CN106873595A - A kind of is recognition methods with garage based on Timed Automata - Google Patents
A kind of is recognition methods with garage based on Timed Automata Download PDFInfo
- Publication number
- CN106873595A CN106873595A CN201710146212.4A CN201710146212A CN106873595A CN 106873595 A CN106873595 A CN 106873595A CN 201710146212 A CN201710146212 A CN 201710146212A CN 106873595 A CN106873595 A CN 106873595A
- Authority
- CN
- China
- Prior art keywords
- car
- state
- sub
- behavior
- following model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 230000006399 behavior Effects 0.000 claims abstract description 37
- 239000000284 extract Substances 0.000 claims abstract description 4
- 239000012141 concentrate Substances 0.000 claims abstract 2
- 230000001133 acceleration Effects 0.000 claims description 10
- 230000006641 stabilisation Effects 0.000 claims description 7
- 238000011105 stabilization Methods 0.000 claims description 7
- 238000012546 transfer Methods 0.000 claims description 7
- 238000003064 k means clustering Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 2
- 230000009471 action Effects 0.000 description 9
- 238000011160 research Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention relates to it is a kind of based on Timed Automata with garage be recognition methods, comprise the following steps：S1, concentrates from traffic data and extracts original with car data, by the |input paramete symbolism of car-following model；S2, is trained using Timed Automata learning algorithm to car-following model, obtains car-following model automatic machine；S3, using the hidden state of car-following model automatic machine as the sub- state with car, to sub- state clustering；S4, probability is removed less than the sub- state of setting value, and then into multiple classifications, each classification correspondence one kind is with car behavior for merger；S5, obtains reality with car data as input, is obtained with car behavior by car-following model automatic machine.Compared with prior art, the present invention by the symbolism of multidimensional continuous time series and learn generation one have very strong interpretation with car behavior model, symbol is learnt by automatic machine, obtain hidden state, and clustered, the sub-line that can preferably embody with car is.
Description
Technical field
The present invention relates to one kind with garage be generation method, be with garage based on Timed Automata more particularly, to a kind of
Recognition methods.
Background technology
Automatic driving vehicle (hereinafter referred to as unmanned vehicle) is a kind of intelligentized mobile traffic, and it can replace the mankind
Driver completes a series of driving behaviors, is related to that environment sensing, navigator fix and intelligent decision control etc. is many multi-disciplinary to grind
Study carefully field.New ideas, the development trend, demonstrating computer science, the mould that lead automotive industry of the unmanned vehicle as modernized war
Formula recognizes the Important Platform with artificial intelligence technology level, all the time by Defence business, auto industry and colleges and universities and scientific research
The concern of mechanism.The research purpose of unmanned vehicle is exactly to replace the human driver to carry out vehicle autonomous driving work, then normal
Traffic downward driving during, be essential with interacting for other vehicles.And the interaction between vehicle includes：By row
Unmanned vehicle itself behavior is set to be recognized by other vehicles by decision-making, while unmanned vehicle can be allowed to recognize the behavior of other vehicles
And then make rational driving behavior.
With continuing to develop for unmanned vehicle technology, unmanned vehicle has been realized in some basic functions, can specifically tie
Travelled on structure road.Unmanned vehicle technology is along direction that is intelligent, personalizing and develops.The research of unmanned vehicle control problem is burnt
Point is progressively transferred to influencing each other between unmanned vehicle and running environment, the wherein research of vehicle social action exactly from functional realiey
The new problem of unmanned vehicle research attention and new challenge.
The social action of vehicle refers to driver's (human driver or unmanned vehicle control) when vehicle is driven, with week
Enclose vehicle and collectively constitute a colony, collaboration completes a kind of interbehavior of traveling task.This interaction both includes driver's energy
The behavior of surrounding vehicles is enough recognized, is also recognized by other drivers including vehicle itself behavior.When thering is vehicle to be close to,
Experienced mankind driver will produce reaction, and the driving side of oneself is determined by distinguishing the social action of Adjacent vehicles
Formula, such as accelerate overtake other vehicles, slows down give precedence to or Stop and give way.And for unmanned vehicle, only by nearby vehicle position and
It is incomplete that attitude information (such as spacing, acceleration, side drift angle) is controlled, it must be understood that other vehicle locations and attitude are believed
Cease the vehicle behavior (such as allow Che Huobing roads) expressed by change.Effective identification, base only are carried out to other vehicle social actions
Appropriate Driving control is taken in vehicle social action, and other vehicles around is capable of identify that the social action of unmanned vehicle,
Can cause that unmanned vehicle keeps safe, quick and stable traveling in wagon flow.
At present, the basic function of unmanned vehicle Driving control is more perfect, and for the identification of surrounding vehicles social action
Research with application also in the starting stage.Identification vehicle social action is a uncertain problem complicated and changeable, can be by
Social action can be played an important role in being added to the Ride Control System of unmanned vehicle to the development that personalizes of unmanned vehicle.
With garage to be most common behavior during the daily traveling of vehicle, vehicle is referred mainly to car with car driving behavior
Need to keep certain safe distance with front truck during traveling, it is to avoid rear-end collision occurs.The method for building up of car-following model mainly divides
It is two major classes：Physiological psychology model and stimulation-action model.According to the physiological psychology model that Wei Deman is proposed, with garage
To can be largely classified into：Free traveling behavior, close to front truck behavior, stabilization with garage it is and brake hard behavior.For the mankind
For driver, because the driving habit and driving style of driver are different, it is difficult to determine unified threshold value divide with
The sub-line that garage is is.Stimulation-action model is largely used in traffic flow analysis software, is driven with car according to a large amount of drivers
Data demarcate the parameter in car-following model, with obtain general driver with car behavior model.This class model often can only be thick
Expression driver slightly is with garage, for unmanned vehicle, it is necessary to more accurate car-following model judges and predicts surrounding vehicles
Behavior.
The content of the invention
The purpose of the present invention provides a kind of raising unmanned vehicle for the defect for overcoming above-mentioned prior art to exist
The intelligent level that personalizes based on Timed Automata with garage be recognition methods.
The purpose of the present invention can be achieved through the following technical solutions：
Extract original with car data, including rear vehicle speed, rear car acceleration, rear car from the track of vehicle of traffic data collection
With the relative distance and relative velocity of front truck, using the relative distance and relative velocity of rear vehicle speed, rear car and front truck as with car
The |input paramete of model, using k-means clustering algorithms by |input paramete symbolism；
S2, is trained using Timed Automata learning algorithm to car-following model, obtains car-following model automatic machine, with car mould
The output valve of type is rear car acceleration；
S3, using the hidden state of car-following model automatic machine as the sub- state with car, to sub- state clustering, for represent with
Car behavior, hidden state refers to the state that can not be observed, such as：The speed of car, acceleration etc. can be observed, shape is hidden
State refers to just the behavior that acceleration is reflected by speed, such as：With car, overtake other vehicles or lane change；
S4, probability is removed less than the sub- state of setting value, and then merger is into multiple classifications, each classification correspondence it is a kind of with
Car behavior；
S5, obtains reality with car data as input, is obtained with car behavior by car-following model automatic machine.
In described step S2, described Timed Automata contains four elements<A,ε,T,H>, wherein ε is event set, T
It is time-constrain collection, H is mapping ensemblen of the state to time-constrain, A is four-dimension tuple, Α=<Q,Σ,Δ,q0>, wherein Q is have
The intersection of limit state, ∑ is the limited intersection of symbol, and Δ is the limited intersection of state transfer, q0It is original state.
In described step S3, using hierarchical clustering method to sub- state clustering, use apart from computing formula be Jaro-
Distance, it is specific as follows：
Wherein, JS is similarity of character string, and character string is more similar, and JS represents string length, footnote i and j closer to 1, L
Expression will calculate two sequence numbers of character string of distance, NmatchRepresent two character numbers of string matching, NTRepresent dislocation
Character number half, take 1-JS carries out hierarchical clustering as character Distance conformability degree, and character string is more similar, 1-JS closer to
0。
Described includes that stabilization is near with car behavior, stabilization with distance in car behavior, stabilization over long distances with car behavior
Distance is with car behavior and intermediate transfer sub-line.
In described step S4, the setting value of probability is located between 1%~2%.
Compared with prior art, the present invention has advantages below：
(1) by the symbolism of multidimensional continuous time series and learning generation one, there is very strong interpretation to be with garage
Model, is learnt by automatic machine to symbol, obtains hidden state, and is clustered according to hidden state, can be more preferable
Embody and be with the sub-line of car.
(2) propose and the status switch of Timed Automata is clustered and automatic machine is carried out into modular method, i.e.,
Automatic machine is clustered, the corresponding sub- state set of each classification is referred to as module, preferably Explanation-based Learning And It can obtained
Automaton model, can also correspond into the sub-line for being with garage is.
(3) the Timed Automata model that this method is generated, can not only according to vehicle be presently in sub- state obtain with
Car behavior, meanwhile, the state transfer in automaton model has corresponded to probability distribution, so by the sub- state being presently in,
The probability that next sub- state occurs can be predicted.
Brief description of the drawings
Fig. 1 is the flow chart of the inventive method；
Fig. 2 is that the present embodiment ELBOW methods choose cluster data result；
Fig. 3 is the present embodiment Timed Automata training result；
Fig. 4 is the present embodiment hierarchical clustering result；
Fig. 5 is that the present embodiment is automatic machine hidden state cluster result with garage；
Fig. 6 (a) -6 (e) be the present embodiment with car Activity recognition result, wherein, 6 (a) is the rear car car that each sub-line is
Speed, relative distance and relative speed relationship, 6 (b) is the location and time relation that each sub-line is, 6 (c) be each sub-line be it is relative
Distance and relative speed relationship, 6 (d) is the relative velocity and time relationship that each sub-line is, 6 (e) be each sub-line be it is relative away from
From with time relationship.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, give detailed implementation method and specific operating process, but protection scope of the present invention is not limited to
Following embodiments.
Embodiment
A kind of is recognition methods with garage based on Timed Automata, first from disclosed traffic data collection (Next
Generation SIMulation, NGSIM) in extract vehicle with car data.Based on k-means clustering algorithms by car-following model
|input paramete (relative distance, relative velocity and rear car speed) carry out symbolism, using Timed Automata learning algorithm (RTI+
Learning Algorithm) car-following model is trained, obtain car-following model automatic machine.Then using the side of hierarchical clustering
Method is clustered to status switch, a series of sub- behavior pattern in obtaining being with garage, with this come to garage to carry out more
It is careful sub- Activity recognition and prediction.This patent has novelty practicality concurrently, be capable of real-time detection surrounding vehicles with garage
For and to vehicle behavior is predicted.
Specific steps to each step as shown in figure 1, be explained as follows：
1. with car data symbolism
Data used in the present embodiment are the highway data sets (NGSIM) that Bureau of Public Road announces.It is first
First, according to the feature for being with garage, initial data is carried out to be extracted with wheel paths.Trace information is included：The speed of target vehicle,
The relative distance and relative velocity of the longitudinal acceleration, target vehicle and front truck of target vehicle.Target vehicle is rear car, due to grinding
What is studied carefully is to be with garage, it is believed that in same track, can not consider lateral motion.Longitudinal acceleration is headstock direction of advance
Acceleration.The sample frequency of initial data is 10Hz, and the present embodiment proposes a kind of symbolism method based on k-means, makes
Cluster data is chosen with ELBOW methods, the result of ELBOW methods is as shown in Fig. 2 the corresponding central point of symbol is as shown in the table：
The corresponding central point of the symbol of table 1
2. Timed Automata study
Timed Automata used in this patent contains 4 elements：
<A,ε,T,H>
Wherein A is that one 4 dimension tuple is as follows：
Α=<Q,Σ,Δ,q0>
Q is the limited intersection of state, and ∑ is the limited intersection of symbol, and Δ is the limited intersection of state transfer, q0
It is original state.ε and T are the probability distribution of event and time.Timed Automata is one and is set up by observable data
The model of hidden state transfer, while considering the time-constrain of event generation, can preferably process multidimensional time-series.Profit
Algorithm is practised with time robotics to be trained the track sets after symbolism, obtain the automatic machine such as Fig. 3.
3. clustered with car behavior
After the automatic machine that expression is with garage is obtained, hidden state represents the sub- state with car, is used in the present embodiment
The method of hierarchical clustering is clustered to similar state subsequence, is represented with this with car behavior.Hierarchical clustering is used
Apart from computing formula be Jaro-distance, it is as follows：
Wherein, L represents string length, and i and j represents two character strings of distance to be calculated, NmatchRepresent two characters
The character number of String matching, NTThe half of the character number of dislocation is represented, the implication of the distance is exactly more similar two character strings,
Closer to 1,1 has been taken in this patent subtract JS distances carries out hierarchical clustering to its distance values, and cluster result is as shown in Figure 4.
4. Timed Automata model optimization and with car behavior explanation
Fig. 3 automatic machines are optimized in the present embodiment, eliminate the very low state of part probability of occurrence, and to cluster
Sub- state afterwards is sorted out and has been merged, and has obtained the Timed Automata model such as Fig. 5, and is to understand to each sub-line
Release, it is as shown in the table：
The sub- state of the automaton model of table 2 is explained
Automatic machine training method and automaton model proposed in the present embodiment can be used to identification with car behavior, will
Vehicle speed, relative velocity and relative distance are used as input afterwards, and partial results such as Fig. 5 and Fig. 6 (a) -6 (e) are shown.
Claims (5)
1. a kind of is recognition methods with garage based on Timed Automata, it is characterised in that comprised the following steps：
S1, concentrates from traffic data and extracts original with car data, including rear vehicle speed, rear car acceleration, rear car and front truck phase
Adjust the distance and relative velocity, using the relative distance and relative velocity of rear vehicle speed, rear car and front truck as car-following model input
Parameter, using k-means clustering algorithms by |input paramete symbolism；
S2, is trained using Timed Automata learning algorithm to car-following model, obtains car-following model automatic machine, car-following model
Output valve is rear car acceleration；
S3, using the hidden state of car-following model automatic machine as the sub- state with car, to sub- state clustering；
S4, probability is removed less than the sub- state of setting value, and then into multiple classifications, each classification correspondence one kind is with car for merger
Behavior；
S5, obtains reality with car data as input, is obtained with car behavior by car-following model automatic machine.
2. it is according to claim 1 it is a kind of based on Timed Automata with garage be recognition methods, it is characterised in that it is described
The step of S2 in, described Timed Automata contains four element ＜ Α, ε, Τ, H>, wherein ε is event set, and T is time-constrain
Collection, H be state to time-constrain mapping ensemblen, A for the four-dimension tuple, Α=<Q,Σ,Δ,q0>, wherein Q is limited state
Intersection, ∑ is the limited intersection of symbol, and Δ is the limited intersection of state transfer, q0It is original state.
3. it is according to claim 1 it is a kind of based on Timed Automata with garage be recognition methods, it is characterised in that it is described
The step of S3 in, using hierarchical clustering method to sub- state clustering, use apart from computing formula be Jaro-distance, specifically
It is as follows：
Wherein, JS is similarity of character string, and L represents string length, and footnote i and j represent two character strings of distance to be calculated
Sequence number, NmatchRepresent two character numbers of string matching, NTThe half of the character number of dislocation is represented, 1-JS conducts are taken
Character Distance conformability degree carries out hierarchical clustering.
4. it is according to claim 1 it is a kind of based on Timed Automata with garage be recognition methods, it is characterised in that it is described
With car behavior include stabilization over long distances with car behavior, stabilization in distance with car behavior, stabilization closely with car row
It is to be with intermediate transfer sub-line.
5. it is according to claim 1 it is a kind of based on Timed Automata with garage be recognition methods, it is characterised in that it is described
The step of S4 in, the setting value of probability is located between 1%~2%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710146212.4A CN106873595B (en) | 2017-03-13 | 2017-03-13 | A kind of follow the bus Activity recognition method based on Timed Automata |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710146212.4A CN106873595B (en) | 2017-03-13 | 2017-03-13 | A kind of follow the bus Activity recognition method based on Timed Automata |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106873595A true CN106873595A (en) | 2017-06-20 |
| CN106873595B CN106873595B (en) | 2019-09-27 |
Family
ID=59170760
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710146212.4A Active CN106873595B (en) | 2017-03-13 | 2017-03-13 | A kind of follow the bus Activity recognition method based on Timed Automata |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106873595B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109976188A (en) * | 2019-03-12 | 2019-07-05 | 广东省智能制造研究所 | A kind of cricket control method and system based on Timed Automata |
| CN110161848A (en) * | 2019-03-12 | 2019-08-23 | 广东省智能制造研究所 | A kind of single order straight line inverted pendulum control method and system based on Timed Automata |
| CN110516746A (en) * | 2019-08-29 | 2019-11-29 | 吉林大学 | A kind of driver's follow the bus behavior genre classification method based on no label data |
| CN110533610A (en) * | 2019-08-20 | 2019-12-03 | 东软医疗***股份有限公司 | The generation method and device of image enhancement model, application method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103531042A (en) * | 2013-10-25 | 2014-01-22 | 吉林大学 | Rear-end collision pre-warning method based on driver types |
| CN104915514A (en) * | 2015-06-25 | 2015-09-16 | 华东师范大学 | Time requirement modeling and verification method based on problem frame method |
| CN105260188A (en) * | 2015-10-24 | 2016-01-20 | 北京航空航天大学 | Time characteristic model and modeling method thereof |
| FR3031820A1 (en) * | 2015-01-15 | 2016-07-22 | Dassault Aviat | METHOD FOR DETERMINING THE VALIDITY OF A MODEL |
-
2017
- 2017-03-13 CN CN201710146212.4A patent/CN106873595B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103531042A (en) * | 2013-10-25 | 2014-01-22 | 吉林大学 | Rear-end collision pre-warning method based on driver types |
| FR3031820A1 (en) * | 2015-01-15 | 2016-07-22 | Dassault Aviat | METHOD FOR DETERMINING THE VALIDITY OF A MODEL |
| CN104915514A (en) * | 2015-06-25 | 2015-09-16 | 华东师范大学 | Time requirement modeling and verification method based on problem frame method |
| CN105260188A (en) * | 2015-10-24 | 2016-01-20 | 北京航空航天大学 | Time characteristic model and modeling method thereof |
Non-Patent Citations (1)
| Title |
|---|
| 张智勇 等: "《城市快速道路跟车行为的状态划分》", 《北京工业大学学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109976188A (en) * | 2019-03-12 | 2019-07-05 | 广东省智能制造研究所 | A kind of cricket control method and system based on Timed Automata |
| CN110161848A (en) * | 2019-03-12 | 2019-08-23 | 广东省智能制造研究所 | A kind of single order straight line inverted pendulum control method and system based on Timed Automata |
| CN109976188B (en) * | 2019-03-12 | 2022-01-07 | 广东省智能制造研究所 | Cricket control method and system based on time automaton |
| CN110533610A (en) * | 2019-08-20 | 2019-12-03 | 东软医疗***股份有限公司 | The generation method and device of image enhancement model, application method and device |
| CN110516746A (en) * | 2019-08-29 | 2019-11-29 | 吉林大学 | A kind of driver's follow the bus behavior genre classification method based on no label data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106873595B (en) | 2019-09-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Driving style analysis using primitive driving patterns with Bayesian nonparametric approaches | |
| Kruber et al. | An unsupervised random forest clustering technique for automatic traffic scenario categorization | |
| US11899411B2 (en) | Hybrid reinforcement learning for autonomous driving | |
| Dogan et al. | Autonomous driving: A comparison of machine learning techniques by means of the prediction of lane change behavior | |
| CN107492251A (en) | It is a kind of to be identified and driving condition supervision method based on the driver identity of machine learning and deep learning | |
| CN107229973A (en) | The generation method and device of a kind of tactful network model for Vehicular automatic driving | |
| CN107169567A (en) | The generation method and device of a kind of decision networks model for Vehicular automatic driving | |
| CN106971194A (en) | A kind of driving intention recognition methods based on the double-deck algorithms of improvement HMM and SVM | |
| CN106873595A (en) | A kind of is recognition methods with garage based on Timed Automata | |
| CN111275249A (en) | Driving behavior optimization method based on DQN neural network and high-precision positioning | |
| d'Agostino et al. | Learning-based driving events recognition and its application to digital roads | |
| Lin et al. | MOHA: A multi-mode hybrid automaton model for learning car-following behaviors | |
| Chu et al. | A review of driving style recognition methods from short-term and long-term perspectives | |
| Zhang et al. | An ensemble learning–online semi-supervised approach for vehicle behavior recognition | |
| Moosavi et al. | Driving style representation in convolutional recurrent neural network model of driver identification | |
| Li et al. | Personalized driver braking behavior modeling in the car-following scenario: An importance-weight-based transfer learning approach | |
| Zhang et al. | Multi-vehicle interaction scenarios generation with interpretable traffic primitives and gaussian process regression | |
| CN107045639B (en) | Following behavior generation method based on time automaton | |
| Li et al. | Research on lane change prediction model based on GBDT | |
| Gao et al. | Discretionary cut-in driving behavior risk assessment based on naturalistic driving data | |
| Wang et al. | A survey on datasets for the decision making of autonomous vehicles | |
| Zhang et al. | Car-following behavior model learning using timed automata | |
| Li et al. | Automated Vehicle Identification Based on Car-Following Data With Machine Learning | |
| Chen et al. | Platoon separation strategy optimization method based on deep cognition of a driver’s behavior at signalized intersections | |
| Zhang et al. | An Embedded Driving Style Recognition Approach: Leveraging Knowledge in Learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |