CN114528040A - Environment self-adaption method, device, medium and roadside perception and calculation system - Google Patents
Environment self-adaption method, device, medium and roadside perception and calculation system Download PDFInfo
- Publication number
- CN114528040A CN114528040A CN202210096605.XA CN202210096605A CN114528040A CN 114528040 A CN114528040 A CN 114528040A CN 202210096605 A CN202210096605 A CN 202210096605A CN 114528040 A CN114528040 A CN 114528040A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- algorithms
- environment
- calculation
- equipment
- 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.)
- Pending
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 83
- 238000000034 method Methods 0.000 title claims abstract description 46
- 230000008447 perception Effects 0.000 title claims abstract description 15
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 335
- 230000007613 environmental effect Effects 0.000 claims abstract description 39
- 238000012549 training Methods 0.000 claims description 47
- 230000003044 adaptive effect Effects 0.000 claims description 28
- 238000004590 computer program Methods 0.000 claims description 13
- 238000005286 illumination Methods 0.000 claims description 10
- 230000006978 adaptation Effects 0.000 claims description 8
- 238000009434 installation Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 230000006855 networking Effects 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 102000003712 Complement factor B Human genes 0.000 description 1
- 108090000056 Complement factor B Proteins 0.000 description 1
- 241000497429 Obus Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/445—Program loading or initiating
- G06F9/44505—Configuring for program initiating, e.g. using registry, configuration files
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/445—Program loading or initiating
- G06F9/44568—Immediately runnable code
- G06F9/44578—Preparing or optimising for loading
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention discloses an environment self-adaptive method, which comprises the following steps: traversing the calculation requirements of roadside computing equipment, and deploying multiple groups of algorithms for the same calculation requirement; configuring a plurality of groups of algorithm strategies corresponding to different computing requirements according to the environmental factors of the equipment; after configuration is completed, acquiring a demand to be calculated, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence corresponding to each group of algorithms; and obtaining the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms. The algorithm combination is configured according to each calculation requirement, a plurality of groups of different algorithms with the same calculation target are deployed to deal with different equipment deployment environments or weather environments, and the algorithm which is matched with the actual environment preferentially is adopted, so that the problems that the equipment applicability is poor, the accuracy rate is easy to deteriorate and the specific scene cannot be considered when the same algorithm is adopted in a roadside perception and calculation system in the prior art are effectively solved.
Description
Technical Field
The invention relates to the technical field of Internet of vehicles and artificial intelligence, in particular to an environment self-adaption method, device and medium and a roadside sensing and computing system.
Background
In the application of car networking and intelligent transportation, single perception equipment, for example laser radar, millimeter wave radar, camera all have respective short slab, are difficult to be applicable to all scenes.
The existing roadside perception and computing system intelligently fuses data acquired from different perception devices to realize advantage complementation. Specifically, firstly, calibrating a reference point in a sensing range of a sensor, finding a coordinate system conversion relation among multiple sensors, and realizing space synchronization among different sensors; setting a unique clock source to provide the same reference time for each device, and calibrating the clock time of each system by each device according to the reference time so as to realize time synchronization among different devices; then fusing data which come from different sensors and realize space-time synchronization; and finally, extracting the target object or detecting the traffic event by utilizing algorithms such as machine vision, point cloud segmentation, target detection, classification and tracking and the like to form structured data. Algorithms used by different devices are different, but generally the same device adopts the same algorithm under different deployment environments, different illumination conditions, different visibility and different meteorological conditions, so that the device applicability is poor, the accuracy is rapidly deteriorated once the deployment position, the illumination conditions and the climatic conditions are changed, and the specific scene cannot be considered.
Disclosure of Invention
The embodiment of the invention provides an environment self-adaption method, an environment self-adaption device, a medium and a roadside sensing and calculating system, and aims to solve the problems that in the prior art, when the same algorithm is adopted in the roadside sensing and calculating system, the equipment applicability is poor, the accuracy rate is easy to deteriorate, and a specific scene cannot be considered at the same time.
An environment adaptive method, the method comprising:
traversing the calculation requirements of the roadside computing equipment, and deploying multiple groups of algorithms for the same calculation requirement;
configuring a plurality of groups of algorithm strategies corresponding to different calculation requirements according to the environmental factors of the equipment;
after configuration is completed, acquiring a demand to be calculated, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence corresponding to each group of algorithms;
and obtaining the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms.
Optionally, the environmental factors include a device deployment factor and a weather factor;
the multiple sets of algorithm configuration algorithm strategies corresponding to different computing requirements according to the environmental factors of the equipment comprise:
aiming at each calculation demand, acquiring a plurality of groups of algorithms corresponding to the calculation demand;
traversing each group of algorithms, and configuring algorithm strategies corresponding to the algorithms according to the equipment deployment factors and the weather factors;
the algorithm strategy corresponding to each algorithm comprises an algorithm parameter set and an algorithm weight.
Optionally, before configuring algorithm policies for multiple sets of algorithms corresponding to different computing requirements according to the environmental factors of the device, the method includes:
respectively setting threshold intervals for the environmental factors of the equipment;
when multiple groups of algorithms corresponding to the same calculation requirement are trained, combining the environment factors according to different threshold intervals to obtain at least one environment factor combination set;
classifying preset training samples according to the environment factor combination set to obtain a training sample subset corresponding to each environment factor combination set;
and aiming at each environment factor combination set, training and optimizing parameters of each group of algorithms by adopting the corresponding training sample subset.
Optionally, after completing training and parameter tuning of the algorithm, the method comprises:
calculating the average accuracy of the algorithm result of each group of algorithms aiming at each environment factor combination set;
acquiring the algorithm weight of the algorithm according to the average accuracy;
wherein, the higher the average accuracy, the larger the algorithm weight.
Optionally, after calculating an average accuracy of the algorithm results of each set of algorithms for each set of combinations of environmental factors, the method further comprises:
traversing each group of algorithms, and comparing the average accuracy of the algorithm results with a preset accuracy threshold;
if the average accuracy is smaller than the preset accuracy threshold, acquiring sensing information acquired by the road side sensing equipment from the actual environment;
and adopting the perception information to carry out adaptive training on the algorithm.
Optionally, the equipment deployment factor includes an equipment installation height and an equipment pitch angle;
the weather factors comprise illumination quantitative indexes, visibility quantitative indexes and weather types.
Optionally, when the to-be-calculated requirement is a target classification, the obtaining the calculation result of the to-be-calculated requirement according to the algorithm results and the confidence degrees of the multiple sets of algorithms includes:
calculating an election factor of each group of algorithms, wherein the election factor is the product between the confidence coefficient corresponding to the algorithm and the algorithm weight;
if the target types output by the multiple groups of algorithms corresponding to the requirements to be calculated are different, acquiring the algorithm corresponding to the election factor with the largest value as a target algorithm, wherein the target type obtained by election is the target type output by the target algorithm, and the confidence coefficient obtained by election is the confidence coefficient corresponding to the target algorithm;
if the target types output by the multiple groups of algorithms corresponding to the requirements to be calculated are not completely the same, adding the election factors of the multiple algorithms outputting the same target type to obtain election factors and values, comparing the election factors and values of different target types, obtaining the target type corresponding to the election factor and value with the largest value as the target type obtained by election, and obtaining the confidence coefficient obtained by election as the ratio of the election factors and values to the sum of the algorithm weights of the algorithms corresponding to the target types.
An environment adaptive apparatus, the apparatus comprising:
the deployment module is used for traversing the calculation requirements of the roadside computing equipment and deploying a plurality of groups of algorithms for the same calculation requirements;
the configuration module is used for configuring a plurality of groups of algorithm strategies corresponding to different computing requirements according to the environmental factors of the equipment;
the calculation module is used for acquiring a demand to be calculated after configuration is completed, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence coefficient corresponding to each group of algorithms;
and the acquisition module is used for acquiring the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms.
A computer-readable storage medium storing a computer program which, when executed by a processor, implements an environment adaptation method as described above.
A roadside awareness and computing system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor implementing the environment adaptation method as described above when executing the computer program.
According to the embodiment of the invention, multiple groups of algorithms are deployed for the same calculation requirement by traversing the calculation requirement of the roadside calculation equipment; configuring a plurality of groups of algorithm strategies corresponding to different calculation requirements according to the environmental factors of the equipment; after configuration is completed, acquiring a demand to be calculated, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence; and obtaining the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms. The algorithm combination is configured according to each calculation requirement, a plurality of groups of different algorithms with the same calculation target are deployed to deal with different equipment deployment environments or weather environments, and the algorithm which is matched with the actual environment preferentially is adopted, so that the problems that the equipment applicability is poor, the accuracy rate is easy to deteriorate and the specific scene cannot be considered when the same algorithm is adopted in a roadside perception and calculation system in the prior art are effectively solved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments of the present invention will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive labor.
FIG. 1 is a flow chart of an environment adaptation method provided by an embodiment of the invention;
fig. 2 is a flowchart of step S102 in the environment adaptive method provided by an embodiment of the present invention;
FIG. 3 is a flowchart illustrating algorithm training and parameter tuning in an environment adaptive method according to an embodiment of the present invention;
fig. 4 is a flowchart illustrating an implementation of obtaining a calculation result of a demand to be calculated in the environment adaptive method according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an environment adaptive apparatus according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of a roadside sensing and computing system according to an embodiment of the present invention;
FIG. 7 is a schematic diagram of a computer device according to an embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention optimizes all-weather adaptive capacity of the roadside computing equipment in different deployment environments, and deploys a plurality of groups of algorithms for the same computing requirement by traversing the computing requirement of the roadside computing equipment; configuring a plurality of groups of algorithm strategies corresponding to different calculation requirements according to the environmental factors of the equipment; after configuration is completed, acquiring a demand to be calculated, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence; obtaining the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms; therefore, the problems that equipment applicability is poor, accuracy rate is easy to deteriorate and a specific scene cannot be considered when the same algorithm is adopted in a roadside sensing and computing system in the prior art are effectively solved, and the accuracy rate of detecting, classifying and tracking target objects such as traffic participants, obstacles and the like and the accuracy rate of detecting traffic events by the roadside sensing and computing system under different deployment environments, different illumination conditions, different visibility and different meteorological conditions are effectively improved.
The environment adaptive method provided by the present embodiment is described in detail below. As shown in fig. 1, the environment adaptive method includes:
in step S101, the computation requirements of the roadside computing devices are traversed, and multiple sets of algorithms are deployed for the same computation requirements.
Here, the roadside computing device is an important component of a roadside sensing and computing system. In order to provide the road condition sensing capability beyond the visual range for the intelligent networked vehicles, the roadside computing equipment performs space-time synchronization and information fusion on information collected by sensing equipment arranged on the roadside, extracts, classifies and tracks target objects such as traffic participants and obstacles, detects traffic events and the like, and forms structured data. Including but not limited to laser radar, millimeter wave radar, cameras, etc. The roadside computing device splits the above-described functions into multiple fine-grained computing demands.
According to the embodiment of the invention, multiple groups of algorithms are deployed on the roadside computing equipment aiming at the same computing requirement. The calculation targets of each group of algorithms with the same calculation requirement are the same, but the calculation accuracy of each group of algorithms is respectively superior under different deployment environments or different illumination, visibility and meteorological conditions. The method and the device can effectively solve the problem of accuracy degradation of a single algorithm when equipment deployment environments are different or weather environments are changed by configuring the algorithm combination according to each computing requirement and deploying a plurality of groups of different algorithms with the same computing target.
In step S102, algorithm strategies are configured for a plurality of sets of algorithms corresponding to different computing requirements according to the environmental factors of the device.
Here, the environmental factors of the devices vary according to different computing needs, including but not limited to device deployment factors and weather factors. The equipment deployment factor is basically fixed after the equipment deployment is completed and does not change, including but not limited to the equipment installation height and the equipment pitch angle. The weather factors change in real time, including but not limited to illumination quantitative indicators, visibility quantitative indicators, and weather types. The weather factors can be acquired through sensing equipment deployed on the road side, and can also be acquired from a third-party meteorological platform.
As a preferred example of the present invention, as shown in fig. 2, the multiple sets of algorithm configuration algorithm policies corresponding to different computation requirements according to the environmental factors of the device in step S102 include:
in step S201, for each computation demand, a plurality of sets of algorithms corresponding to the computation demand are obtained.
In step S202, each group of algorithms is traversed, and algorithm policies corresponding to the algorithms are configured according to the device deployment factor and the weather factor.
Herein, the embodiment of the present invention adapts, for each algorithm, an algorithm policy matched with the current environment factor combination in combination with the input environment factor, where the algorithm policy corresponding to each algorithm includes an algorithm parameter set and an algorithm weight. The algorithm strategy set of multiple groups of algorithms aiming at the same calculation requirement forms an algorithm combination configuration strategy of the calculation requirement.
Optionally, in order to avoid frequent changes of the algorithm strategy, in the embodiment of the present invention, Kq threshold intervals are respectively set for each environmental factor, and q represents the number of the environmental factor. Therefore, when the threshold interval corresponding to the device deployment factor or the weather factor changes, the embodiment of the present invention triggers the adjustment of the algorithm policy.
Here, Kq is not set too high in order to reduce the complexity of algorithm training. For a certain calculation requirement, if an environmental factor has a small influence on the accuracy of the calculation result, Kq corresponding to the environmental factor may be 1.
The algorithm parameter set and the algorithm weight in the algorithm strategy are obtained by training and parameter tuning of the algorithm. After a plurality of sets of algorithms are deployed according to different computing requirements, the embodiment of the invention trains and optimizes the parameters of the plurality of sets of algorithms corresponding to each computing requirement under the action of different environmental factors. As shown in fig. 3, the training and parameter tuning of the multiple sets of algorithms corresponding to each computation requirement includes:
in step S301, threshold sections are set for the environmental factors of the devices, respectively.
The threshold interval is set according to practical applications of different environmental factors, and is not limited herein.
In step S302, when training multiple sets of algorithms corresponding to the same computation requirement, the environmental factors are combined according to different threshold intervals to obtain at least one environmental factor combination set.
The embodiment of the invention combines the preset environmental factors according to respective threshold value intervals to form different environmental factor combination sets. For example, the environment factor a has a threshold interval a1, the environment factor B has a threshold interval B1 and a threshold interval B2, and the environment factor combination set formed by the two includes a threshold interval a1, a threshold interval B1, a threshold interval B2, a threshold interval a1+ a threshold interval B1, and a threshold interval a1+ a threshold interval B2.
In step S303, the preset training samples are classified according to the environment factor combination set, so as to obtain a training sample subset corresponding to each environment factor combination set.
Here, the preset training sample is a set of data samples used for training. The training samples are classified according to the environment factor combination set, the training samples are divided into a plurality of subsets, the training sample subset corresponding to each environment factor combination set is obtained, and the training sample subset is used for training and parameter tuning of the algorithm under the influence of the corresponding environment factor combination set.
In step S304, for each set of environment factor combinations, each set of algorithms is trained and parameter-tuned using the corresponding training sample subset.
Here, a set of environment factors is combined to correspond to an application scenario of a specific environment state, such as environment states determined by different deployment environments, different illumination, visibility, and weather conditions. And the training sample subset corresponding to the environment factor combination set represents data collected by the perception equipment under the specific environment state. For the same calculation requirement, the embodiment of the invention adopts the corresponding training sample subset to train and optimize the parameters of each group of algorithms aiming at each environment factor combination set, so as to obtain the calculation result of each group of algorithms under the specific environment state.
Optionally, as a preferred example of the present invention, after completing training and parameter tuning of the algorithm, the method includes:
in step S305, an average accuracy of the algorithm results for each set of algorithms is calculated for each set of combinations of environmental factors.
In the embodiment of the invention, for a calculation requirement, under different environment factor combination sets, each group of algorithm outputs one algorithm result once for operation, the accuracy of the algorithm result is calculated, and then the average accuracy is calculated according to the accuracy of the algorithm result obtained after a plurality of times of algorithm operation. Under a specific set of environment factors, each set of algorithms under a computational demand corresponds to an average accuracy.
In step S306, the algorithm weight of the algorithm is obtained according to the average accuracy.
Aiming at the same calculation requirement, the embodiment of the invention compares the average accuracy rates corresponding to a plurality of groups of algorithms under the calculation requirement, sorts the algorithms according to the accuracy rates, and then gives corresponding algorithm weights. Here, the algorithm weight is used as a judgment factor of the roadside calculation device responding to the calculation demand and outputting the response result. The higher the average accuracy rate is, the larger the algorithm weight is, so as to ensure that the calculation result of the algorithm with higher accuracy rate in the training process under the environment is preferentially adopted.
Optionally, due to the specificity of the actual operating environment, the roadside perception and computing system has an average accuracy of the algorithm result lower than expected in some weather environments of some deployment environments, and the algorithm can be adaptively trained from data acquired from the actual environment, so that the algorithm combination configuration strategy in the specific environment is further optimized. As a preferred example of the present invention, after calculating the average accuracy of the algorithm results of each set of algorithms for each set of environment factor combinations in step S305, the method further comprises:
in step S307, each set of algorithms is traversed, and the average accuracy of the algorithm results is compared with a preset accuracy threshold.
The preset accuracy threshold is used as a criterion for determining whether to trigger the adaptive training, and may be set according to practical applications, which is not limited herein. One set of algorithms corresponds to a predetermined accuracy threshold, and the predetermined accuracy thresholds corresponding to different sets of algorithms may be the same or different.
In step S308, if the average accuracy is smaller than the preset accuracy threshold, acquiring sensing information acquired by the roadside sensing device from the actual environment.
If the average accuracy is smaller than the preset accuracy threshold, it indicates that the average accuracy of the calculation result of the roadside sensing and calculating system is lower than expected in some weather environments of some deployment environments, and at this time, adaptive training is triggered, and sensing information is collected from actual environments.
In step S309, the algorithm is adaptively trained using the perceptual information.
The embodiment of the invention utilizes the perception information collected from the actual environment to carry out adaptive training on the algorithm, further optimizes the algorithm combination configuration strategy in the specific environment so as to improve the accuracy of the calculation result in the specific scene by adopting the personalized algorithm combination configuration strategy, and then issues the training result to carry out algorithm deployment so as to ensure that the algorithm combination configuration strategy is more matched with the actual scene.
Here, the result of the adaptive training includes, but is not limited to, a new algorithm combination, a new environmental factor threshold interval partition, a new algorithm parameter set, and a new algorithm weight ratio, and implements maintenance and management of algorithm combination configuration policies deployed on all roadside computing devices, so as to ensure that when a roadside computing device recovers from a fault or a new device is replaced, a matched algorithm combination configuration policy can be quickly obtained.
In step S103, after configuration is completed, a demand to be calculated is obtained, and a plurality of sets of algorithms corresponding to the demand to be calculated are run to obtain a result and a confidence of each set of algorithms.
Here, after configuration is complete, each algorithm has adapted an algorithm strategy that matches the current combination of environmental factors. For the demand to be calculated, the embodiment of the invention calls a plurality of groups of algorithms corresponding to the demand to be calculated, and takes actual acquired data of sensing equipment in the roadside sensing and calculating system or output data of the last calculating link as algorithm input to execute the algorithm. Each set of algorithms will output an algorithm result after running, and confidence. The algorithm confidence degree refers to the confidence degree corresponding to the algorithm result.
In step S104, a calculation result of the demand to be calculated is obtained according to the algorithm results and the confidence degrees of the plurality of sets of algorithms.
Here, the embodiment of the present invention selects the final calculation result and the confidence level for the algorithm result and the confidence level output by the algorithm and the algorithm weight.
Alternatively, the following description will be given taking the object classification as an example. As a preferred example of the present invention, as shown in fig. 4, when the to-be-calculated requirement is a target category, the step S104 of obtaining the calculation result of the to-be-calculated requirement according to the algorithm results and the confidence degrees of the multiple sets of algorithms includes:
in step S401, an election factor of each set of algorithms is calculated, where the election factor is a product between a confidence corresponding to an algorithm and a weight of the algorithm.
Here, assume that there are a series of objects, and n sets of algorithms are matched for the object classification, i.e., the requirement to be calculated, in combination with the environment factor, and the corresponding algorithm combination configuration policy is (algorithm i, algorithm parameter set i, algorithm weight i), i represents an algorithm number, and i is 1 to n. When one of the objects is classified, the n groups of algorithms and the matched algorithm parameter set are adopted to obtain a calculation result (object type i, confidence coefficient i), wherein the object type i represents the classification of the object calculated by the algorithm i, and the confidence coefficient i is the credibility of the classification of the object calculated by the algorithm i. And calculating an election factor Vi of each group of algorithms, namely confidence i and algorithm weight i.
In step S402, if the target types output by the multiple sets of algorithms corresponding to the needs to be calculated are different, the algorithm corresponding to the election factor with the largest value is obtained as the target algorithm, the target type obtained by election is the target type output by the target algorithm, and the confidence obtained by election is the confidence corresponding to the target algorithm.
Here, if the target types calculated by all the algorithms are different, the target type i calculated by the algorithm corresponding to the election factor Vi with the largest value is the elected target type, and the confidence coefficient i calculated by the algorithm is the elected confidence coefficient.
In step S403, if the target types output by the multiple sets of algorithms corresponding to the needs to be calculated are not completely the same, adding the election factors of the multiple algorithms outputting the same target type to obtain election factors and values, comparing the election factors and values of different target types, obtaining the target type corresponding to the election factor and value with the largest value as the target type obtained by election, and obtaining a confidence level obtained by election as a ratio of the election factor and value to the sum of the algorithm weights of the algorithms corresponding to the target types.
Here, if there are multiple sets of algorithms that calculate the same object type, the same object type will be calculatedVi addition of multiple groups of algorithms obtains election factor sum value VSum,j(j is the object type number), and the election factor and the value V of different object types are combinedSum,jA comparison is made, wherein the election factor with the largest value and the value VSum,jThe corresponding target type j is the selected target type, and the confidence coefficient of the selection is VSum,jThe calculation result is the sum of the algorithm weights of the algorithm of the target type j.
Illustratively, assuming there is a target classification, 3 algorithms are matched:
the algorithm weight of the algorithm 1 is 0.6, the calculated target type is A, and the confidence coefficient is 0.95;
the algorithm weight of the algorithm 2 is 0.1, the calculated target type is A, and the confidence coefficient is 0.8;
algorithm 3 has an algorithm weight of 0.3 and the calculated target type is B with a confidence of 0.85.
Then the target type is the election factor and value V corresponding to ASum,A0.6 × 0.95+0.1 × 0.8 ═ 0.65; election factor and value V corresponding to target type BSum,B0.3 × 0.85 ═ 0.255. Compared with VSum,A>VSum,BSo the final elected target type is a, with confidence VSum,AAnd 0.93, which is the calculation result of the demand to be calculated.
In summary, in the embodiments of the present invention, multiple sets of algorithms are deployed for the same computation demand by traversing the computation demand of the roadside computing device; configuring a plurality of groups of algorithm strategies corresponding to different calculation requirements according to the environmental factors of the equipment; after configuration is completed, acquiring a demand to be calculated, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence of each group of algorithms; and obtaining the calculation result of the demand to be calculated according to the algorithm results of the multiple groups of algorithms and the corresponding confidence degrees. The algorithm combination is configured according to each calculation requirement, a plurality of groups of different algorithms with the same calculation target are deployed to deal with different equipment deployment environments or weather environments, and the algorithm which is matched with the actual environment preferentially is adopted, so that the problems that the equipment applicability is poor, the accuracy rate is easy to deteriorate and the specific scene cannot be considered when the same algorithm is adopted in a roadside perception and calculation system in the prior art are effectively solved.
It should be understood that, the sequence numbers of the steps in the foregoing embodiments do not imply an execution sequence, and the execution sequence of each process should be determined by functions and internal logic of the process, and should not limit the implementation process of the embodiments of the present invention in any way.
In an embodiment, the present invention further provides an environment adaptive apparatus, which corresponds to the environment adaptive method in the above embodiments one to one. As shown in fig. 5, the environment adaptive apparatus includes a deployment module 51, a configuration module 52, a calculation module 53, and an acquisition module 54. The functional modules are explained in detail as follows:
the deployment module 51 is used for traversing the calculation requirements of the roadside computing equipment and deploying multiple groups of algorithms for the same calculation requirements;
the configuration module 52 is configured to configure algorithm strategies for a plurality of sets of algorithms corresponding to different computing requirements according to the environmental factors of the equipment;
the calculation module 53 is configured to obtain a requirement to be calculated after configuration is completed, and run a plurality of sets of algorithms corresponding to the requirement to be calculated to obtain an algorithm result and a confidence corresponding to each set of algorithms;
and an obtaining module 54, configured to obtain a calculation result of the requirement to be calculated according to the algorithm results and the confidence degrees of the multiple sets of algorithms.
Optionally, the environmental factors include a device deployment factor and a weather factor;
the configuration module 52 includes:
the acquiring unit is used for acquiring a plurality of groups of algorithms corresponding to the calculation requirements aiming at each calculation requirement;
the configuration unit is used for traversing each group of algorithms and configuring the algorithm strategies corresponding to the algorithms according to the equipment deployment factors and the weather factors;
the algorithm strategy corresponding to each algorithm comprises an algorithm parameter set and an algorithm weight.
Optionally, before configuring algorithm strategies for multiple sets of algorithms corresponding to different computing requirements according to the environmental factors of the device, the apparatus includes a training and tuning module 55, configured to:
respectively setting threshold intervals for the environmental factors of the equipment;
when multiple groups of algorithms corresponding to the same calculation requirement are trained, combining the environment factors according to different threshold intervals to obtain at least one environment factor combination set;
classifying preset training samples according to the environment factor combination set to obtain a training sample subset corresponding to each environment factor combination set;
and aiming at each environment factor combination set, training and optimizing parameters of each group of algorithms by adopting the corresponding training sample subset.
Optionally, the training and tuning module is further configured to: after the training and parameter tuning of the algorithm is completed,
calculating the average accuracy of the algorithm result of each group of algorithms aiming at each environment factor combination set;
acquiring the algorithm weight of the algorithm according to the average accuracy;
wherein, the higher the average accuracy, the larger the algorithm weight.
Optionally, the training and tuning module is further configured to:
traversing each group of algorithms, and comparing the average accuracy of the algorithm results with a preset accuracy threshold;
if the average accuracy is smaller than the preset accuracy threshold, acquiring sensing information acquired by the road side sensing equipment from the actual environment;
and adopting the perception information to carry out adaptive training on the algorithm.
Optionally, the equipment deployment factor includes an equipment installation height and an equipment pitch angle;
the weather factors comprise illumination quantitative indexes, visibility quantitative indexes and weather types.
Optionally, when the demand to be calculated is a target category, the obtaining module 54 includes:
the computing unit is used for computing election factors of each group of algorithms, wherein the election factors are products between confidence degrees corresponding to the algorithms and algorithm weights;
the first obtaining unit is used for obtaining the algorithm corresponding to the election factor with the largest numerical value as the target algorithm if target types output by a plurality of groups of algorithms corresponding to the requirements to be calculated are different, wherein the target type obtained by election is the target type output by the target algorithm, and the confidence coefficient obtained by election is the confidence coefficient corresponding to the target algorithm;
and the second acquisition unit is used for adding the election factors of a plurality of algorithms outputting the same target type to obtain election factors and values if the target types output by a plurality of groups of algorithms corresponding to the requirements to be calculated are not completely the same, comparing the election factors and values of different target types, acquiring the target type corresponding to the election factor and value with the largest value as the target type obtained by election, and obtaining the confidence coefficient obtained by election as the ratio of the election factor and value to the sum of the algorithm weights of the algorithms corresponding to the target types.
For specific limitations of the environment adaptive apparatus, reference may be made to the above limitations of the environment adaptive method, which are not described herein again. The respective modules in the above-described environment adaptive apparatus may be wholly or partially implemented by software, hardware, and a combination thereof. The modules can be embedded in a hardware form or independent of a processor in a computer device, and can also be stored in a memory in a software form, so that the processor can call and execute operations corresponding to the modules.
Optionally, as a preferred example of the present invention, fig. 6 is a roadside sensing and computing system provided in an embodiment of the present invention, including: roadside computing equipment, a V2X platform, a laser radar, a millimeter wave radar, a video camera, clock synchronization equipment, a roadside communication unit (RSU) and an intelligent networked vehicle/vehicle-mounted unit (OBU).
The roadside computing equipment carries out space-time synchronization and information fusion on information collected by sensing equipment arranged on the roadside in order to provide beyond-the-horizon road condition sensing capability for intelligent networked vehicles, extracts, classifies and tracks target objects such as traffic participants and obstacles, detects traffic events and the like, and forms structured data. The roadside computing equipment splits the functions into a plurality of fine-grained computing requirements, each computing requirement can be simultaneously realized by a plurality of different algorithms, the computing target of each algorithm is the same, and the computing accuracy is respectively superior under different deployment environments or different illumination, visibility and meteorological conditions. According to the embodiment of the invention, multiple groups of algorithms can be deployed on roadside computing equipment aiming at the same computing requirement, and the algorithm which is most matched with the current actual environment is adopted for the computing requirement through the mutual cooperation of the configuration module 52, the computing module 53 and the acquisition module 54, so that the all-weather self-adaptive capacity under different deployment environments is realized.
The V2X platform is composed of a deployment module 51 and a training and tuning module, and is used for training and deploying an algorithm deployed on a roadside computing device. In addition, V2X may also provide information including road condition information and traffic guidance information to the OBU through communication methods such as 3G/4G/5G.
The sensing equipment arranged on the roadside includes but is not limited to a laser radar, a millimeter wave radar, a camera and the like. The laser radar is used for obtaining relevant information of the target object to be measured, including but not limited to distance, direction, speed and shape. The millimeter wave radar is used for acquiring distance, direction and relative speed information of a single target object to be detected, and detecting average speed of traffic flow, lane occupancy rate, queuing length and time analysis. The video camera is used for acquiring audio and video information of the target object to be detected.
The all-time synchronization equipment is used for providing the same reference time for each equipment in the system by setting a unique clock source, and assisting each equipment in the system to calibrate respective clock time according to the provided reference time so as to realize time synchronization.
The road side communication unit RSU supports a communication mode of C-V2X and is used for providing various information including road structural information, traffic participant information, obstacle information, traffic event information, sign information, signal lamp state information, perception sharing information, road side coordination information and the like for the intelligent networked vehicle/vehicle-mounted unit OBU.
The intelligent networked vehicle/vehicle-mounted unit OBU supports a C-V2X communication mode and is used for acquiring information including road condition information, other traffic participant information, traffic guidance information and the like from nearby road side equipment RSUs or other intelligent networked vehicle/vehicle-mounted units OBUs; and supports a V2X platform through a 3G/4G/5G communication mode, and is used for acquiring information including road condition information, traffic guidance information and the like; the system also supports the acquisition of the running data of the vehicle through the CAN bus, including the current speed, the running direction and the like; the current position of the vehicle is acquired in real time through a built-in GNSS function, and the broadcasting of the vehicle state and the driving intention to nearby intelligent networking equipment is supported.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 7. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement an environment adaptation method.
In one embodiment, a computer device is provided, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the computer program:
traversing the calculation requirements of the roadside computing equipment, and deploying multiple groups of algorithms for the same calculation requirement;
configuring a plurality of groups of algorithm strategies corresponding to different calculation requirements according to the environmental factors of the equipment;
after configuration is completed, acquiring a demand to be calculated, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence corresponding to each group of algorithms;
and obtaining the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, databases, or other media used in embodiments provided herein may include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-mentioned division of the functional units and modules is illustrated, and in practical applications, the above-mentioned function distribution may be performed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-mentioned functions.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present invention, and are intended to be included within the scope of the present invention.
Claims (10)
1. An environment adaptive method, the method comprising:
traversing the calculation requirements of roadside computing equipment, and deploying multiple groups of algorithms for the same calculation requirement;
configuring a plurality of groups of algorithm strategies corresponding to different calculation requirements according to the environmental factors of the equipment;
after configuration is completed, acquiring a demand to be calculated, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence corresponding to each group of algorithms;
and obtaining the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms.
2. The environment adaptive method of claim 1, wherein the environment factors comprise a device deployment factor and a weather factor;
the multiple sets of algorithm configuration algorithm strategies corresponding to different computing requirements according to the environmental factors of the equipment comprise:
aiming at each calculation demand, acquiring a plurality of groups of algorithms corresponding to the calculation demand;
traversing each group of algorithms, and configuring algorithm strategies corresponding to the algorithms according to the equipment deployment factors and the weather factors;
the algorithm strategy corresponding to each algorithm comprises an algorithm parameter set and an algorithm weight.
3. The environment adaptive method of claim 2, wherein prior to configuring algorithm policies for sets of algorithms corresponding to different computational requirements according to environmental factors of a device, the method comprises:
respectively setting threshold intervals for the environmental factors of the equipment;
when multiple groups of algorithms corresponding to the same calculation requirement are trained, combining the environment factors according to different threshold intervals to obtain at least one environment factor combination set;
classifying preset training samples according to the environment factor combination set to obtain a training sample subset corresponding to each environment factor combination set;
and aiming at each environment factor combination set, training and optimizing parameters of each group of algorithms by adopting the corresponding training sample subset.
4. The environment adaptation method of claim 3, wherein after completion of training and parameter tuning of the algorithm, the method comprises:
calculating the average accuracy of the algorithm result of each group of algorithms aiming at each environment factor combination set;
acquiring the algorithm weight of the algorithm according to the average accuracy;
wherein, the higher the average accuracy, the larger the algorithm weight.
5. The environment adaptive method of claim 4, wherein after calculating an average accuracy rate of algorithm results for each set of algorithms for each set of combinations of environmental factors, the method further comprises:
traversing each group of algorithms, and comparing the average accuracy of the algorithm results with a preset accuracy threshold;
if the average accuracy is smaller than the preset accuracy threshold, acquiring sensing information acquired by the road side sensing equipment from the actual environment;
and adopting the perception information to carry out adaptive training on the algorithm.
6. The environment adaptation method of any one of claims 2 to 6, wherein the equipment deployment factor comprises an equipment installation height, an equipment pitch angle;
the weather factors comprise an illumination quantitative index, a visibility quantitative index and a weather type.
7. The environment adaptive method of claim 1, wherein when the to-be-calculated requirement is a target classification, the obtaining the calculation result of the to-be-calculated requirement according to the algorithm results and the confidence degrees of the plurality of sets of algorithms comprises:
calculating an election factor of each group of algorithms, wherein the election factor is the product between the confidence coefficient corresponding to the algorithm and the algorithm weight;
if the target types output by the multiple groups of algorithms corresponding to the requirements to be calculated are different, acquiring the algorithm corresponding to the election factor with the largest value as a target algorithm, wherein the target type obtained by election is the target type output by the target algorithm, and the confidence coefficient obtained by election is the confidence coefficient corresponding to the target algorithm;
if the target types output by the multiple groups of algorithms corresponding to the requirements to be calculated are not completely the same, adding the election factors of the multiple algorithms outputting the same target type to obtain election factors and values, comparing the election factors and values of different target types, obtaining the target type corresponding to the election factor and value with the largest value as the target type obtained by election, and obtaining the confidence coefficient obtained by election as the ratio of the election factors and values to the sum of the algorithm weights of the algorithms corresponding to the target types.
8. An environment adaptive apparatus, characterized in that the apparatus comprises:
the deployment module is used for traversing the calculation requirements of the roadside computing equipment and deploying a plurality of groups of algorithms for the same calculation requirements;
the configuration module is used for configuring a plurality of groups of algorithm strategies corresponding to different computing requirements according to the environmental factors of the equipment;
the calculation module is used for acquiring a demand to be calculated after configuration is completed, and operating a plurality of groups of algorithms corresponding to the demand to be calculated to obtain an algorithm result and confidence coefficient corresponding to each group of algorithms;
and the acquisition module is used for acquiring the calculation result of the demand to be calculated according to the algorithm results and the confidence degrees of the multiple groups of algorithms.
9. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the environment adaptation method of any one of claims 1 to 7.
10. A roadside awareness and computing system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the environment adaptation method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210096605.XA CN114528040A (en) | 2022-01-26 | 2022-01-26 | Environment self-adaption method, device, medium and roadside perception and calculation system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210096605.XA CN114528040A (en) | 2022-01-26 | 2022-01-26 | Environment self-adaption method, device, medium and roadside perception and calculation system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114528040A true CN114528040A (en) | 2022-05-24 |
Family
ID=81622101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210096605.XA Pending CN114528040A (en) | 2022-01-26 | 2022-01-26 | Environment self-adaption method, device, medium and roadside perception and calculation system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114528040A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116985803A (en) * | 2023-09-26 | 2023-11-03 | 赛奎鹰智能装备(威海)有限责任公司 | Self-adaptive speed control system and method for electric scooter |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107025505A (en) * | 2017-04-25 | 2017-08-08 | 无锡中科智能农业发展有限责任公司 | A kind of paddy water requirement prediction method based on principal component analysis and neutral net |
| CN111259680A (en) * | 2020-02-13 | 2020-06-09 | 支付宝(杭州)信息技术有限公司 | Two-dimensional code image binarization processing method and device |
| CN111796925A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Method and device for screening algorithm model, storage medium and electronic equipment |
| CN112598084A (en) * | 2021-03-02 | 2021-04-02 | 深圳金三立视频科技股份有限公司 | Vehicle type identification method and terminal based on image processing |
| CN112633321A (en) * | 2020-11-26 | 2021-04-09 | 北京瑞友科技股份有限公司 | Artificial intelligence recommendation system and method |
-
2022
- 2022-01-26 CN CN202210096605.XA patent/CN114528040A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107025505A (en) * | 2017-04-25 | 2017-08-08 | 无锡中科智能农业发展有限责任公司 | A kind of paddy water requirement prediction method based on principal component analysis and neutral net |
| CN111796925A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Method and device for screening algorithm model, storage medium and electronic equipment |
| CN111259680A (en) * | 2020-02-13 | 2020-06-09 | 支付宝(杭州)信息技术有限公司 | Two-dimensional code image binarization processing method and device |
| CN112633321A (en) * | 2020-11-26 | 2021-04-09 | 北京瑞友科技股份有限公司 | Artificial intelligence recommendation system and method |
| CN112598084A (en) * | 2021-03-02 | 2021-04-02 | 深圳金三立视频科技股份有限公司 | Vehicle type identification method and terminal based on image processing |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116985803A (en) * | 2023-09-26 | 2023-11-03 | 赛奎鹰智能装备(威海)有限责任公司 | Self-adaptive speed control system and method for electric scooter |
| CN116985803B (en) * | 2023-09-26 | 2023-12-29 | 赛奎鹰智能装备(威海)有限责任公司 | Self-adaptive speed control system and method for electric scooter |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111127931B (en) | Vehicle road cloud cooperation method, device and system for intelligent networked automobile | |
| WO2018141403A1 (en) | System, device and method for managing traffic in a geographical location | |
| CN111382768A (en) | Multi-sensor data fusion method and device | |
| CN111462484A (en) | Congestion state determination method, device, equipment and computer readable storage medium | |
| CN110441790A (en) | Method and apparatus crosstalk and multipath noise reduction in laser radar system | |
| CN114670852A (en) | Method, device, equipment and medium for identifying abnormal driving behaviors | |
| US20230228593A1 (en) | Method, device and system for perceiving multi-site roadbed network and terminal | |
| CN114528040A (en) | Environment self-adaption method, device, medium and roadside perception and calculation system | |
| WO2022242465A1 (en) | Method and apparatus for fusing data of multiple sensors | |
| Hawas et al. | Optimized multistage fuzzy-based model for incident detection and management on urban streets | |
| CN109120916B (en) | Camera fault detection method and device and computer equipment | |
| CN116434523A (en) | Vehicle active safety control method and device based on constraint degree in information perception scene | |
| CN111681430B (en) | Method for predicting number of stop lines of signal lamp intersection in future in real time | |
| CN115273468A (en) | Traffic congestion control strategy generation method and device | |
| CN111483463B (en) | Vehicle-mounted unit and road side unit based pre-judging overtaking method and storage medium | |
| CN113936476A (en) | Signal lamp control method and device, computer equipment and storage medium | |
| CN115830860B (en) | Traffic accident prediction method and device | |
| CN110428612A (en) | Vehicle management-control method, device, computer equipment and storage medium | |
| CN117593708B (en) | Traffic digital twin method, equipment and storage medium containing vehicle identity information | |
| CN113790761B (en) | Driving end positioning method, device, computer equipment and storage medium | |
| RU2724596C1 (en) | Method, apparatus, a central device and a system for recognizing a distribution shift in the distribution of data and / or features of input data | |
| WO2022115993A1 (en) | Method and apparatus for tuning sensor fusion weights | |
| US20230322255A1 (en) | Multiple source mec assistance strategy for autonomous vehicles | |
| CN115620251A (en) | Target identification method, target identification device, electronic equipment and medium | |
| CN116486623A (en) | Traffic flow prediction method, prediction system and urban traffic signal control method |
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 |