper edge of express highway+when the y direction of principal axis moves.Owing to do not consider correlativity between the object for each object generates track independently in the first embodiment, the track of different objects may this time intersect in the air.

In Fig. 6, the trajectory set { P in each zone of space-time _k(n _k) density of the per unit volume of (k=1,2) provides object O _kThe density that has probability in each zone of described space-time (hereinafter referred to as " space-time probability density ").Therefore, handle the space-time environment Env (P that makes up by utilizing generating of step S2 by track ₁, P ₂), can determine object O _kPass the probability of the presumptive area in the described three-dimensional space-time.Because the notion of above-mentioned space-time probability density aerial probability when on stricti jurise being, the summation of its value may not be 1 in the air when relevant with object.

If the track rise time, T should be redefined for fixed value, if the distribution of aerial probability density will be that the same there is no need calculated during thereby its occurrence preferably makes track overtime T generate.If for example described object is four wheeler and this four wheeler cruising, then T can be set at about 5 (sec) at the most.In this situation, if be about 0.1 to 0.5 (sec), then in order to generate a track P at the operating time of step S204 Δ t _k(m) one group of processing from step S203 to step S207 of carrying out repeats 10 to 50 times.

Have again, after being to set track rise time T such as the different roads of express highway, ordinary road and two-lane road, preferably based on utilize by it position data read from map datum current operation road type method or come T transfer time by the method that the road Identification device by application image identification etc. reads road type.

Also preferably carry out adaptive control, wherein, the track that calculates in utilization is estimated the distribution of probability density aerial when described with statistical, up to track rise time T, if described distribution is uniform, then reduce track rise time T, and if described distribution be not uniformly, then increase the described rise time.

Further, the probability that also can intersect by the route of preparing a plurality of routes that can be taked by main vehicle and the route that utilizes main vehicle and another object in advance permanent timer-operated time that becomes makes a prediction as track rise time T.In this situation, also can adopt risk increment after predicted time only increases Δ t, to become the permanent timer-operated time as end condition when each route that main vehicle can be taked.In this mode, the end points on the future of route that main vehicle can be taked side is set to naturally spatially and extensively distributes to obtain in order to ensure the foundation of safety to the judgement of the current route that will take.

After the track at above-mentioned each object generated processing, the route that prediction section 5 can be taked each object was made probability prediction (step S3).To illustrate below from object O _kTrajectory set { the P that generates _k(n _k) the middle intended trajectory P that selects _k(m) probability, it is handled for concrete prediction and calculation that the prediction and calculation portion 51 that predicts in the section 5 carries out, but this prediction and calculation is an example certainly.

As object O _kN _kWhen individual track generates, calculating N as shown below _kA track P in the individual track _k(m) become the probability of actual path.At first, if realize object O _kTrack P _k(m) the sequence of operation { u _Km(t) } be { u _Km(0), u _Km(Δ t), u _Km(2 Δ t) ...,, u _Km(T) }, then select operation u at time t _Km(t) probability is p (u _Km(t)), therefore, at the sequence of operation { u of time t=0 to T execution _Km(t) } probability draws by formula 1:

[formula 1]

$p\left(u_{\mathrm{km}}\left(0\right)\right)\·p\left(u_{\mathrm{km}}\left(\mathrm{\Δt}\right)\right)\·p\left(u_{\mathrm{kn}}\left(2\mathrm{\Δt}\right)\right)\·\·\·p\left(u_{\mathrm{km}}\left(T\right)\right)=\underset{t=0}{\overset{T}{\mathrm{\Π}}}p\left(u_{\mathrm{km}}\left(t\right)\right)---\left(1\right)$

Therefore, when providing for object O _kN _kIndividual trajectory set { P _k(n _k) time, selected object O _kA track P that can follow _k(m) Probability p (P _k(m)) draw by formula 2:

[formula 2]

$p\left(P_k\left(m\right)\right)=\frac{\underset{t=0}{\overset{T}{\mathrm{\Π}}}p\left(u_{\mathrm{km}}\left(t\right)\right)}{\underset{n=1}{\overset{N_k}{\mathrm{\Σ}}}\left(\underset{t=0}{\overset{T}{\mathrm{\Π}}}p\left(u_{\mathrm{kn}}\left(t\right)\right)\right)}---\left(2\right)$

If select all operations u with the Probability p 0 (0＜p0＜1 herein) that equates _Km(t), formula 1 can be simplified to:

$\underset{t=0}{\overset{T}{\mathrm{\Π}}}p\left(u_{\mathrm{km}}\left(t\right)\right)=p_0^s---\left(3\right)$

Here, s is the sum of operating time Δ t from t=0 to T, that is, number of operations.Therefore, be included in object O _kThe N that can follow _kTrack P in the individual track _kThe summation of probability (m) becomes N _kP ₀ ^sAnd by formula (3) substitution formula (2) being obtained a selected track P in them _k(m) Probability p (P _k(m)).

[formula 4]

$p\left(P_k\left(m\right)\right)=\frac{1}{N_k}---\left(4\right)$

That is, Probability p (P _k(m)) do not depend on track P _k(m).

If the track number that will generate at all objects in formula (4) identical (N) is then from N ₁=N ₂=...=N _k=N (constant) draws p (P _k(m))=and 1/N, it illustrates described probability and is independent of object O _kBe constant.By making Probability p (P in this case _k(m)) value is standardized as 1, can simplify the prediction and calculation that prediction and calculation portion 51 carries out, thereby carries out predetermined prediction and calculation quickly.Have again, can be by making operation u from 2 inputs of input section _Km(t) selecteed Probability p (u _Km(t)) can suitably set or change.

Based at each object O _k(k=1,2 ..., the probability that K) calculates, per unit volume object O in each zone of described three-dimensional space-time judges in prediction and calculation portion 51 _kHave a probability.This exist probability with at trajectory set { P _k(n _k) three-dimensional space-time in the space-time probability density corresponding and have the higher probability that exists usually through the high zone of density of track.

After the operation that so far illustrated prediction and calculation portion 51 carries out, image production part 52 generated and the image-related graphicinformation that will be shown by the display part 61 of deferent segment 6 according to resulting operating result to deferent segment 6 in the transmission graphicinformation.

After above-mentioned steps S3, the information that demonstration/output is consistent with the operating result of prediction and calculation portion 51, that is (step S4) predicts the outcome.Fig. 7 be illustrate display part 61 the demonstration/output example that predicts the outcome chart and be schematically illustrated when by utilizing by two articles O ₁(main vehicle) and O ₂The space-time environment Env (P1, the demonstration/output example that predicts the outcome when P2) (see figure 6) is made a prediction that constitute.More specifically, Fig. 7 illustrates by at object O ₁Translucent stacked another object O that shows on the Windshield F of (main vehicle) ₂At the fixed time exist probability to surpass predetermined threshold regional the time situation.The region D of translucent demonstration _aAnd region D _bHave different illumination (region D _aBrighter).This differences of illumination intensities reflected predicting the outcome of prediction and calculation portion 51 and according to the value of being judged that has probability by the stacked translucent area that shows on Windshield F with different illumination.

By image that image production part 52 is generated from being arranged on object O ₁The projector (it is the part of deferent segment 6 and not shown) on top, operator's saddle rear be projected in Windshield F and go up and realize above-mentioned stacked demonstration.Correspondingly, object O ₁The driver when the working direction of visual main vehicle is driven, can identify the zone that might cause danger in the near future immediately.Therefore, can be by described recognition result being reflected in driving and averting danger exactly.

Yet the demonstration/output example of deferent segment 6 is not limited thereto, and for example, the function that can have display part 61 by the read-out CN (seeing figure) that makes auto-navigation system that predicts the outcome of prediction section 5 shows.In this situation, with region D shown in Figure 8 _aAnd D _bThe same, available color gradient demonstrates each zone on the two dimensional surface that is presented on the read-out CN.Or by making sound export information, alarm tone or the sound consistent from deferent segment 6 with surrounding road condition via microphone generating.

Above-mentioned first embodiment according to the present invention, have the position of storage object and comprise that the computing machine of memory cell of internal state of the speed of object reads the position and the internal state of object from memory cell, the variation of the position that can take according to the object of passing in time based on the position of the object that is read and internal state is at the time aerial track of being made up of time and space that generates, even and also can guarantee safety by utilizing the track probability ground that is generated to predict that the route of object is feasible in the situation that reality can take place.

Equally, according to first embodiment, be formed on the route prediction that the time aerial space-time environment of being made up of time and space is made object by utilization, not only can accurately make the route prediction of stationary object, and can accurately make the route prediction of dynamic object.

Further, according to first embodiment,, specified object (for example main vehicle) and other objects can be made a distinction owing to generate the track of detected object independently.So, can be in the time period of practicality easily and calculate to a nicety and may be hidden in danger between described specified object and other objects.

In addition, according to first embodiment owing to can utilize space-time environment prediction result prompting to comprise dangerous information by output, the driver of main vehicle can rapidly and during accurately avoiding driving in the near future in contingent danger.

Have, as previously discussed, first embodiment also is applicable in the four-dimensional spacetime (three dimensional space and one dimension time) again.First embodiment can be applicable to the automobile that moves on the road of diff-H having certainly, uses in the time of also can carrying out the route prediction to other movable bodies that similarly move in addition at the movable body that moves such as aircraft or helicopter aloft aloft.

Here, the non-patent literature 1 that explanation is quoted in the above-mentioned background technology and the difference of first embodiment.Although these two kinds of technology all utilize the probability notion to make the route prediction of object, the technology in the non-patent literature 1 is not to predict the route of object independently and only make probability calculation based on interrelation in preset range.Therefore, when any two articles in a plurality of objects collided, the route prediction of these two objects stopped when two articles collides.When in three-dimensional space-time, considering, this means the collision judgment that can not carry out after the track cross of two different objects is handled.

On the contrary, in first embodiment, generate object trajectory at each object independently, therefore, even the track of different objects intersects in described three-dimensional space-time, collision judgment is handled and also can be continued till the process schedule time.Therefore, complete different according to non-patent literature 1 space-time environment that generates and the space-time environment that generates according to first embodiment in nature.In addition, owing to look at each object execution path independently under the situation of not considering object collision in the first embodiment, calculated amount is littler than the situation in the non-patent literature 1.

In addition, according to non-patent literature 1, even measurable collision accident, but can not predict when such collision takes place.This is because the technology in the non-patent literature 1 is conceived at each state search collisionless is being arranged all the time, rather than the probability of judgment object collision in time stream.In other words, in non-patent literature 1, do not use space-time environment and yet do not adopt the notion of space-time probability density clearly.

Although make first embodiment and non-patent literature 1 give the impression of similar techniques at prima facie owing to all utilizing the probability notion to make the route prediction, but its know-why is different fully in itself, even and also be difficult to those skilled in the art realize first embodiment from non-patent literature 1.

(remodeling of first embodiment)

The operation selection portion 41 of track generating unit 4 can only be that main vehicle is kept present operation.In this situation, keep main vehicle prediction constantly internal state and will continue to carry out unique operation, the operation selection probability of therefore selecting described operation be 1 and the time only generate the trajectory set of a track in the air as main vehicle.

Fig. 9 illustrates the space-time environment that is generated when the operation of main vehicle such as above-mentioned keeping, and it is and the cooresponding chart of Fig. 6.Space-time environment Env ' (P shown in Figure 9 ₁, P ₂) in, object O ₁In (theme vehicle) trajectory set in described three-dimensional space-time only by a straight path form (with at object O ₂Fig. 6 similar).As previously discussed, keep main vehicle O by application ₁The pattern of operation, can simplify prediction situation during object when for example having around a lot, thereby the calculated amount that generates in section and the prediction section at track reduces.

(second embodiment)

Second embodiment of the invention is characterised in that: when generating track for each object, but carry out all selection operations for track generates.According to the functional structure of the object route prediction unit of second embodiment and the identical (see figure 1) of functional structure according to the object route prediction unit 1 of first embodiment.Except generate at the track of each object handle, according to the object route Forecasting Methodology of second embodiment with identical according to the method for first embodiment.

Figure 10 illustrates the diagram of circuit that the track of carrying out according to the object route Forecasting Methodology of second embodiment generates the overview of handling (corresponding with step S2 among Fig. 2).During the track that illustrates in the drawings generates and handles, at first, carry out initialization and be set at 1 (step S21) with the value of the counting machine k that will be used to discern each object.Equally, in second embodiment, will be assumed to K at its object sum that generates track.

Next, track generation section 4 reads by the result of sensor segments 3 detections and with the testing result that is read from memory paragraph 7 and is set at initial condition (step S22).More specifically, time t is set at 0 and with initial position (x _k(0), y _kAnd initial condition (v (0)) _k(0), θ _k(0)) is set at the input information (x of sensor segments 3 respectively _K0, y _K0) and (v _K0, θ _K0).

Subsequently, generate object O _kAt three-dimensional space-time (x, y, t) track in (step S23).Figure 11 is illustrated in the diagram of circuit that track that step S23 carries out generates the details of handling.Below in the explanation, suppose that the operating time Δ t that track rise time T carries out with each operation represents with T=J Δ t (J is a natural number).

At first, circular treatment (circulation 1) begins (step S231-1) at time t=0.In this circulation 1, the operation u when t=0 _k(0) forms in the time of Δ t.The same with first embodiment, according to object O _kType determine operation u _kParticular content (0) (if object is a vehicle, then can be by acceleration/accel or cireular frequency assigned operation, and, if object is the people, then can be by the speed assigned operation).Operational group { u _KcForm by limited element, and, if but selection operation is a continuous quantity, operational group { u then _KcThe discretization of element by appropriate intervals constitute.

With the more specifically processing of explanation at step S231-1.At first, operation selection portion 41 is selected an operation u _Kc(0) and object operating portion 42 make selected operation u _Kc(0) time of execution Δ t.After this operation, judging part 43 judgment object O _kThe position and internal state whether satisfy with first embodiment in similarly controlled condition and mobile condition.If all conditions all satisfies (OK), then judging part 43 proceeds to circular treatment (circulation 2) at next moment t=Δ t.On the contrary, if any of described controlled condition or described mobile condition do not satisfy (NG), the operation u that just carried out of front then _Kc(0) after proceeding to step S233-1, is cancelled.Because operation selection portion 41 is selected all operations in second embodiment, the order of selecting each operation is arbitrarily.This also be applicable to follow-up circular treatment (circulation 1, the circulation 2 ..., cyclic J).

To at first illustrate below up to the processing that generates a track.In the circulation 2, with the same in above-mentioned circulation 1, operation selection portion 41 is selected an operation and operation u _KcThe time that (Δ t) only carries out Δ t.Then, if object O _kThe position after described operation, satisfy controlled condition similar to the above and mobile condition (OK), then this processing proceeds to circular treatment (circulation 3) at 2 Δ t constantly.On the other hand, if do not satisfy described controlled condition or mobile condition any (NG), the operation u that just carried out of front then _Kc(Δ t) is cancelled after proceeding to step S233-2.

Hereinafter, by repeat with in above-mentioned circulation 1 or circulate and similarly handle in 2, carry out circular treatment J time continuously.That is, as long as object O _kOperate in times 3 Δ t, 4 Δ t ... satisfy controlled condition and mobile condition after the execution, then described processing sequence proceed to the circulation 4, the circulation 5 ....Then, if object O _kSatisfy controlled condition and mobile condition cyclic J to the last, then described processing proceeds to step S232 subsequently.At step S232, will export and be stored in the memory paragraph 7 from the track of t=0 to t=T (=J Δ t).This track and the same three-dimensional space-time that passes of track shown in Figure 4.

Step S233-J after step S232 is at nearest performed operation u of time _Kc(T-Δ t)=u _Kc((J-1) Δ t) is cancelled, and if cyclic J will continue (cyclic J continuation), then handle being back to step S231-J.On the other hand, if cyclic J will stop (cyclic J termination), then handle the step S233-(J-1) that proceeds to subsequently.

At step 233-(J-1), the operation u that in circulation (J-1), carries out _Kc(T-2 Δ t) is cancelled, and, if circulation (J-1) will continue (circulation (J-1) continues), that is, if will be implemented as operation u arbitrarily _KcThe element of (T-2 Δ t) keeps, and then handles being back to step S231-(J-1) with reprocessing.On the other hand, if circulation (J-1) will stop (circulation (J-1) stops), that is, there is not to be implemented as operation u _KcThe element of (T-2 Δ t) keeps, and then handles the step S233-(J-2) that proceeds to subsequently.

Hereinafter, with circulation (J-2) ..., circulation 2, circulation 1 order repeats with above-mentioned cyclic J or circulation (J-1) in similarly handle.Thus, when in the end when step S233-1 finishes circulation and proceeds to processing at step S24 after 1, generated object O _kThe track that might follow, that is generate trajectory set { P _k(n _k), as shown in Figure 5.

Next, situation when not satisfying any of (NG) controlled condition or mobile condition at step S231-1 will be described.In this situation, has just carried out the front operates in and is cancelled after proceeding to step S233-1.Then, 1 will continue, then handle being back to step S231-1 if circulate, and, 1 will stop if circulate, then processing proceeds to step S24 subsequently.

If make step S231-2, S231-3 ... or back object O is carried out in the selected operation of S231-J _kDo not satisfy any controlled condition or mobile condition, then carry out with the processing of above-mentioned steps S231-1 and similarly handle.That is, if common object O _kStep S231-j (j=2,3 ..., J) do not satisfy any executive condition or mobile condition, then after proceeding to step S233-j, can cancel the operation that has just carried out the front.Correspondingly, if carve t at a time _kDo not satisfy any condition, then can omit t _kAfter track generate to handle, realize the minimizing of calculated amount.

Above-mentioned track generates the algorithm of handling and equals based on depth-first search, utilizes the recursive call method to search for the algorithm of all possible operation.Therefore, in this situation, up to object O _kTrack generate finish dealing with till, at an object O _kTrajectory set { P in the terminal point generation _k(n _k) number of elements---that is track number---all be unknown.

Can utilize BFS rather than carry out above-mentioned track and generate to handle and search for all possible operation.In order to generate the track that can follow by each object by searching for all enforceable operations, can select a kind of following searching method, described searching method has according to operation u _Kc(t) number of elements in operating time Δ t is (as operation u _KcDispersion when (t) being continuous quantity) optimal computed amount.

Also do not reach K (being "No") if after the track of above-mentioned steps S23 generates processing, be used for the counting machine k of object identification at step S24, then the value of counting machine adds 1 (step S25), testing result based on sensor segments 3 after being back to step S22 is carried out initialization, and to another object O _K+1Carry out above-mentioned track and generate processing (step S23).On the other hand, reach K (is "Yes" at step S24), then finished track generation processing, therefore stop track and generate processing (corresponding) with step S2 among Fig. 2 to all objects that exist in the preset range if be used for the counting machine k of object identification.Thus, generate and the similar space-time environment Env of space-time environment (P shown in Figure 6 ₁, P ₂) and it is stored in the memory paragraph 7.

After this processing, that is the object route probability prediction (corresponding with step S3 among Fig. 2) that prediction section 5 is carried out and the output that predicts the outcome (corresponding with step S4 among Fig. 2) of deferent segment 6 are similar with first embodiment.

Above-mentioned second embodiment according to the present invention, have the position of storage object and comprise that the computing machine of memory cell of internal state of the speed of object reads the position and the internal state of object from memory cell, the variation of the position that can take according to the object of passing in time based on the position of the object that is read and internal state is at the time aerial track of being made up of time and space that generates, even and make by the route that utilizes the track probability ground prediction object that generates also can guarantee safety in the situation of meeting reality generation.

Equally, according to second embodiment, be formed on the route prediction that the time aerial space-time environment of being made up of time and space is made object by utilization, not only can accurately make the route prediction of stationary object, and can accurately make the route prediction of dynamic object.

Have again, in second embodiment, carry out the time utilize the track of recursive call method to generate in the air to handle, but can not say blood surely that when comparing with the track generation processing in above-mentioned first embodiment, this track generates handles the calculated amount that needs still less.In other words, this be used for the time aerial resultant body track algorithm depend on that various conditions change, described condition comprises the number of elements in operating time Δ t, track rise time T and the operational group.Therefore, can adopt optimal algorithm according to the condition of making a prediction.

(the 3rd embodiment)

Figure 12 is the block diagram that illustrates according to the functional structure of the object route prediction unit of third embodiment of the invention.Object route prediction unit 101 shown in the figure is a kind of being installed in such as the device on the movable body of four wheeler, is present in the degree of disturbance between the route of estimating route that can---main vehicle---can be taked by described specified object as the route of detected object of object, forecasting institute in the preset range of the main vehicle periphery of specified object and main vehicle and based on predicting the outcome quantitatively and can being taked by other objects in order to detection.

Object route prediction unit 101 comprises the input section 2 of importing various information from the outside, be used to detect the position that is present in the object in the preset range and the sensor segments 3 of internal state, the variation that is used for the position of taking according to the object of passing in time based on sensor segments 3 detected results generates section 4 at the time aerial track that generates track that is being made of time and space, be used to utilize track to generate section 4 tracks that generate are made the probability prediction to the route of object a prediction section 105, be used for degree of disturbance calculating part 106 based on the degree of disturbance of predicting the annoyance level between the route that section route that 105 calculating of making that predict the outcome represent that quantitatively main vehicle can be taked and other objects can be taked, be used to export the deferent segment 107 of the various information of the evaluation result that comprises degree of disturbance calculating part 106, and the memory paragraph 108 that is used to store the various information of the position that comprises sensor segments 3 detected objects and internal state.Among Figure 12, give same reference numerals to having with assembly according to the object route prediction unit 1 identical functions structure of first embodiment shown in Figure 1.

Deferent segment 107 has and is used for according to the display part 171 of the evaluation result demonstration/output image of degree of disturbance calculating part 106 and is used for sending portion 172 according to the give the alarm alarm tone of sound of its evaluation result.Alarm tone sends portion 172 by utilizing realizations such as loud speaker.

Except that the testing result of sensor segments 3, memory paragraph 108 also storage track predicting the outcome of generating that tracks, prediction section 105 that section 4 generates make, the degree of disturbance result of calculation of degree of disturbance calculating part 106, track generates operation that the operation selection portion 41 in the section 4 selects etc.

Next, with the object route Forecasting Methodology of explanation according to third embodiment of the invention.Figure 13 is the diagram of circuit that illustrates according to the processing overview of the object route Forecasting Methodology of the 3rd embodiment.In the 3rd embodiment, also the object that all bands are predicted all is assumed to move on two dimensional surface and describes.

In the 3rd embodiment, to the detection of the position of each object and internal state handle (step S31), the time track carried out at each object in the air generate and handle (step S32) and utilize the probability prediction processing (step S33) of object route of track execution identical with above-mentioned steps S1, step S2 respectively with step S3.Below in the explanation, suppose that the track by selecting probability method to carry out at step S32 based on the operation of explanation in the first embodiment generates processing, but also can adopt the method that in second embodiment, illustrates, that is, but by carrying out the method that all selection operations generate track.

Have again, when step S32 the time carry out track at each object in the air and generate when handling, with regard to know-why, importantly stop prediction and calculation, rather than whether arrived predeterminated position (destination or be similar to the midway location of destination) based on main vehicle based on track rise time T.On ordinary road, there is not to guarantee in advance the place of safety.For example, as shown in figure 14, as the main vehicle O that operates in by supposition on the three-lane road Rd ₁Arrive predeterminated position Q successively ₁, Q ₂And Q ₃When predicting, consider main vehicle O ₁Almost there is another vehicle Q in the situation of travelling to described predeterminated position as the crow flies ₂In order to avoid taking route B ₃Another vehicle O ₃The danger that causes and take route B ₂Move to main vehicle O ₁Danger on the moving traffic lane that is just being travelled.Correspondingly, conventional route prediction and calculation does not reach and guarantees main vehicle O in advance ₁The degree of travelling safely to predeterminated position.

Owing to all be in the 3rd embodiment at every turn not default such as main vehicle O ₁Judge best route in the situation of the position of the destination that arrives etc., for example, with Figure 14 in can select route B shown in Figure 15 under the identical situation ₁As main vehicle O ₁Route, making can be at main vehicle O ₁Guarantee safety by accurately averting danger when travelling.

Replace track rise time T, the track of length that can be by the track that will generate is shown generates length and judges the condition that is used to stop prediction and calculation.In this situation, preferably the speed (or speed of main vehicle) according to object generates length to change track adaptively.

To be described in detail in step S34 and subsequent processing below.At step S34, by the degree of disturbance (step S34) between the degree of disturbance calculating part 106 main vehicles of calculating and another vehicle.Figure 16 is the diagram of circuit that the details of degree of disturbance computing is shown.In the 3rd embodiment, suppose object O ₁It is main vehicle.For convenience of explanation, with other objects O _k(k=2,3 ..., K) all be assumed to and also be four wheeler and be called other vehicles O _kDegree of disturbance computing shown in Figure 16 is handled by four on-cycles and is formed and with other vehicles O _kAll trajectory set { P _k(n _k) calculate the main vehicle O that judges at step S33 seriatim ₁Trajectory set { P ₁(n ₁) the degree of disturbance of all elements.

The input that degree of disturbance calculating part 106 receives at step S34 comprises main vehicle O ₁Trajectory set { P ₁(n ₁), other vehicles O _kAll trajectory set { P _k(n _k) and in order to estimate main vehicle O ₁With other vehicles O _kBetween the degree of disturbance evaluation function of degree of disturbance.Supposition degree of disturbance calculating part 106 comprises the degree of disturbance evaluation function in the 3rd embodiment, but this degree of disturbance evaluation function can be imported from the outside.Perhaps, can make the degree of disturbance evaluation function can be according to road type and main vehicle O ₁Velocity adaptive ground change.

Degree of disturbance between the trajectory set of trajectory set by estimating other vehicles and main vehicle with mutually different end points, as utilize as described in Figure 15, each all situations in the position that not have to preset the destination that will arrive such as main vehicle etc. best route that judges accurately averts danger with in main vehicle ' the time, thereby can guarantee safety.Thus, as shown in figure 14,, the predeterminated position of main vehicle on road also can solve the fatal problem of not guaranteeing safety even travelling.

Among Figure 16, degree of disturbance calculating part 106 at first begins main vehicle O ₁The reprocessing (circulation 1) (step S401) of all tracks.Based on this purpose, from trajectory set { P ₁(n ₁) in select a track and selected track carried out subsequent treatment.

Next, degree of disturbance calculating part 106 begins other vehicles O _kReprocessing (circulation 2) (step S402).In this circulation 2, the value with k at every turn when the counting machine k that will be used to discern other vehicles is initialized as k=2 and finishes reprocessing increases progressively.

In circulation 2, carry out being other vehicles O in step 33 _kTrajectory set { the P that is generated _k(n _k) the reprocessing (circulation 3) (step S403) of all elements.In this reprocessing, will be used for repetitive cycling 1 that is be used to be identified as main vehicle O ₁The counting machine n of the track that generates ₁The degree of disturbance of determining with the counting machine k that is used to discern other vehicles is set at r ₁(n ₁, k) and with r ₁(n ₁, value k) is set at 0 (step S404).

Subsequently, the 106 beginning reprocessings (circulation 4) of degree of disturbance calculating part are to estimate main vehicle O ₁Track P ₁(n ₁) and other vehicles O _kTrack P _k(n _k) between interference (step S405).In this circulation 4, successively at moment t=0, Δ t ..., T determines track P ₁(n ₁) and track P _k(n _k) between in the distance of synchronization.Because with the location definition of each track in two-dimension time-space is the center of each vehicle, if the space length between two tracks becomes than predetermined value little (for example, the normal width of vehicle or length), then can suppose main vehicle and other vehicles O _kCollide.On this meaning, even the inconsistent also decidable of the coordinate figure of two vehicles two articles collides.Hereinafter, the maxim (space length that two articles interferes with each other) with the distance that allows to think that two articles has collided is called interference distance.

Figure 17 is schematically illustrated main vehicle O ₁Track P ₁(n ₁) and other vehicles O _kTrack P _k(n _k) the time aerial relation chart.In the situation shown in the figure, track P ₁(n ₁) and track P _k(n _k) at two some C ₁And C ₂Intersect.Therefore, at these two some C ₁And C ₂Near exist in distance between described two tracks of synchronization less than the regional A of interference distance ₁And A ₂That is, as track P ₁(n ₁) and track P _k(n _k) be included in regional A respectively ₁And A ₂Make main vehicle O when interior ₁With other vehicles O _kThe judgement of collision.In other words, at moment t=0, Δ t ..., T passes regional A ₁And A ₂Number of times be main vehicle O ₁With other vehicles O _kBetween the collision number of times.

Can manifest from Figure 17, in the space-time environment that in the 3rd embodiment, forms, even the collision of two tracks also can generate the later track of collision.This is because generate the cause of track independently for each object.

If to main vehicle O ₁With other vehicles O _kBetween the judged result of distance on above-mentioned meaning, making main vehicle O ₁With other vehicles O _kThe judgement of having collided (is "Yes" at step S406), then degree of disturbance calculating part 106 is with degree of disturbance r ₁(n ₁, value k) is set at:

[formula 5]

r ₁(n ₁，k)←r ₁(n ₁，k)+c _1k·p(P _k(n _k))·F(t)(5)

(step S407).Here, second c will be described _1kP (P _k(n _k)) F (t).Coefficient c _1kBe positive constant and for example can be set at c _1k=1.P (P _k(n _k)) be the amount of formula 2 definition and be from other vehicles O _kSelect a track P _k(n _k) probability.Last F (t) is the amount that provides the time dependence between the object in the primary collision.Therefore, if do not allow the free dependence of interference between the object, then the value of F (t) can be set to constant.On the contrary, if allow the free dependence of interference between the object, as legend as shown in Figure 18, F (t) can the value of being defined as be passed and the function that reduces in time.F shown in Figure 18 (t) is suitable for when authorizing importance to last collision.

If time t does not also reach T behind step S407, degree of disturbance calculating part 106 repetitive cycling 4 (being "No" in step S408) then.In this situation, the value of t increases Δ t (step S409) and repetitive cycling 4 after being back to step S405.On the other hand, if time t reaches T behind step S407, then stop circulation 4 (are "Yes" at step S408).If main vehicle O ₁With other vehicles O _kIn not collision of t sometime, then whether degree of disturbance calculating part 106 proceeds directly to the judgment processing (step S408) of repetitive cycling 4.

By the reprocessing of above-mentioned circulation 4, degree of disturbance r ₁(n ₁, value k) increases with collision frequency.After circulation 4 is finished, whether carry out the judgment processing of repetitive cycling 3 at step S410.That is, if be other vehicles O _kThere are any and main vehicle O in the track that generates ₁A track P ₁(n ₁) the also unenforced track of interference evaluation (being "No" in step S410), then with n _kBe incremented to n _k+ 1 (step S411) and after being back to step S403 repetitive cycling 3.

On the contrary, if be other vehicles O _kThe track and the main vehicle O that generate ₁A track P ₁(n ₁) all disturb to estimate and all to carry out (in step S410, being "Yes"), then will estimate main vehicle O ₁Track P ₁(n ₁) and other vehicles O _kAll tracks between resultant interference degree r ₁(n ₁, k) enclose (step S412) and (step S413) in the memory paragraph 108 exported and be stored in to the value of being enclosed.

At the resultant interference degree r of step S413 from 106 outputs of degree of disturbance calculating part ₁(n ₁, value k) depends on from other vehicles O _kAll tracks in select a track P _k(n _k) Probability p (P _k(n _k)).Therefore, if the middle coefficient c of postulation formula (5) _1kDo not depend on k and be constant (for example, c _1k=1), F (t) is constant (for example, 1), and main vehicle O ₁Track P ₁(n ₁) and other vehicles O _kTrack P _k(n _k) between collision frequency be M _1k(n ₁, n _k), by with each track P _k(n _k) Probability p (P _k(n _k)) multiply by M _1k(n ₁, n _k) and with all trajectory set { P _k(n _k) the element addition obtain degree of disturbance r ₁(n ₁, value k).Gained and main vehicle O just ₁Track P ₁(n ₁) and other vehicles O _kThe collision probability of the track collision that can follow.Therefore, finally obtain as degree of disturbance r ₁(n ₁, value k) and main vehicle O ₁Track P ₁(n ₁) and other vehicles O _kThe collision probability of collision increases pro rata.

After the step S413, whether degree of disturbance calculating part 106 carries out the judgment processing of repetitive cycling 2.If also remaining have should carry out and main vehicle O ₁Another vehicle O of estimating of interference _k(being "No" in step S414), then degree of disturbance calculating part 106 adds 1 (step S415) and repetitive cycling 2 after being back to step S402 with the value of k.On the other hand, if there be not remaining should the execution and main vehicle O ₁The vehicle O that estimates of interference _k(being "Yes" in step S414), then degree of disturbance calculating part 106 proceeds to step S416 subsequently.

At step S416, whether carry out the judgment processing of repetitive cycling 1.More specifically, if main vehicle O ₁Trajectory set { P ₁(n ₁) in remaining have should carry out the track of disturb estimating (is "No" at step S416), then degree of disturbance calculating part 106 is with n ₁Value add 1 (step S417) and repetitive cycling 1 after being back to step S401.On the other hand, if main vehicle O ₁Trajectory set { P ₁(n ₁) in do not had remainingly should carry out the track of disturb estimating (is "Yes" at step S416), then degree of disturbance calculating part 106 stopped circulation 1 (step S34) before stopping the degree of disturbance computing.

Then, deferent segment 107 according to the degree of disturbance output information that calculates at step S34 as evaluation result (step S35).To illustrate that below the display part 171 in the deferent segment 107 passes through as shown in Figure 7 at main vehicle O ₁Windshield F on stacked and carry out the situation of translucent demonstration.Fig. 7 can be interpreted as a kind of following situation that illustrates in this situation, that is, and by main vehicle O ₁With another vehicle O ₂Space-time environment Env (the P that constitutes ₁, P ₂) in, demonstrate at main vehicle O by stacked ₁According to main vehicle O ₁With other vehicles O ₂Between degree of disturbance r ₁(n ₁, 2) and degree of disturbance r in the route that can on two dimensional surface, take ₁(n ₁, 2) value surpass the zone of preset threshold value.

Display part 171 has according to degree of disturbance r ₁(n ₁, 2) value change the function of illumination.For example, if set with degree of disturbance r ₁(n ₁, 2) value increase and increase illumination, then in region D shown in Figure 7 _aAnd D _bIn, region D _aIllumination bigger.In this situation, main vehicle O ₁The driver can be by recognize immediately with reference to the stacked demonstration on the Windshield F when driving can be by taking towards degree of disturbance r eyeing to the front ₁(n ₁, 2) the less relatively region D of value _bRoute drive to avert danger.By this recognition results is reflected in the driver behavior, the driver can accurately avoid may occurring in the near future main vehicle O ₁On danger.

Except that display part 171 shows, alarm tone send portion 172 can at the resulting degree of disturbance r of the cooresponding desired path of current operation ₁(n ₁, 2) the value sound (comprising sound) that gives the alarm when surpassing predetermined threshold.

Equally, with the same in first embodiment, the function that the interference evaluation result of degree of disturbance calculating part 106 can have display part 171 by the read-out CN (see figure 8) that makes auto-navigation system shows.

Above-mentioned the 3rd embodiment according to the present invention, have the position of storing a plurality of objects at least and comprise that the computing machine of memory cell of internal state of the speed of each object reads the position and the internal state of a plurality of objects from memory cell, the time aerial generation track that the variation of the position that can take according to each object in a plurality of objects of passing in time based on the position of the object that is read and internal state is being made up of time and space, and by utilizing the route of a plurality of objects of track probability ground prediction that generate, and connect and calculate the degree of disturbance represent the annoyance level between the track that track that specified object can be followed and other objects can follow quantitatively based on predicting the outcome, in the situation of the actual generation of meeting, also can guarantee safety even make.

Equally, according to the 3rd embodiment, by applications exploiting the time aerial collision probability definition degree of disturbance, probability with other object collisions can calculate to a nicety in feasible period.

Further, according to the 3rd embodiment, be formed on the route prediction that the time aerial space-time environment of being made up of time and space is made object by utilization, not only can accurately make the route prediction of stationary object, and can accurately make the route prediction of dynamic object.

In addition, according to the 3rd embodiment,, specified object (for example, main vehicle) and other objects can be made a distinction owing to generate the track of detected object independently.So, can be easily and calculate to a nicety and be hidden in danger between specified object and other objects.

Have, formula is used to increase degree of disturbance r in (5) again ₁(n ₁, the coefficient c of value k) _1kCan not constant.For example, coefficient c _1kCan be main vehicle O ₁With other vehicles O _kBetween in the size of the relative velocity in when collision.Usually, when the size of relative velocity increased, the impact during the collision increased.Therefore, if coefficient c _1kBe the size of the relative velocity when collision between the vehicle, the shock extent that collides between the vehicle will be increased to degree of disturbance r ₁(n ₁, k) on.

Perhaps, the value branch of expression well damage degree can be tasked coefficient c _1kIn this situation, for example, task coefficient c at the storing value branch that reads from memory paragraph 108 _1kBefore, can estimate the damage grade evaluation value of the damage grade that causes of collision or the size of the loss from spoilage amount that caused by the collision relative velocity between the vehicle when colliding is stored in memory paragraph 108 accordingly in the numerical value mode with being used for.If sensor segments 3 have can the inspected object type function, can determine to damage grade evaluation value or loss from spoilage amount according to object type.In this situation, for example, when the object of collision with it when being people or vehicle, preferably by for example being used for the coefficient c with people's collision _1kValue be set at much larger than the coefficient c that is used for other object collisions _1kValue and will drop to minimum with people's possibility of collision.

Have, in the 3rd embodiment, degree of disturbance can be kept main vehicle O by supposition again ₁Operation and calculate.In this situation, the display part 171 in the deferent segment 107 not only can show main vehicle O ₁The prediction route, and can be presented at the danger of other vehicles that move in the preset range according to the result of calculation of degree of disturbance.

As previously discussed, keep the pattern of the operation of main vehicle O1, can simplify prediction situation during object when for example having around a lot, thereby the calculated amount that track is generated in section, prediction section and the degree of disturbance calculating part reduces by application.

Here, non-patent literature 1 that explanation is quoted in the above-mentioned background technology and the difference between the 3rd embodiment.Since in the 3rd embodiment under the situation of the correlativity of not considering object at each object independently execution path look for, calculated amount is less than non-patent literature 1.Particularly, the number of times that calculates the degree of disturbance of every track in the 3rd embodiment passes through

[formula 6]

${\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="H2007800070159E00031.png,H2007800070159E00101.png"\; state="\{\{state\}\}">n</figure-callout>}}_1\&times;\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}{n}_k$

Draw and, approximately the track number square calculated amount just enough have nothing to do with the object number that constitutes space-time environment.On the contrary, when disturbing according to 1 pair of non-patent literature when estimating, specified object (main vehicle) is not distinguished with other objects (other vehicles) and be used for the calculated amount (corresponding with the number of times of first embodiment calculating degree of disturbance) that mutual interference is mutually estimated is passed through

[formula 7]

${\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="H2007800070159E00031.png,H2007800070159E00101.png"\; state="\{\{state\}\}">n</figure-callout>}}_1\&times;\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}{n}_k$

Draw, therefore, need the calculated amount of K power of about track number.Thus, when the object number that constitutes space-time environment increases, and the difference of the calculated amount between the 3rd embodiment will be bigger significantly.

Except that above-mentioned difference, the difference between non-patent literature 1 and the 3rd embodiment is with identical at the described difference of first embodiment.Therefore, and in the same manner, realize the 3rd embodiment from non-patent literature 1 even also be difficult to for a person skilled in the art at first embodiment.

(the 4th embodiment)

Four embodiment of the invention is characterised in that by utilizing minimum collision time between main vehicle and other vehicles to define to estimate the degree of disturbance when disturbing.Identical according to the functional structure of the object route prediction unit of the 4th embodiment with object route prediction unit 101 (seeing Figure 12) according to the 3rd embodiment.Except that the degree of disturbance computing, identical with the 3rd embodiment according to the object route Forecasting Methodology of the 4th embodiment.

Figure 19 is the diagram of circuit that illustrates by the details of the degree of disturbance computing of carrying out according to the object route Forecasting Methodology of the 4th embodiment (corresponding with step S34 among Figure 13).In the 4th embodiment, also suppose object O ₁It is main vehicle.For convenience of explanation, with other objects O _k(k=2,3 ..., K) all be assumed to and also be four wheeler and be called other vehicles O _k

Degree of disturbance calculating part 106 at first begins main vehicle O ₁The reprocessing (circulation 1) (step S421) of all tracks.Based on this purpose, from trajectory set { P ₁(n ₁) in select a track and selected track carried out subsequent treatment.

Next, degree of disturbance calculating part 106 begins other vehicles O _kReprocessing (circulation 2) (step S422).In this circulation 2, the counting machine k that will be used to discern other vehicles is initialized as k=2 and when finishing reprocessing the value of k is increased progressively at every turn.

Degree of disturbance calculating part 106 is carried out the trajectory set { P that generates at step S33 _k(n _k) all elements reprocessing (circulation 3) and also carry out other vehicles O _kReprocessing (step S423).In this reprocessing, will be used for repetitive cycling 1 that is be used to be identified as main vehicle O ₁The counting machine n of the track that generates ₁With the counting machine N that is used to discern other vehicles ₁The degree of disturbance of determining is set at r ₁(n ₁, k) and with track rise time T be set at r ₁(n ₁, value k) (step S424).

Subsequently, the 106 beginning reprocessings (circulation 4) of degree of disturbance calculating part are to estimate main vehicle O ₁Track P ₁(n ₁) and other vehicles O _kTrack P _k(n _k) between interference (step S425).In this circulation 4, successively moment t=0, Δ t ..., T determines track P ₁(n ₁) and track Pk (n _k) between in the distance of synchronization, to judge main vehicle O ₁With other vehicles O _kWhether collide.For this judgement, the definition of collision is with identical in the 3rd embodiment and at main vehicle O ₁With other vehicles O _kBetween distance make the judgement that collision has taken place in short-term than interference distance.

If to main vehicle O ₁With other vehicles O _kBetween the judged result of distance be that degree of disturbance calculating part 106 is judged main vehicle O ₁With other vehicles O _kCollide (is "Yes" at step S426), and degree of disturbance r ₁(n ₁, value k) then is set at r with t greater than at the time in this moment t (from described initial position to required time of collision) (is "Yes" at step S427) ₁(n ₁, value k) and then t is set at T (step S428).Therefore, in this situation, circulation 4 stops (is "Yes" at step S429).

On the contrary, if main vehicle O ₁With other vehicles O _kCollision (is "Yes" at step S426) and degree of disturbance r ₁(n ₁, value k) is equal to or less than at the time in this moment t (is "No" at step S427), and then degree of disturbance calculating part 106 proceeds to step S429 to judge whether to stop circulation 4.As main vehicle O ₁With other vehicles O _kWhen not colliding (is "No" at step S426), degree of disturbance calculating part 106 also proceeds to step S429.

If also do not reach T at step S429 time t, repetitive cycling 4 (is "No" at step S429) then.In this situation, degree of disturbance calculating part 106 increases the value of t Δ t (step S430) and be back to step S425 before repetitive cycling 4.On the other hand, if reached T at step S429 time t, then degree of disturbance calculating part 106 stops circulation 4 (are "Yes" at step S429).

Under the situation of the reprocessing of above-mentioned circulation 4, degree of disturbance r ₁(n ₁, value k) will be minimum collision time, this minimum collision time is at main vehicle O ₁With other vehicles O _kBetween in the collision that takes place from described initial position to the required shortest time of collision.

After stopping circulation 4, whether degree of disturbance calculating part 106 carries out the judgment processing of repetitive cycling 3.That is, if be other vehicles O _kThere are any and main vehicle O in the track that generates ₁A track P ₁(n ₁) the also unenforced track of interference evaluation (being "No" in step S431), with n _kBe incremented to n _k+ 1 (step S432) and after being back to step S423 repetitive cycling 3.

On the other hand, if promising other vehicles O _kThe track and the main vehicle O that generate ₁A track P ₁(n ₁) the interference evaluation all carry out (in step S431 for "Yes"), then with other vehicles O _kA track P _k(n _k) the interference evaluation finish.Therefore, in this situation, degree of disturbance calculating part 106 is assigned the main vehicle O of evaluation ₁Track P ₁(n ₁) and other vehicles O _kAll tracks between the resultant interference degree r of interference ₁(n ₁, k) (step S433) and the output value of being assigned are to be stored in it (step S434) in memory paragraph 108.

Following step S435 to S438 relates to repetitive cycling 2 and circulation 1 judgment processing and identical with the step S414 to S417 that is used at the degree of disturbance computing of the 3rd embodiment explanation.

According to the invention described above the 4th embodiment, have the position of storing a plurality of objects at least and comprise that the computing machine of memory cell of internal state of the speed of each object reads the position and the internal state of a plurality of objects from memory cell, the time aerial generation track that the variation of the position that can take according to each object in a plurality of objects of passing in time based on the position of the object that is read and internal state is being made up of time and space, by utilizing the route of a plurality of objects of track probability ground prediction that generate, and based on prediction result, calculate the degree of disturbance be illustrated in the annoyance level between the track that specified object can be followed in the space-time track and other objects can follow quantitatively, in the actual situation that takes place of meeting, also can guarantee safety even make.

Equally, according to the 4th embodiment, by the degree of disturbance of the minimum collision time definition of applications exploiting, possibility with other object collisions can calculate to a nicety in feasible period.

(the 5th embodiment)

Fifth embodiment of the invention is characterised in that by adding up to the main vehicle that obtains in the mode identical with the 3rd embodiment and the result of the degree of disturbance between other vehicles to estimate main vehicle and the interference between the space-time environment on every side.Identical according to the functional structure of the object route prediction unit of the 5th embodiment with object route prediction unit 101 (seeing Figure 12) according to the 3rd embodiment.Except that the degree of disturbance computing, identical with the 3rd embodiment according to the object route Forecasting Methodology of the 5th embodiment.

Figure 20 is the diagram of circuit that the details of the degree of disturbance computing of carrying out according to the object route Forecasting Methodology of fifth embodiment of the invention (corresponding with step S34 among Figure 13) is shown.Degree of disturbance calculating part 106 at first begins main vehicle O ₁The reprocessing (circulation 1) (step S441) of all tracks.Based on this purpose, from trajectory set { P ₁(n ₁) in select a track and selected track carried out subsequent treatment.

In the 5th embodiment, to other vehicles O _kReprocessing (circulation 2), to the trajectory set { P of other vehicles _k(n _k) all elements reprocessing (circulation 3) and be used to estimate main vehicle O ₁Track P ₁(n ₁) and other vehicles O _kTrack P _k(n _k) between the reprocessing (circulation 4) of interference identical with the 3rd embodiment.That is the step S402 to S415 that illustrates in the processing of step S442 to S455 shown in Figure 20 and degree of disturbance computing at the 3rd embodiment is identical.

After the reprocessing of circulation 2 stopped, degree of disturbance calculating part 106 was according to other vehicles O _kTo from circulate 2 to the circulation the 4 degree of disturbance r that obtain ₁(n ₁, k) assign flexible strategy α (k) (〉 0), by:

[formula 8]

$R_1\left(n_1\right)=\underset{k=2}{\overset{K}{\mathrm{\&Sigma;}}}\mathrm{\&alpha;}\left(k\right)r_1({\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="H2007800070159E00031.png,H2007800070159E00101.png"\; state="\{\{state\}\}">n</figure-callout>}}_1,k)---\left(6\right)$

Calculate total degree of disturbance as the summation of these degree of disturbances and export its result of calculation it is stored in (step S456) in the memory paragraph 108.The value of flexible strategy α (k) can all equate and be constant (for example, 1), maybe can assign basis such as other vehicles O _kThe value of conditions such as type.Thus, can be to main vehicle O ₁Track P ₁(n ₁) and comprise every other vehicle O ₂..., O _kIntegrated environment between interference estimate.

Total degree of disturbance R ₁(n ₁) can pass through:

[formula 9]

$R_1\left(n_1\right)=\underset{k}{\max }\left(\mathrm{\&alpha;}\left(k\right)r_1({\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="H2007800070159E00031.png,H2007800070159E00101.png"\; state="\{\{state\}\}">n</figure-callout>}}_1,k)\right)---\left(7\right)$

Definition.In this situation, the most dangerous object O _kDegree of risk will be treated to total degree of disturbance.If according to the definition of formula (6), then ought for example main vehicle O ₁Disturb mutually with the minority object but low with most total degree of disturbances that remaining object is irrelevant might to be calculated when disturbing.Therefore, even wherein because the close main vehicle O of minority vehicle ₁The fortune people to exercise is feeling the situation of danger close instinctively, might this situation be judged as safety with counter-intuitive ground.On the other hand, estimate, can reduce the possibility of making as previously discussed with the judgement of counter-intuitive by carrying out to disturb based on the definition of formula (7) etc.

Subsequently, whether degree of disturbance calculating part 106 carries out the judgment processing of repetitive cycling 1.That is, if main vehicle O ₁Trajectory set { P ₁(n ₁) in also remaining have should carry out the track of disturb estimating (is "No" at step S457), then degree of disturbance calculating part 106 is with n ₁Value add 1 (step S458) and repetitive cycling 1 after being back to step S411.On the other hand, if main vehicle O ₁Trajectory set { P ₁(n ₁) in should not carry out the track of disturb estimating (is "Yes" at step S457), then degree of disturbance calculating part 106 stopped circulation 1 (step S34) before stopping the degree of disturbance computing.

Figure 21 is the schematically illustrated chart of having used according to the structure of the space-time environment of the object route Forecasting Methodology of the 5th embodiment.Space-time environment Env (P shown in the figure ₁, P ₂, P ₃) illustrate with respect to main vehicle O ₁The situation that has two other vehicles.Main vehicle O ₁Track P ₁(n ₁) represent the second vehicle O with solid line ₂Track P ₂(n ₂) be represented by dotted lines the 3rd vehicle O ₃Track P ₃(n ₃) represent with thick line.By utilizing and space-time environment Env (P ₁, P ₂, P ₃) total degree of disturbance R ₁(n ₁) carry out to disturb and estimate, rather than the separate processes and the second vehicle O ₂Degree of disturbance r ₁(n ₁, 2) and with the 3rd vehicle O ₃Degree of disturbance r ₁(n ₁, 3), can avoid main vehicle O according to surrounding environment ₁Danger.

Above-mentioned the 5th embodiment according to the present invention, have the position of storing a plurality of objects at least and comprise that the computing machine of memory cell of internal state of the speed of each object reads the position and the internal state of a plurality of objects from memory cell, the time aerial generation track that the variation of the position that can take according to each object in a plurality of objects of passing in time based on the position of the object that is read and internal state is being made up of time and space, by utilizing the route of a plurality of objects of track probability ground prediction that generate, and based on prediction result, calculate the degree of disturbance represent the annoyance level between the track that track that specified object can be followed and other objects can follow quantitatively, in the situation of the actual generation of meeting, also can guarantee safety even make.

According to the 5th embodiment,, when the object quantity that constitutes described space-time environment is big, also can accurately carry out to disturb and estimate by utilizing total degree of disturbance.

Have again, in the 5th embodiment, as degree of disturbance r ₁(n ₁, k) since collision also can adopt when increasing to the 3rd embodiment in the similar various definition arbitrarily definition as coefficient c _1kOr the value of F (t).Degree of disturbance r ₁(n ₁, 3) also can with the 4th embodiment in similarly be defined as minimum collision time.

In addition, in the 5th embodiment, carry out and disturb when estimating, can be in conjunction with total degree of disturbance R ₁(n ₁) and single degree of disturbance r ₁(n ₁, k) disturb evaluation.

(the 6th embodiment)

Figure 22 illustrates by the block diagram of utilization according to the functional structure of the automatic operation system of the object route prediction unit formation of sixth embodiment of the invention.Automatic operation system 1000 shown in the figure is installed in such as on the movable body of four wheeler and comprise: object route prediction unit 201 is used for can and can setting the route that will be taked by main vehicle by the route that other objects (comprising vehicle, people and obstacle) are taked by the route of taking as the main vehicle of specified object by prediction; And actuating device 211 is used for operating the route that main vehicle realizes that object route prediction unit 201 is set according to actuated signal.

Object route prediction unit 201 comprises the input section 2 of importing various information from the outside, be used to detect the position that is present in the object in the preset range and the sensor segments 3 of internal state, the variation that is used for the position that can take according to the object of passing in time based on the testing result of sensor segments 3 generates section 4 at the time aerial track that generates track that is being made of time and space, be used to utilize track to generate section 4 tracks that generate are made the probability prediction to the route of object a prediction section 105, be used for degree of disturbance calculating part 106 based on the degree of disturbance of predicting the annoyance level between the route that section route that 5 calculating of making that predict the outcome represent that quantitatively main vehicle can be taked and other objects can be taked, be used for selecting the route selection portion 207 of the route that main vehicle should take according to the degree of disturbance that degree of disturbance calculating part 106 calculates, be used for behind the cooresponding actuated signal of result that generates the selection made from route selection portion 207, actuated signal being transferred to the actuated signal transport part 208 of actuating device 211, be used for deferent segment 209 with the various information outputs relevant with the processing of object route prediction unit 201 execution, and the memory paragraph 210 that is used to store the various information of the position that comprises sensor segments 3 detected objects and internal state.Among Figure 22, authorize identical reference number to having with assembly according to the object route prediction unit 101 identical functions structures of the 3rd embodiment shown in Figure 12.

Deferent segment 209 has: display part 291 is used for the information relevant with the processing of prediction section 105, degree of disturbance calculating part 106 and 207 execution of route selection portion is shown/be output as the information that comprises image; And alarm tone sends portion 292, is used for according to predicting the outcome of making of prediction section 105 or result of calculation that degree of disturbance calculating part 106 the is made sound that gives the alarm.

Except that the testing result of sensor segments 3, memory paragraph 210 also route selection result, the track of the execution of degree of disturbance result of calculation, the route selection portion 207 of storage track predicting the outcome of generating that tracks, prediction section 105 that section 4 generates make, degree of disturbance calculating part 106 generates operation that the operation selection portion 41 in the section 4 selects etc.

Next, with the object route Forecasting Methodology of explanation according to the 6th embodiment.Figure 23 is the diagram of circuit that illustrates according to the overview of the processing of the object Forecasting Methodology of the 6th embodiment.In the 6th embodiment, object also that all are to be predicted all is assumed to move on two dimensional surface and describes.

In the 6th embodiment, the position of each object and the detection of internal state handle (step S61), the time aerial each object track generate and handle (step S62), utilize track the probability prediction processing (step S63) of object route and identical with the step S31 that in the 3rd embodiment, illustrates, step S32, step S33 and step S34 respectively to the degree of disturbance computing (step S64) between main vehicle and other vehicles based on predicting the outcome.

To be described in detail in the processing of step S65 and subsequent step thereof below.At step S65, carry out according to the route selection of the degree of disturbance that in the degree of disturbance computing of step S64, calculates and handle (step S65).In automatic operative technique such as the movable body that in wide region, moves of automobile, as previously discussed, except that not considering that wherein the influence or the not calculative in practice route of its influence of other dynamic barriers are looked for the technology at least, also need a kind of route calculation technology, in the time of practicality, realize avoiding colliding required calculating when needs move, to avert danger with dynamic barrier by this technology.In the 6th embodiment, in handling, route selection utilize two evaluation numbers to avert danger.First evaluation number is the degree of disturbance that calculates in the degree of disturbance computing and handles by utilizing degree of disturbance to carry out first via line options.If the result who selects a plurality of routes to handle as first via line options utilizes second evaluation number that is stored in the memory paragraph 210 to carry out second route selection and handles.In second route selection is handled, except judging whether route has along the choice criteria of the one-tenth branch foundation in the path that arrives the destination, preferably by the additional choice criteria of combination (explanation in the back) suitably as second evaluation number with further restriction route.

Trajectory set by estimating other vehicles and have between the trajectory set of main vehicle of different mutually end points degree of disturbance and as previously discussed according to the degree of disturbance selection schemer of being estimated out, accurately avert danger with in main vehicle ' the time and feasiblely can guarantee safety (seeing Figure 15) at the situation that not have the default position that will arrive such as main vehicles such as the destinations best route that judges at every turn.Thus,, the predeterminated position of main vehicle on road also can solve problem shown in Figure 14 even travelling, that is, do not guarantee the fatal problem of safety.

Figure 24 is the diagram of circuit that the details of route selection processing is shown.Among Figure 24, route selection portion 207 is chosen in the degree of disturbance r that calculates in the degree of disturbance computing ₁(n ₁, the route (step S501) of value minimum k).

If owing to selected degree of disturbance r ₁(n ₁K) the feasible only remaining track (being "No" in step S502) of Zui Xiao track, then route selection portion 207 reads with the record of the cooresponding position of selected track (x (t), y (t)) and the sequence of operation when t=0 to T { u (t) } and with them from memory paragraph 210 and exports actuated signal transport part 208 (step S504) to.On the contrary, if owing to selected degree of disturbance r ₁(n ₁, k) the feasible also remaining a plurality of tracks (is "Yes" at step S502) that have of Zui Xiao track, then route selection portion 207 proceeds to step S503).

At step S503, track (step S503) with additional choice criteria optimum matching is selected by utilizing route and comprise the choice criteria of the assembly institute foundation along route to the destination and set in advance and be stored in additional choice criteria in the memory paragraph 210 by route selection portion 207 from a plurality of tracks of selecting at step S501.The condition enactment that may have repetition values in track hardly can be additional choice criteria.

Provide several examples of additional choice criteria below:

(1) at the main vehicle O of (behind Δ t) after the operation ₁Each position in, at the position at the center in the track that approaching main vehicle travelled (the x coordinate in Fig. 4 etc.) on the road width direction.In this situation, be chosen in the track of the position that operation is expected most on the road.If selected during t=0 to T, to take the track of stable position, can select to make in each operation back of t=0 to T along the position and track minimum of road width direction.

(2) at the main vehicle O of (behind Δ t) after the operation ₁Each position in, the position maximum on travel direction (in Fig. 4 etc. is the y coordinate direction).In this situation, select the fastest track.Perhaps, the track of position maximum in the time of can being chosen in t=T.

(3) acceleration/accel is minimum in the size of initial time (t=0).In this situation, select to quicken pulsation-free track.Quicken pulsation-free track in the sequence of operation of t=0 to T if be chosen in, select size and track minimum of each the operation post-acceleration that can select to make when t=0 to T.

(4) cireular frequency is minimum in the size of initial time (t=0).In this situation, select to turn to pulsation-free track.Turn to pulsation-free track in the sequence of operation of t=0 to T if be chosen in, then similar to (3), can select to make size and track maximum at each operation back cireular frequency of t=0 to T.

Handle according to above-mentioned route selection, by selecting most possibly to avoid main vehicle O ₁Danger track as the time aerial track, thus, selected main vehicle O ₁The route that should on actual two dimensional surface, take.

If at the only remaining track of step S503, route selection portion 207 proceeds to above-mentioned steps S504.Even if after using additional choice criteria still remaining a plurality of tracks, for example can set and make automatic gated counter n ₁Or the value of k is got minimum or peaked that track.

After above-mentioned route selection is handled, actuated signal transport part 208 generates and the corresponding position of the track (x (t) that exports according to the selection result of step S65, y (t)) record and corresponding to the actuated signal of the sequence of operation { u (t) } of t=0 to T, and they are transferred to actuating device 211 (step S66).

The structure that depends on actuating device 211 by actuated signal transport part 208 at the actuated signal of step S66 generation and transmission.For example, if actuating device 211 is a kind of mechanicals device, such as steering gear, accelerator and drg, then the record of the 208 exportable positions that receive from route selection portion 207, actuated signal transport part (x (t), y (t)) and operation { u (t) } are directly as actuated signal.

On the contrary, if actuating device 211 is a kind of being used for increase the device of operation moment of torsion such as the mechanical device of steering gear, accelerator and drg, the record of the position (x (t), y (t)) that receives from route selection portion and operation { u (t) } can be transmitted directly as operation moment of torsion actuated signal or will increase to this mechanical device during operation after calculating in actuated signal transport part 208.In the previous case, the operation moment of torsion will calculate in actuating device 211 sides.

If actuating device 211 is a kind of devices that are used to increase the operation moment of torsion, and the driver can be by increasing than the big operation goes through torque conversion of the operation moment of torsion of actuating device 211 to M/C, that is, this device can be by driver's override, and then automatic operation system 1000 accessory device that is also applicable as operation makes the operation that can realize reflecting driver's purpose when selection schemer.

Have again, detect main vehicle O by utilizing sensor segments 3 ₁The ground-surface condition of being travelled can be based on controlling actuating device 211 according to the feedback of pavement conditions.

Figure 25 be illustrated in route selection result in the display part 291 of deferent segment 209 etc. the demonstration output example chart and be to be shown schematically in by main vehicle O ₁With another vehicle O ₂Space-time environment the Env ' (P that constitutes ₁, P ₂) show the chart of output example when carrying out route selection in the (see figure 6).More specifically, Figure 25 illustrates a kind of following situation, wherein, and by route being set with stacked demonstration of translucent arrow H on Windshield F, and at main vehicle O ₁According to main vehicle O ₁With other vehicles O ₂Between degree of disturbance r ₁(n ₁, 2) and degree of disturbance r in the route that can on two dimensional surface, take ₁(n ₁, 2) value surpass the zone of preset threshold value also by being stacked in object O ₁The Windshield F of (main vehicle) goes up and shows translucently.The schematically illustrated a kind of following situation of situation shown in Figure 25, wherein, actuating device 211 makes bearing circle ST produce the operation moment of torsion, makes operation moment of torsion that bearing circle ST is rotated in a clockwise direction to take to be provided with route thereby produced.

Two zones have been shown among Figure 25, region D also translucently _aAnd region D _bHas different illumination (region D, here _aBrighter).Difference on this illumination and degree of disturbance r ₁(n ₁, 2) the corresponding and described illumination of the difference of value mean if selects close region D _aRoute degree of disturbance r then ₁(n ₁, 2) value increase.From this meaning, Figure 25 visually represents to take to lead to degree of disturbance r ₁(n ₁, 2) the less region D of value _bRoute can make that driving averts danger travels.

Except showing by display part 291, alarm tone sends portion 292 can be at the resulting degree of disturbance r of route according to current behaviour's expection ₁(n ₁, 2) the value sound (comprising sound) that gives the alarm when surpassing predetermined threshold.

The demonstration output example of deferent segment 209 is not limited thereto, and the function that for example can have display part 291 by the read-out CN (seeing Figure 26) that makes auto-navigation system shows and route to be set and to disturb evaluation result.In this situation, as shown in figure 26, the described route that is provided with shows and two zones---region D by arrow H _aAnd region D _b---between the difference of degree of disturbance also show by each regional color gradient on the two dimensional surface that shows on the read-out CN.

Above-mentioned the 6th embodiment according to the present invention, have the position of storing a plurality of objects at least and comprise that the computing machine of memory cell of internal state of the speed of each object reads the position and the internal state of a plurality of objects from memory cell, the time aerial generation track that the variation of the position that can take according to each object in a plurality of objects of passing in time based on the position of the object that is read and internal state is being made up of time and space, by utilizing the route of a plurality of objects of track probability ground prediction that generate, based on prediction result, calculate the degree of disturbance represent the annoyance level between the track that track that specified object can be followed and other objects can follow quantitatively, and, in the actual situation that takes place of meeting, also can guarantee safety in utility time even make according to the route that the degree of disturbance that calculates selects designated vehicle to take.

Equally, according to the 6th embodiment, by between applications exploiting specified object and other objects the time aerial collision probability definition degree of disturbance and when described the aerial track of selecting the degree of disturbance minimum, can from the route that specified object can be taked, accurately set the minimum route of probability with other object collisions.

Further, according to the 6th embodiment, by utilizing the route prediction of the object of making, not only can accurately make the route prediction of stationary object, and can accurately make the route prediction of dynamic object at the time aerial space-time environment of forming by time and space that forms.

In addition, according to the 6th embodiment,, specified object (for example main vehicle) and other objects can be made a distinction because the track of detected object is independent the generation.Thus, can be easily and calculate to a nicety and to be hidden in danger between described specified object and other objects.

Have, in the 6th embodiment, formula is used to increase degree of disturbance r in (5) again ₁(n ₁, the coefficient c of value k) _1kCan not constant, and coefficient c for example _1kCan be main vehicle O ₁With other vehicles O _kBetween in the size of the relative velocity in when collision.Perhaps, the value branch of expression well damage degree can be tasked coefficient c _1k

In addition, in the 6th embodiment, degree of disturbance can be kept main vehicle O by supposition ₁Operation and calculate.Thereby, can situation be simplified during object around many predicting for example having, calculated amount reduces in section, prediction section, degree of disturbance calculating part and the route selection portion thereby generate at track.

(the 7th embodiment)

Seventh embodiment of the invention is characterised in that the route by adding up to the main vehicle (specified object) that obtains in the mode identical with the 6th embodiment and the degree of disturbance result of calculation between other vehicles to estimate main vehicle and the interference between the space-time environment on every side and select main vehicle to take based on evaluation result.According to the 7th embodiment the functional structure of object route prediction unit identical with functional structure according to the object route prediction unit 201 (seeing Figure 22) of the 6th embodiment.Except that the degree of disturbance computing, according to the object route Forecasting Methodology of the 7th embodiment with identical according to the method for the 6th embodiment.

Figure 27 is the diagram of circuit that the details of the degree of disturbance computing of carrying out according to the object route Forecasting Methodology of the 7th embodiment (corresponding with step S64 among Figure 23) is shown.Degree of disturbance calculating part 106 at first begins main vehicle O ₁The reprocessing (circulation 1) (step S461) of all tracks.Based on this purpose, from trajectory set { P ₁(n ₁) in select a track and selected track carried out subsequent treatment.

Then, degree of disturbance calculating part 106 is identical with the processing of the step S422 to S432 of the 4th embodiment in the processing that step S462 to S472 carries out.Therefore, processing at step S473 and subsequent step thereof will be described below.

Circulation 3 repeat to finish the time (is "Yes" at step S471) execution in step S473, that is, carry out with respect to other vehicles O _kMain vehicle O in the track that generates ₁A track P ₁(n ₁) all disturb to estimate.With this step S473, degree of disturbance calculating part 106 will be finished other vehicles O _kA track P _k(n _k) the interference evaluation.Therefore, in this situation, 106 fens party masters of degree of disturbance calculating part vehicle O ₁Track P ₁(n ₁) and other vehicles O _kAll tracks between the degree of disturbance r that estimates of interference ₁(n ₁, k) (step S473) and with the value of being assigned output it is stored in (step S474) in the memory paragraph 210.

Then, whether degree of disturbance calculating part 106 carries out the judgment processing of repetitive cycling 2.If also remaining have should carry out and main vehicle O ₁Another vehicle O of estimating of interference _k(is "No" at step S475), degree of disturbance calculating part 106 adds 1 (step S476) and repetitive cycling 2 after being back to step S462 with the value of k.On the other hand, if there be not remaining should the execution and main vehicle O ₁The vehicle O that answers of interference evaluation _k(is "Yes" at step S475), degree of disturbance calculating part 106 proceeds to subsequent step S477 after the repetition of finishing circulation 2.

At step S477, degree of disturbance calculating part 106 utilizes the degree of disturbance r that obtains to the circulation 4 in circulation 2 ₁(n ₁, k) total degree of disturbance R of providing of computing formula (6) ₁(n ₁) and export result of calculation it is stored in (step S477) in the memory paragraph 210.Have again, also can adopt formula (7) as total degree of disturbance R ₁(n ₁).

Subsequently, whether degree of disturbance calculating part 106 carries out the judgment processing of repetitive cycling 1.That is, if main vehicle O ₁Trajectory set { P ₁(n ₁) in should carry out the track of disturb estimating (is "No" at step S478) in addition, then degree of disturbance calculating part 106 is with n ₁Value add 1 (step S479) and repetitive cycling 1 after being back to step S461.On the other hand, if main vehicle O ₁Trajectory set { P ₁(n ₁) in should not carry out the track of disturb estimating (is "Yes" at step S478), then degree of disturbance calculating part 106 is stopping stopping circulation 1 (step S64) between the degree of disturbance computing.

Next, route selection processing (corresponding with step S65 among Figure 23) will be described.Figure 28 is the diagram of circuit that the details of the route selection processing of carrying out according to the object route Forecasting Methodology of the 7th embodiment is shown.Among Figure 28, route selection portion 207 is chosen in the total degree of disturbance R that calculates in the degree of disturbance computing ₁(n ₁) maximum track (step S511).

If selected the result of the track of degree of disturbance maximum a track (is "No" at step S512) to be arranged for only remaining, then route selection portion 207 reads and the cooresponding position of selected track (x (t) from memory paragraph 210, y (t)) record and in the sequence of operation of t=0 to T, and export them to actuated signal transport part 208 (step S514).On the contrary, be the also remaining a plurality of tracks (is "Yes" at step S512) that have if selected the result of the track of degree of disturbance maximum, then route selection portion 207 proceeds to step S513.

At step S513, route selection portion 207 is stored in default choice criteria in the memory paragraph 210 from select the track (step S513) with additional choice criteria optimum matching a plurality of tracks that step S511 selects by utilization.Can with to the 6th embodiment in similar condition enactment be additional choice criteria.

If at the only remaining track of step S513, then route selection portion 207 proceeds to above-mentioned steps S514.Even if after using additional choice criteria still remaining a plurality of tracks, then for example can set and make automatic gated counter n ₁Or the value of k is got minimum or peaked that track.

Handle according to above-mentioned route selection, by selecting most possibly to avoid theme vehicle O ₁Danger track as the time aerial track, thus, selected main vehicle O ₁The route that should on actual two dimensional surface, take.

After the route selection of step S65 is handled by identical in the processing (step S66) of actuated signal transport part 208 execution and the 6th embodiment.The processing that actuating device 211 is carried out after actuated signal transport part 208 receives actuated signals also with the 6th embodiment identical.

Above-mentioned the 7th embodiment according to the present invention, have the position of storing a plurality of objects at least and comprise that the computing machine of memory cell of internal state of the speed of each object reads the position and the internal state of described a plurality of objects from memory cell, the time aerial generation track that the variation of the position that can take according to each object in a plurality of objects of passing in time based on the position of the object that is read and internal state is being made up of time and space, by utilizing the route of the described a plurality of objects of track probability ground prediction that generate, calculate the degree of disturbance represent the annoyance level between the route that route that specified object can be followed and other objects can follow quantitatively based on prediction result, and the route of selecting designated vehicle to take according to the degree of disturbance that calculates in utility time equally with the 6th embodiment also can be guaranteed safety even make in the situation of meeting reality generation.

Equally, according to the 7th embodiment, assign total degree of disturbance that the degree of disturbance addition (being weighted the back addition) that will utilize minimum collision time definition after the flexible strategy obtains and the track of aerial total degree of disturbance maximum when being chosen in by being applied in to each object, even when the object quantity of forming space-time environment is big, also can accurately set and the minimum route of the possibility of other object collisions.

In the 7th embodiment, can similarly define degree of disturbance r with the 6th embodiment ₁(n ₁, k) and as degree of disturbance r ₁(n ₁, can adopt similar to the 3rd embodiment arbitrarily various definition as coefficient c when k) increasing owing to collision _1kOr the value of F (t).In this situation, in handling, route selection can select R ₁(n ₁) become minimum track.

In addition, in the 7th embodiment, carry out equally and disturb when estimating, can be with total degree of disturbance R ₁(n ₁) and independent degree of disturbance r ₁(n ₁, k) combination is to disturb evaluation.

(other embodiments)

So far, described first to the 7th embodiment in detail, still the invention is not restricted to these embodiments as implementing best mode of the present invention.For example, in object route Forecasting Methodology of the present invention, except that the object that detect to exist by sensor segments, can make route prediction by dummy object is set to set dummy object.More specifically, can carry out the route prediction by making up to show the dummy object model of unfavorable behavior and this object model is arranged on the desired location to main vehicle.Utilize this dummy object model, when to because for example existence of hovel and can not provide near the vehicle (main vehicle) that moves the four corners at broad visual angle to carry out route when prediction, can can't detected position make and to predict the danger such as object collision that may fly out from four corners by model being arranged on main vehicle.In advance the information storage relevant with the dummy object model can be arranged on the condition enactment of this information according to the input section on the position of expectation behind the memory paragraph.

When object route prediction unit according to the present invention is applied in such as the such field of the express highway of supposing operational vehicle only, by every vehicle being had be used for the communication device of car to car communication, nearby Yun Hang vehicle can be intercoursed operation conditions by car to car communication.In this situation, can select probability and will select the relevant information transfer of probability to other vehicles by the operating record of every vehicle of storage in the vehicle storage section and based on the operation of additional each operation of operating record with operation.Can avert danger more reliably thereby improve when the route accuracy of predicting makes operation.

Further, among the present invention, can adopt GPS (global positioning system) as position detection unit.In this situation, can be stored in the location information and the mobile message of the detected object of three-dimensional map information correction sensor segments among the described GPS by reference.The mutual communication of output that further, can be by GPS makes GPS have the function of sensor segments.In all situations, can realize the prediction of high precision route, further improve the reliability that predicts the outcome by adopting GPS.

The present invention is applicable to the object that moves in three dimensional space.The present invention also is applicable to the object (for example, the object as the manipulator with six-freedom degree) with a plurality of degree of freedom degree.

Can manifest from the above description, the present invention can comprise not at the various embodiments of this explanation and not depart under the prerequisite of scope of the clear and definite know-why of claims and can make various design developments.

Industrial usability

Be suitable for guaranteeing the technology of security as a kind of by averting danger when the movable body of driving such as four-wheel car according to object path prediction method of the present invention, device and program and automatic operation system.

Claims (57)

1. object route Forecasting Methodology that is used for by the route of computer forecast object, described computing machine has memory cell, and described memory cell is stored the position of described object at least and is comprised the internal state of the speed of described object, and described method comprises:

Variation in the position that can take according to the described object of passing in time based on the position of the described object that is read and internal state after position of reading described object from described memory cell and the internal state generates step at the time aerial track that generates track that is being made of time and space; And

By utilizing the prediction steps that generates the route of the described object of track probability ground prediction that generates in the step at described track, wherein

Described track generates step and comprises

Step is selected in the operation that is chosen in the operation that described object carries out from a plurality of operations;

Make the operation of in described operation selection step, selecting carry out the object control step of a scheduled time slot; And

Whether the described object in selected operation back is carried out in judgement in described object control step position and internal state satisfy the determining step of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein

Repeat one group of processing till reaching the track rise time that generates described track from described operation selection step to described determining step.

2. object route Forecasting Methodology as claimed in claim 1, wherein, described operation selects step to select probability to select operation according to the operation that each operation in described a plurality of operations is authorized, and

If judged result is that the position and the internal state of described object satisfies described controlled condition and described mobile condition in described determining step, then increases and be back to described operation after the described period and select step.

3. object route Forecasting Methodology as claimed in claim 2 wherein, is selected probability by utilizing random number to define described operation.

4. object route Forecasting Methodology as claimed in claim 2 wherein, preestablishes and will generate the track number that generates in the step at described track.

5. object route Forecasting Methodology as claimed in claim 1 wherein, if judged result is for satisfying described controlled condition and described mobile condition in described determining step, but is then carried out recursive call after the described period all selection operations is performed by increasing.

6. object route Forecasting Methodology as claimed in claim 1, wherein, the position and the internal state of a plurality of objects of described cell stores, and

Described track generates step aerial track that generates each object in described a plurality of objects when described.

7. object route Forecasting Methodology as claimed in claim 6, wherein, described prediction steps is specified an object and is calculated the aerial probability that exists when described of object except that specified object from described a plurality of objects.

8. object route Forecasting Methodology as claimed in claim 1 comprises that further output comprises the output step of the information that predicts the outcome in the described prediction steps.

9. object route Forecasting Methodology that is used for by the route of a plurality of objects of computer forecast, described computing machine has memory cell, described memory cell is stored the position of described a plurality of objects at least and is comprised the internal state of the speed of each object, and described method comprises:

The variation of the position that can take according to each object the described a plurality of object of passing in time based on the position of the described object that is read and internal state after position of reading described a plurality of objects from described memory cell and internal state generates step at the time aerial track that generates track that is being made of time and space;

By utilizing the prediction steps that generates the route of the described a plurality of objects of track probability ground prediction that generate in the step at described track; And

Calculate the degree of disturbance calculation procedure of degree of disturbance based on predicting the outcome in the described prediction steps, described degree of disturbance is represented the annoyance level between the route that route that specified object can be taked and other objects can take quantitatively, and wherein, described track generates step and comprises

Step is selected in the operation that is chosen in the operation that described object carries out from a plurality of operations;

Make the operation of in described operation selection step, selecting carry out the object control step of a scheduled time slot; And

Whether the described object in selected operation back is carried out in judgement in described object control step position and internal state satisfy the determining step of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein

Repeat one group of processing till reaching the track rise time that generates described track from described operation selection step to described determining step.

10. object route Forecasting Methodology as claimed in claim 9, wherein, described degree of disturbance calculation procedure

Lean on to such an extent that the value of the degree of disturbance between described specified object and each other object is increased or reduce specified amount than the nearer number of times of interference distance according to described specified object and each other object, described interference distance is the space length that object interferes with each other.

11. object route Forecasting Methodology as claimed in claim 10, wherein, described degree of disturbance calculation procedure

When described specified object and one of other objects lean on closelyer than described interference distance, and the probability of the two articles that draws closer together aerial collision when described increases the value of the degree of disturbance between the described two articles pro rata.

12. object route Forecasting Methodology as claimed in claim 10, wherein, described degree of disturbance calculation procedure

When described specified object and one of other objects lean on closelyer than described interference distance, and the value that increases the degree of disturbance between the described two articles of the relative velocity in moment of drawing closer together of the two articles that draws closer together with being in proportion.

13. object route Forecasting Methodology as claimed in claim 10, wherein, described memory cell will be used to estimate the size of the relative velocity when colliding between the damage grade evaluation value of the damage grade that collision causes or loss from spoilage amount that collision causes and the different object stores accordingly, and

Described degree of disturbance calculation procedure reads described damage grade evaluation value or described loss from spoilage amount in the size of the relative velocity in the moment that draws closer together from described memory cell according to two articles when described specified object and one of other objects lean on closelyer than described interference distance, and and described damage grade evaluation value or described loss from spoilage amount increase degree of disturbance between the described two articles pro rata.

14. object route Forecasting Methodology as claimed in claim 10, wherein, described degree of disturbance calculation procedure

To be the value of described degree of disturbance from the initial position of each object from the required described time set of described initial position at one of described specified object and other objects when required time is less than the value of the degree of disturbance the two articles that draws closer together when leaning on closelyer than described interference distance.

15. object route Forecasting Methodology as claimed in claim 10, wherein, described degree of disturbance calculation procedure

Value to each degree of disturbance between described specified object and other objects is weighted the back addition.

16. object route Forecasting Methodology as claimed in claim 9, wherein, described operation selects step to select probability to select operation according to the operation that each operation in described a plurality of operations is authorized, and

If judged result is that the position and the internal state of described object satisfies described controlled condition and described mobile condition in described determining step, then increases and be back to described operation after the described period and select step.

17. object route Forecasting Methodology as claimed in claim 16 wherein, is selected probability by utilizing random number to define described operation.

18. object route Forecasting Methodology as claimed in claim 16 wherein, preestablishes and will generate the track number that generates in the step at described track.

19. object route Forecasting Methodology as claimed in claim 9 further comprises the output step of output corresponding to the information of the degree of disturbance that calculates in the described degree of disturbance calculation procedure.

20. object route Forecasting Methodology as claimed in claim 9 further comprises

Select to be included in the route selection step of the route that the described specified object within described a plurality of object will take according to the degree of disturbance that in described degree of disturbance calculation procedure, calculates.

21. object route Forecasting Methodology as claimed in claim 20, wherein, when described degree of disturbance being defined as the collision probability of the track collision that track that described specified object can take and other objects can take, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just more little, and

Described route selection step is selected the route of degree of disturbance minimum.

22. object route Forecasting Methodology as claimed in claim 21, wherein, described route selection step

When being arranged, the route of a plurality of degree of disturbance minimums from described a plurality of routes, selects route with predetermined additional choice criteria optimum matching.

23. object route Forecasting Methodology as claimed in claim 20, wherein, when described degree of disturbance is defined as minimum collision time, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just big more, described minimum collision time is to collide the required shortest time from initial position in the collision that takes place between described specified object and described other objects, and

Described route selection step is selected the route of degree of disturbance maximum.

24. object route Forecasting Methodology as claimed in claim 23, wherein, described route selection step

When being arranged, the route of a plurality of degree of disturbance maximums from described a plurality of routes, selects route with predetermined additional choice criteria optimum matching.

25. object route Forecasting Methodology as claimed in claim 21 further is included in according to the record of the position of selected route in the described route selection step and the sequence of operation that is used to realize described route is transferred to exterior actuated signal transmitting step with the actuated signal that is produced after producing actuated signal.

26. object route Forecasting Methodology as claimed in claim 20, wherein, described degree of disturbance calculation procedure

Lean on to such an extent that the value of the degree of disturbance between described specified object and each other object is increased or reduce specified amount than the nearer number of times of interference distance according to described specified object and each other object, described interference distance is the space length that object interferes with each other.

27. object route Forecasting Methodology as claimed in claim 20, wherein, described degree of disturbance calculation procedure

Value to each degree of disturbance between described specified object and other objects is weighted the back addition.

28. object route Forecasting Methodology as claimed in claim 20 further comprises the output step of exporting the information relevant with the route of selecting in described route selection step.

29. an object route prediction unit comprises:

At least store the position of object and the memory cell of the internal state of the speed that comprises described object;

The time aerial track generation unit that generates track that variation in the position that can take according to the described object of passing in time based on the position of the described object that is read and internal state after position of reading described object from described memory cell and the internal state is being made of time and space; And

Predict the predicting unit of the route of described object by the track probability ground that utilizes described track generation unit to generate, wherein

Described track generation unit comprises

From a plurality of operations, be chosen in the operation selected cell of the operation that described object carries out;

The object operating unit of a scheduled time slot is carried out in the operation that described operation selected cell is selected; And

Whether the described object in selected operation back is carried out in judgement at described object operating unit position and internal state satisfy the judging unit of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein

Repeat one group of operation of carrying out from described operation selected cell and select to handle the processing of the judgment processing of carrying out to described judging unit till reaching the track rise time that generates described track.

30. object route prediction unit as claimed in claim 29, wherein, described operation selected cell selects probability to select operation according to the operation that each operation in described a plurality of operations is authorized, and

Satisfy described controlled condition and mobile condition if the judged result of described judging unit is the position and the internal state of described object, then increase and be back to the operation that described operation selected cell carries out after the described period and select to handle.

31. object route prediction unit as claimed in claim 30 wherein, is selected probability by utilizing random number to define described operation.

32. object route prediction unit as claimed in claim 30, wherein, preestablishing will be by the track number of described track generation unit generation.

33. object route prediction unit as claimed in claim 29, wherein, if the result of the judgement that described judging unit is carried out is for satisfying described controlled condition and described mobile condition, but then carries out recursive call after the described period all selection operations are performed by increasing.

34. object route prediction unit as claimed in claim 29, wherein, the position and the internal state of a plurality of objects of described cell stores, and

Described track generation unit is the aerial track that generates each object in described a plurality of objects when described.

35. object route prediction unit as claimed in claim 34, wherein, described predicting unit is specified an object and is calculated the aerial probability that exists when described of object except that the object of described appointment from described a plurality of objects.

36. object route prediction unit as claimed in claim 29 comprises that further output comprises the output unit of the information that predicts the outcome of described predicting unit.

37. an object route prediction unit comprises:

At least store the position of a plurality of objects and the memory cell of the internal state of the speed that comprises each object;

The time aerial track generation unit that generates track that the variation of the position that can take according to each object the described a plurality of object of passing in time based on the position of the described object that is read and internal state after position of reading described a plurality of objects from described memory cell and internal state is being made of time and space;

Predict the predicting unit of the route of described a plurality of objects by the track probability ground that utilizes described track generation unit to generate; And

Calculate the degree of disturbance calculating unit of degree of disturbance based on predicting the outcome of described predicting unit, described degree of disturbance is represented the annoyance level between the route that route that specified object can be taked and other objects can take quantitatively, and wherein, described track generation unit comprises

From a plurality of operations, be chosen in the operation selected cell of the operation that described object carries out;

The object operating unit of a scheduled time slot is carried out in the operation that described operation selected cell is selected; And

Judge whether position that described object operating unit carries out the described object in selected operation back and internal state satisfy the judging unit of controlled condition relevant with the control of described object and the mobile condition relevant with the moving area of described object, wherein

Repeat one group of operation of carrying out from described operation selected cell and select to handle the processing of the judgment processing of carrying out to described judging unit till reaching the track rise time that generates described track.

38. object route prediction unit as claimed in claim 37, wherein, described degree of disturbance calculating unit

Lean on to such an extent that the value of the degree of disturbance between described specified object and each other object is increased or reduce specified amount than the nearer number of times of interference distance according to described specified object and each other object, described interference distance is the space length that object interferes with each other.

39. object route prediction unit as claimed in claim 38, wherein, described degree of disturbance calculating unit

When described specified object and one of other objects lean on closelyer than described interference distance, and the collision probability of the two articles that draws closer together in the air when described increases the value of the degree of disturbance between the described two articles pro rata.

40. object route prediction unit as claimed in claim 38, wherein, described degree of disturbance calculating unit

In described specified object and when moving to such an extent that becomes nearer than described interference distance with one of other objects with the value of the degree of disturbance of the two articles that draws closer together between the described two articles of increase of mobile the relative velocity that becomes nearer moment with being in proportion.

41. object route prediction unit as claimed in claim 38, wherein, described memory cell will be used to estimate the size of the relative velocity when colliding between the damage grade evaluation value of the damage grade that collision causes or loss from spoilage amount that collision causes and the different object stores accordingly, and

Described degree of disturbance calculating unit reads described damage grade evaluation value or described loss from spoilage amount according to the size at the relative velocity in the moment that the two articles that draws closer together draws closer together from described memory cell when described specified object and one of other objects lean on closelyer than described interference distance, and and described damage grade evaluation value or described loss from spoilage amount increase degree of disturbance between the described two articles pro rata.

42. object route prediction unit as claimed in claim 38, wherein, described degree of disturbance calculating unit

To be the value of described degree of disturbance from the initial position of each object from the required described time set of described initial position at one of described specified object and other objects when required time is less than the value of the degree of disturbance the two articles that draws closer together when leaning on closelyer than described interference distance.

43. object route prediction unit as claimed in claim 38, wherein, described degree of disturbance calculating unit

Value to each degree of disturbance between described specified object and other objects is weighted the back addition.

44. object route prediction unit as claimed in claim 37, wherein, described operation selected cell selects probability to select operation according to the operation that each operation in described a plurality of operations is authorized, and

If the result of the judgement that described judging unit is carried out is back to the operation selection processing that described operation selected cell is carried out after, then increasing the described period for the position and the internal state of described object satisfy described controlled condition and described mobile condition.

45. object route prediction unit as claimed in claim 44 wherein, is selected probability by utilizing random number to define described operation.

46. object route prediction unit as claimed in claim 44, wherein, preestablishing will be by the track number of described track generation unit generation.

47. object route prediction unit as claimed in claim 37 further comprises the output unit of the information of the degree of disturbance that output calculates corresponding to described degree of disturbance calculating unit.

48. object route prediction unit as claimed in claim 37 further comprises

Select the route selection unit of the route that described specified object will take according to the degree of disturbance that calculates by described degree of disturbance calculating unit.

49. object route prediction unit as claimed in claim 48, wherein, when described degree of disturbance being defined as the collision probability of the track collision that track that described specified object can take and other objects can take, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just more little, and

The route of degree of disturbance minimum is selected in described route selection unit.

50. object route prediction unit as claimed in claim 49, wherein, described route selection unit

When being arranged, the route of a plurality of degree of disturbance minimums from described a plurality of routes, selects route with predetermined additional choice criteria optimum matching.

51. object route prediction unit as claimed in claim 48, wherein, when described degree of disturbance is defined as minimum collision time, annoyance level between the route that the route that described specified object can be taked and other objects can be taked is more little, the value of described degree of disturbance is just big more, described minimum collision time is to collide the required shortest time from initial position in the collision that takes place between described specified object and described other objects, and

The route of degree of disturbance maximum is selected in described route selection unit.

52. object route prediction unit as claimed in claim 51, wherein, described route selection unit

When being arranged, the route of a plurality of degree of disturbance maximums from described a plurality of routes, selects route with predetermined additional choice criteria optimum matching.

53. object route prediction unit as claimed in claim 49, the record and being used to that further is included in the position of the route of selecting according to described route selection unit realizes that the sequence of operation of described route is transferred to exterior actuated signal transmitting device with the actuated signal that is produced after producing actuated signal.

54. object route prediction unit as claimed in claim 48, wherein, described degree of disturbance calculating unit

Must the value of the degree of disturbance between described specified object and each other object be increased or reduce specified amount than the nearer number of times of interference distance according to described specified object and each other movement of objects, described interference distance be the space length that object interferes with each other.

55. object route prediction unit as claimed in claim 48, wherein, described degree of disturbance calculating unit

Value to each degree of disturbance between described specified object and other objects is weighted the back addition.

56. object route prediction unit as claimed in claim 48 further comprises the output unit of the information that output is relevant with the route of described route selection unit selection.

57. one kind is installed on the vehicle automatically to operate the automatic operation system of described vehicle, comprises:

Object route prediction unit as claimed in claim 48; And

Route that realization is selected by the route selection unit that is located in the described object route prediction unit and the actuating device of operating described vehicle according to actuated signal.

Images