Vehicle identification method, identification system, equipment and medium

Abstract

The present application discloses a vehicle identification method, identification system, device and medium. In the vehicle identification method, a vehicle identification request initiated by a terminal is received, and input parameters of the vehicle to be identified are retrieved according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and stepping data parameters; the input parameters are respectively input into multiple first classification models, and multiple first classification labels of the vehicle to be identified are obtained through multiple first classification models; the predicted probability of the vehicle to be identified is obtained according to the multiple first classification labels; the predicted probability is sent to the terminal, so that the terminal determines the type of the vehicle to be identified according to the predicted probability. The present application can be widely used in the field of vehicle identification technology.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a vehicle identification method, identification system, device and medium. Background Art

[0002] In recent years, with the rapid development of artificial intelligence technology, various types of machine learning models have achieved relatively good application results in image classification, vehicle recognition, autonomous driving and other fields.

[0003] In the related technologies, online ride-hailing vehicles cannot be accurately identified due to incomplete information and irregular statistics of information. Existing online ride-hailing vehicle identification is mostly based on simple rules such as the mileage and driving time of new energy vehicles. However, due to the relatively simple judgment criteria, this method leads to some commercial vehicles with a long mileage being mistakenly identified as online ride-hailing vehicles. There are also some identification methods that use VR technology to identify vehicle models, and judge the probability that the vehicle belongs to an online ride-hailing vehicle based on common online ride-hailing vehicle models and vehicle passenger conditions. This method also faces the problem of low recognition rate and is prone to misjudgment.

[0004] In summary, the problems existing in related technologies need to be solved urgently. Summary of the invention

[0005] The purpose of this application is to solve one of the technical problems existing in the related art to at least a certain extent.

[0006] To this end, an object of an embodiment of the present application is to provide a vehicle identification method, which can improve the accuracy of vehicle identification, reduce the consumption of computing resources, and improve the utilization of hardware resources.

[0007] In order to achieve the above technical objectives, the technical solutions adopted by the embodiments of the present application include:

[0008] On the one hand, an embodiment of the present application provides a vehicle identification method, including:

[0009] Receiving a vehicle identification request initiated by a terminal;

[0010] Retrieving input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters;

[0011] Inputting the input parameters into a plurality of pre-trained first classification models respectively to obtain a plurality of first classification labels of the vehicle to be identified;

[0012] Obtaining a predicted probability of the to-be-identified vehicle according to the plurality of the first classification labels;

[0013] The predicted probability is sent to the terminal, so that the terminal determines the type of the vehicle to be identified according to the predicted probability.

[0014] In addition, a vehicle identification method according to the above embodiment of the present application may also have the following additional technical features:

[0015] Furthermore, in one embodiment of the present application, the step of retrieving input parameters of the vehicle to be identified includes:

[0016] Respectively obtain the WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters of the vehicle to be identified;

[0017] Standardizing the numerical features of the WIFI connection parameter, the vehicle configuration parameter, the historical travel parameter, and the step data parameter to obtain a plurality of first standardized parameters;

[0018] Encoding the category features in the WIFI connection parameter, the vehicle configuration parameter, the historical travel parameter, and the step data parameter to obtain a plurality of first encoded parameters;

[0019] The input parameter is obtained according to a plurality of the first standardized parameters and a plurality of the first encoded parameters.

[0020] Further, in one embodiment of the present application, obtaining the input parameter according to the plurality of the first standardized parameters and the plurality of the first encoded parameters includes:

[0021] Regularizing the plurality of first standardized parameters and the plurality of first encoded parameters to obtain a plurality of first weights;

[0022] Screening out a plurality of the first weights that are less than a first threshold value to obtain a plurality of second weights;

[0023] Deleting a plurality of the first standardized parameters and a plurality of the first encoded parameters corresponding to a plurality of the second weights, thereby obtaining the input parameter

[0024] Furthermore, in one embodiment of the present application, the obtaining of the WIFI connection parameter includes:

[0025] Acquire multiple historical WIFI names of the vehicle to be identified and multiple POI categories corresponding to the multiple historical WIFI names;

[0026] Training a second classification model based on the plurality of historical WIFI names and the plurality of POI categories;

[0027] The WIFI name of the vehicle to be identified is input into the trained second classification model to obtain the WIFI connection parameters of the vehicle to be identified.

[0028] Furthermore, in one embodiment of the present application, the training of the first classification model includes:

[0029] Acquire multiple training results when training the first classification model, and obtain a first ratio of positive numbers and a second ratio of negative numbers in the multiple training results, wherein the sum of the first ratio and the second ratio is 1;

[0030] According to the first ratio and the second ratio, a loss value compensation is given to the output result of the first classification model.

[0031] Further, in one embodiment of the present application, obtaining the predicted probability of the to-be-identified vehicle according to the plurality of the first classification labels includes:

[0032] Determine a first weight of each of the first classification models according to an accuracy parameter obtained when pre-training the first classification model; wherein the first weight of the first classification model with a high accuracy parameter is greater than the first weight of the first classification model with a low accuracy parameter;

[0033] The predicted probability of the vehicle to be identified is obtained according to the first weights of the plurality of the first classification models and the corresponding plurality of first classification labels.

[0034] On the other hand, the embodiment of the present application also provides another vehicle identification method, including:

[0035] Initiate a vehicle identification request to the server;

[0036] receiving the predicted probability sent by the server, so that the type of the vehicle to be identified is loaded and displayed according to the predicted probability;

[0037] The predicted probability is obtained by the server executing the following steps:

[0038] Retrieving input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters;

[0039] Inputting the input parameters into a plurality of pre-trained first classification models respectively to obtain a plurality of first classification labels of the vehicle to be identified;

[0040] The predicted probability of the vehicle to be identified is obtained according to the plurality of the first classification labels.

[0041] On the other hand, the embodiment of the present application further provides a vehicle identification system, including:

[0042] A receiving module, used for receiving a vehicle identification request initiated by a terminal;

[0043] A retrieval module, used to retrieve input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters;

[0044] A classification module, used for inputting the input parameters into a plurality of pre-trained first classification models respectively to obtain a plurality of first classification labels of the vehicle to be identified;

[0045] A prediction module, used for obtaining a prediction probability of the vehicle to be identified according to the plurality of the first classification labels;

[0046] The identification module is used to send the predicted probability to the terminal so that the terminal determines the type of the vehicle to be identified according to the predicted probability.

[0047] On the other hand, an embodiment of the present application provides a computer device, including:

[0048] at least one processor;

[0049] at least one memory for storing at least one program;

[0050] When the at least one program is executed by the at least one processor, the at least one processor implements the above-mentioned vehicle identification method.

[0051] On the other hand, an embodiment of the present application further provides a computer-readable storage medium, which stores a program executable by a processor. When the program executable by the processor is executed by the processor, it is used to implement the above-mentioned vehicle identification method.

[0052] The advantages and benefits of the present application will be partially given in the following description, and partially become apparent from the following description, or be understood through the practice of the present application:

[0053] A vehicle identification method disclosed in an embodiment of the present application is applied to vehicle identification, first receiving a vehicle identification request initiated by a terminal; retrieving input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and stepping data parameters; then respectively inputting the input parameters into a plurality of pre-trained first classification models to obtain a plurality of first classification labels of the vehicle to be identified; then obtaining the predicted probability of the vehicle to be identified based on the plurality of first classification labels; preferably sending the predicted probability to the terminal so that the terminal determines the type of the vehicle to be identified according to the predicted probability. The vehicle identification method uses multiple data sources and also integrates the predicted probabilities of multiple machine learning classification models, thereby making the prediction effect better and improving the accuracy of vehicle identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following introduction is made to the drawings of the embodiments of the present application or the related technical solutions in the prior art. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 A schematic diagram of an implementation environment of the vehicle identification method provided in an embodiment of the present application;

[0056] Figure 2 A schematic diagram of a vehicle identification method according to an embodiment of the present invention;

[0057] Figure 3 A schematic diagram of a process for retrieving input parameters of a vehicle to be identified in a vehicle identification method provided in an embodiment of the present application;

[0058] Figure 4 A schematic diagram of a specific process of obtaining input parameters for the vehicle identification method provided in an embodiment of the present application;

[0059] Figure 5 A schematic diagram of a process for obtaining WIFI connection parameters in a vehicle identification method provided in an embodiment of the present application;

[0060] Figure 6 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] The present application is further described below in conjunction with the accompanying drawings and specific embodiments. The described embodiments should not be regarded as limiting the present application, and all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0062] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0064] Before further describing the embodiments of the present application in detail, the nouns and terms involved in the embodiments of the present application are explained. The nouns and terms involved in the embodiments of the present application are subject to the following interpretations.

[0065] 1) Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines so that machines have the functions of perception, reasoning and decision-making. Artificial intelligence technology is a comprehensive discipline that covers a wide range of fields, including both hardware-level technology and software-level technology. Basic artificial intelligence technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, and mechatronics. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0066] 2) Machine Learning (ML) is a multi-disciplinary subject involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and other disciplines. It specializes in studying how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent. Its applications are spread across all areas of artificial intelligence. Machine learning (deep learning) usually includes artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and learning by teaching.

[0067] 3) Blockchain is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithm. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer. The underlying blockchain platform can include processing modules such as user management, basic services, smart contracts, and operation monitoring. Among them, the user management module is responsible for the identity information management of all blockchain participants, including maintaining the generation of public and private keys (account management), key management, and the maintenance of the corresponding relationship between the user's real identity and the blockchain address (authority management), etc., and under authorization, supervises and audits the transactions of certain real identities and provides risk control rule configuration (risk control audit); the basic service module is deployed on all blockchain node devices to verify the validity of business requests and record the valid requests in the storage after consensus is reached. For a new business request, the basic service first adapts the interface to parse and authenticate (interface adaptation), and then encrypts the business information through the consensus algorithm (consensus management). The smart contract module is responsible for the registration and issuance of contracts, contract triggering and contract execution. Developers can define the contract logic in a programming language and publish it to the blockchain (contract registration). According to the logic of the contract terms, the key or other events are called to trigger the execution and complete the contract logic. It also provides the function of contract upgrade and cancellation. The operation monitoring module is mainly responsible for the deployment, configuration modification, contract setting, cloud adaptation and real-time status visualization output of the product during the product release process, such as alarm, monitoring network status, monitoring node equipment health status, etc. The platform product service layer provides the basic capabilities and implementation framework of typical applications. Developers can implement the business logic on the blockchain based on these basic capabilities and superimpose the characteristics of the business. The application service layer provides application services based on the blockchain solution for business participants to use.

[0068] 4) The criteria for judging the quality of a classifier (prediction model) from AUC:

[0069] AUC = 1, a perfect classifier; AUC = [0.85, 0.95], very good; AUC = [0.7, 0.85], average; AUC = [0.5, 0.7], low, but good enough for predicting stocks; AUC = 0.5, just like random guessing (e.g., flipping a coin), the model has no predictive value. AUC < 0.5, worse than random guessing; but as long as you always go against the prediction, it is better than random guessing.

[0070] 5) LBS: Location Based Services, services based on geographic location data.

[0071] In recent years, with the rapid development of artificial intelligence technology, various types of machine learning models have achieved relatively good application results in image classification, vehicle recognition, autonomous driving and other fields.

[0072] In the related technologies, online ride-hailing vehicles cannot be accurately identified due to incomplete information and irregular statistics of information. Existing online ride-hailing vehicle identification is mostly based on simple calculation rules such as the mileage and driving time of new energy vehicles. However, due to the relatively simple judgment criteria, this method leads to some commercial vehicles with a long mileage being mistakenly identified as online ride-hailing vehicles. There are also some identification methods that use VR technology to identify vehicle models, and judge the probability that the vehicle belongs to an online ride-hailing vehicle based on common online ride-hailing vehicle models and vehicle passenger conditions. This method also faces the problem of low recognition rate and is prone to misjudgment.

[0073] In order to solve the problem of low recognition rate in vehicle identification in the related art, it is easy to make misjudgment, the embodiment of the present application provides a vehicle identification method, identification system, device, medium and program product. In the vehicle identification method, on the one hand, the server can receive a vehicle identification request initiated by a terminal; according to the vehicle identification request, the input parameters and prediction parameters of the vehicle to be identified are retrieved; on the other hand, the server uses multiple first classification models for training and correspondingly obtains multiple first training results according to the input parameters and prediction parameters, and obtains the prediction probability according to the multiple first training results;

[0074] Finally, the server sends the predicted probability to the terminal so that the terminal determines the type of the vehicle to be identified based on the predicted probability. It can use multiple data sources and integrate the predicted probabilities of multiple machine learning classification models to achieve better prediction results, thereby improving the accuracy of vehicle identification.

[0075] Figure 1 Schematic diagram of the implementation environment of a vehicle identification method provided in an embodiment of the present application. Figure 1 The hardware and software components of the implementation environment mainly include a terminal 101 and a server 102, and the terminal 101 is in communication with the server 102. The vehicle identification method can be executed based on the interaction between the terminal 101 and the server 102. In addition, the terminal 101 and the server 102 can be nodes in a blockchain, which is not specifically limited in this embodiment.

[0076] Specifically, the terminal 101 in the present application may include but is not limited to any one or more of a smart camera device, a smart watch, a smart phone, a computer, a personal digital assistant (PDA), an intelligent voice interaction device, a smart home appliance or a vehicle-mounted terminal. The server 102 may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The terminal 101 and the server 102 can establish a communication connection through a wireless network or a wired network. The wireless network or wired network uses standard communication technology and / or protocols. The network can be set to the Internet or any other network, such as but not limited to a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a private network or a virtual private network. Any combination. Of course, it is understandable that Figure 1 The implementation environment in the embodiment of the present application is only an optional application scenario of the vehicle identification method provided in the embodiment of the present application, and the actual application is not fixed to Figure 1 The hardware and software environment shown.

[0077] Combine the following Figure 1 The implementation environment shown is used to describe in detail the vehicle identification method provided in the embodiments of the present application.

[0078] Reference Figure 2 , Figure 2 is a flow chart of a vehicle identification method provided in an embodiment of the present application, Figure 2The vehicle identification method shown is mainly executed on the server 102. The method can be applied to the relevant computer equipment in the server 102, but is not limited to the above form. Figure 2 The method includes but is not limited to steps 210 to 230.

[0079] Step 210: Receive a vehicle identification request initiated by the terminal.

[0080] In this step, the terminal refers to the relevant equipment that performs the vehicle identification task. The specific equipment types have been illustrated in the previous examples and will not be repeated here. The terminal in the embodiment of the present application mainly performs the vehicle identification task based on the vehicle identification model, so the terminal needs to be configured with the relevant model. Specifically, the vehicle identification model in the embodiment of the present application can be uniformly trained in the server, without having to be trained separately in the terminal, which can reduce the number of training times for the same or similar application models and save computing resources; moreover, for the terminal, it does not have to train from scratch to apply the trained vehicle identification model, which can greatly improve the application efficiency of vehicle identification and reduce the pressure on the terminal's hardware equipment.

[0081] In this step, when the server performs vehicle identification for the terminal, it can receive a vehicle identification request initiated by the terminal. On the one hand, the vehicle identification request indicates that the terminal has a need for vehicle identification. On the other hand, the vehicle identification request also includes the hardware architecture information of the terminal, which may include the number of cores of the terminal's processor, graphics card specifications and number, memory size, and other information. Based on the hardware architecture information, the server can determine the file format of the model adapted to the terminal, so as to facilitate the configuration of the trained vehicle identification model for the terminal and perform related vehicle identification tasks.

[0082] It should be noted that in the embodiments of the present application, the vehicle identification request sent to the terminal is not limited to the initial configuration. In other words, the implementation of the present application includes both the initial configuration of the model of the vehicle identification terminal that has not been configured with the vehicle identification model before, and the model update configuration of the terminal that has been configured with the vehicle identification model before. The specific implementation application can be flexibly selected according to the needs.

[0083] Step 220: Retrieve input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters.

[0084] In this step, according to the vehicle identification request in step 210, the input variables of the vehicle to be identified are retrieved. The input variables of this solution come from different categories of data sources, generally including at least WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters. By retrieving different categories of data sources, the imbalance caused by the data set is avoided from affecting the prediction effect of the model, the prediction accuracy is improved, and thus the accuracy of vehicle identification is improved. In the embodiment of the present application, by obtaining different multiple data sources, the specific conditions of each data source can be combined to make the prediction effect better.

[0085] Specifically, in step 220, the vehicle configuration parameters may be common vehicle parameters. For example, data related to new energy vehicle models may include the number of seats, specific parameters of the batteries used, vehicle brand, whether the vehicle is a high-end model, the purchase price of a new car, etc.

[0086] Specifically, in step 220, the historical travel parameters may be historical travel data, such as the age of the new energy vehicle, mileage, fatigue driving time, average speed, number of charging times, number of vehicle failures, etc.

[0087] Specifically, in step 220, the location data parameter may be a category of the vehicle's travel space range. For example, by classifying the map POIs, it is possible to identify whether the geographical locations frequently visited by the driver are near public places such as train stations, airports, shopping malls and hotels, and the approximate length of stay in each public place, thereby obtaining the location data parameters.

[0088] In addition, the specific limitations of the WIFI connection parameters in step 220 are described in detail below and are not specifically limited here.

[0089] Step 230: Input the input parameters into a plurality of pre-trained first classification models respectively to obtain a plurality of first classification labels of the vehicle to be identified.

[0090] In this step, multiple first classification models are used for training according to the input parameters obtained in step 220, and the input parameters of the vehicle to be identified are calculated according to the trained multiple first classification models, and each first classification model inputs a first classification label.

[0091] In the embodiment of the present application, the input parameters and output parameters of the historical vehicles to be identified are used to train multiple first classification models. The input parameters of the historical vehicles to be identified are the same as the data included in the input parameters obtained in step 220, and the output parameters of the historical vehicles to be identified can be the probability that the vehicle to be identified is an online-hailing vehicle, and the probability interval is [0-1].

[0092] In the embodiments of the present application, the classification and recognition models can be built using any machine learning algorithm, without limitation. Machine Learning (ML) is a multi-disciplinary cross-discipline involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and other disciplines. It specializes in how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent. Its applications are spread across all areas of artificial intelligence. Machine learning (deep learning) usually includes artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and teaching learning.

[0093] Specifically, before using the above-mentioned machine learning model, it needs to be trained based on supervised learning. In the embodiment of the present application, it can be trained by training images. After preprocessing the training images, they can be randomly divided into training sets and test sets, for example, using a 5-fold cross-validation method to divide them, that is, taking 80% of the training images as the training set and taking the remaining 20% ​​of the training images as the test set.

[0094] For the training images in the training set, they can be input into the initialized vehicle recognition model to obtain the prediction result output by the vehicle recognition model, which is recorded as the first recognition result. After obtaining the first recognition result, the accuracy of the model recognition can be evaluated based on the result and the aforementioned label to perform back propagation training on the model and update its relevant parameters.

[0095] Specifically, for a machine learning model, the accuracy of its prediction results can be measured by a loss function (LossFunction). The loss function is defined on a single training data and is used to measure the prediction error of a training data. Specifically, the loss value of the training data is determined by the label of a single training data and the prediction result of the model for the training data. In actual training, a training data set has a lot of training data, so the cost function (CostFunction) is generally used to measure the overall error of the training data set. The cost function is defined on the entire training data set and is used to calculate the average value of the prediction error of all training data, which can better measure the prediction effect of the model. For a general machine learning model, based on the aforementioned cost function, plus the regularization term that measures the complexity of the model, it can be used as the objective function of the training. Based on this objective function, the loss value of the entire training data set can be calculated. There are many types of commonly used loss functions, such as 0-1 loss function, square loss function, absolute loss function, logarithmic loss function, cross entropy loss function, etc., which can all be used as loss functions of machine learning models, which will not be explained one by one here. In practical applications, you can choose any one of the loss functions to determine the loss value of the training, that is, the loss value between the prediction result and the label. Based on the training loss value, the back propagation algorithm is used to update the model parameters, and the trained machine learning model can be obtained by iterating the preset rounds. Through the above training method, a trained recognition and classification model can be obtained.

[0096] In an embodiment of the present application, the first classification model may include a Bayesian classifier, a random forest classifier, an Xgboost classifier, a lightgbm classifier, etc.

[0097] Step 240: Obtain the predicted probability of the vehicle to be identified based on the plurality of the first classification labels.

[0098] In this step, the predicted probability of vehicle identification is obtained based on multiple first classification labels, for example, the probability that the vehicle is an online-hailing car. Specifically, the input parameters are respectively input into multiple first classification models, and the multiple first classification models respectively output multiple first classification labels, and the predicted probability of the vehicle to be identified is obtained based on the multiple first classification labels. It can be understood that the predicted probability of the vehicle to be identified can be obtained by taking the average value of the multiple first classification labels.

[0099] Step 250: Send the predicted probability to the terminal, so that the terminal determines the type of the vehicle to be identified according to the predicted probability.

[0100] In this step, the server sends the predicted probability to the terminal, and the display device of the terminal will display the predicted probability that the vehicle to be identified is an online-hailing car. For example, the predicted probability is 80% or 90%. The terminal user can determine whether the vehicle to be identified is an online-hailing car or the possibility of it being an online-hailing car based on the predicted probability.

[0101] The vehicle identification method in the embodiment of the present application adopts input parameters of multiple data sources including WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters in step 220; includes buried data parameters of multiple data sources corresponding to the vehicle to be identified; and also integrates the prediction probabilities of multiple first classification models to make the prediction effect better, thereby improving the accuracy of vehicle identification, and its AUC is as high as 86.54%.

[0102] In some embodiments of this application, see Figure 3 , retrieve the input parameters of the vehicle to be identified, including the following steps:

[0103] Step 321: Obtain WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters;

[0104] Step 322: Standardize the numerical features of the WIFI connection parameter, vehicle configuration parameter, historical travel parameter and step data parameter of the vehicle to be identified to obtain a first standardized parameter;

[0105] Step 323: Encoding the category features in the WIFI connection parameter, the vehicle configuration parameter, the historical travel parameter, and the step data parameter to obtain a first encoded parameter;

[0106] Step 324: Obtain the input parameter according to the first standardized parameter and the first encoded parameter.

[0107] Specifically, after obtaining the WIFI connection parameters, vehicle configuration parameters, historical travel parameters and stepping data parameters, the parameters are classified into numerical features and category features. For numerical features, min-max scaling and other methods can be used for standardization. For category features, dummy variables are used for encoding. Specifically, the python feature engineering toolkit can be used to encode category features.

[0108] In some embodiments of this application, see Figure 4 , obtaining input parameters according to a plurality of first standardized parameters and a plurality of first encoded parameters, including:

[0109] Step 410: Regularizing the plurality of first standardized parameters and the plurality of first encoded parameters to obtain a plurality of first weights;

[0110] Step 420: Filter out a plurality of first weights that are less than a first threshold value to obtain a plurality of second weights;

[0111] Step 430: Delete the multiple first standardized parameters and the multiple first encoded parameters corresponding to the multiple second weights, so as to obtain the input parameters.

[0112] It is worth noting that the first standardized parameter and the first encoded parameter obtained after processing are regularized by the regularization model to obtain the first weight value corresponding to each parameter, and multiple first weight values ​​close to the first threshold are named as second weights, where the first threshold can be 0 or other values ​​close to 0, such as 0.01 or 0.001. The corresponding first standardized parameters and first encoded parameters of multiple second weights and their corresponding feature parameters are obtained, where the feature parameters can include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and stepping data parameters, and these feature parameters are deleted to obtain the filtered input parameters.

[0113] In some embodiments of the present application, see Figure 5 , get WIFI connection parameters, including:

[0114] Step 510: Acquire multiple historical WIFI names of the vehicle to be identified and multiple POI categories corresponding to the multiple historical WIFI names;

[0115] Step 520: training a second classification model according to a plurality of historical WIFI names and a plurality of POI categories;

[0116] Step 530: Input the WIFI name of the vehicle to be identified into the trained second classification model to obtain the WIFI connection parameters of the vehicle to be identified.

[0117] Specifically, when obtaining the WIFI connection parameters, the driver obtains the WiFi-related data, i.e., the WIFI connection parameters, of the WiFi passed by during the historical driving process, wherein the WIFI connection parameters may include the name of the WiFi passed by, the longitude and latitude of the WiFi, etc. It can be understood that the name of the WiFi passed by and the longitude and latitude of the WiFi are only common WIFI connection parameters. In the embodiment of the present application, the WIFI connection parameters may be any commonly used characteristic parameters that can reflect the WIFI connection status, and the present application does not limit this.

[0118] Specifically, firstly, feature extraction is performed based on the WIFI name. In this process, text analysis can be used to simply classify the WIFI category. For example, the WIFI category can be divided into categories related to public places such as restaurants, railway stations, and airports according to the WIFI name. Furthermore, the embodiment of the present application pre-establishes some simple manual rules to identify the category of WIFI. For example, the words "cafe" in the WiFi name belong to the category of restaurants, and the words "railway station" and "airport" belong to the categories of railway stations and airports, respectively.

[0119] In addition to identifying the classification labels of WIFI based on the characters such as WIFI names clearly defined in these texts through predefined classification rules, this solution also uses the classification model in machine learning to classify WIFI, thereby improving the coverage of the classification model. For example, the known POI (abbreviation of Point of Interest) categories and their corresponding WIFI names and other information are first integrated and cleaned, and the cleaned WIFI name is used as the text factor x in the training data, and its corresponding POI classification category is y to train the classification model. The trained classification model is used to identify what type of WIFI corresponds to the remaining unknown POI information. Specifically, for example, when a driver often appears within the WIFI range of a train station or airport, the latitude and longitude and WIFI name can be used to use the classification model to determine that it is a WIFI corresponding to the public transportation category. Then the driver's WIFI information will be classified as a public transportation WIFI, and the modeling will then add the factor of the WIFI category that this driver often appears in. It is worth noting that map POIs are usually called points of interest, which generally refer to point data in Internet electronic maps, which basically include four attributes: name, address, coordinates, and category. According to the category attributes described in the point of interest data, a vector data set of each attribute is constructed based on the point map in the basic surveying and mapping results DLG (Digital Line Graphic) product. According to the vector data set, objects that can be abstracted into points for management, analysis, and calculation in GIS (Geographic Information System), namely, the stepping data parameters. The relevant data of the stepping data parameters of the driver's driving can also be reported for stepping data classification. The stepping data parameters can be obtained by identifying whether the geographical location frequently visited by the driver is located in public places such as railway stations, airports, shopping malls and hotels, as well as the approximate stay time in each public place.

[0120] In some embodiments, training the first classification model includes:

[0121] Acquire multiple training results when training the first classification model, and obtain a first ratio of positive numbers and a second ratio of negative numbers in the multiple training results, wherein the sum of the first ratio and the second ratio is 1;

[0122] According to the first ratio and the second ratio, a loss value compensation is given to the output result of the first classification model.

[0123] Specifically, multiple training results when training the first classification model are obtained, and a first ratio of positive numbers and a second ratio of negative numbers in the multiple training results are obtained, wherein the sum of the first ratio and the second ratio is 1; if the ratio of the first ratio to the second ratio is less than 1, the output result of the first classification model is compensated for the loss value, thereby completing the training of the first classification model. When the input parameters of the vehicle to be identified are input into the trained first classification model, the first classification model will automatically perform weight loss compensation on the output prediction result, thereby obtaining the first classification label.

[0124] In the above embodiment, in order to prevent the class imbalance in the data set from affecting the prediction effect of the classification model, this solution also assigns positive weights to sample data whose prediction results are online ride-hailing cars, and assigns negative weights to sample data whose prediction results are not online ride-hailing cars. The specific basis for the assignment can be the proportion of positive and negative samples in the total sample data. This assignment prevents the classification model from classifying all samples as not online ride-hailing drivers because of the pursuit of too high accuracy.

[0125] In some embodiments, obtaining a predicted probability of a vehicle to be identified according to a plurality of first classification labels includes:

[0126] Determine a first weight of each first classification model according to an accuracy parameter obtained when pre-training the first classification model; wherein the first weight of a first classification model with a high accuracy parameter is greater than the first weight of a first classification model with a low accuracy parameter;

[0127] According to the first weights of the plurality of first classification models and the corresponding plurality of first classification labels, a predicted probability of the vehicle to be identified is obtained.

[0128] Specifically, the first weight of each first classification model is determined according to the accuracy parameter of the training result when training each first classification model; wherein the first weight of the first classification model with a high accuracy parameter is greater than the first weight of the first classification model with a low accuracy parameter; according to the first weights of multiple first classification models and the corresponding multiple first classification labels, the first classification label output by each first classification model and the corresponding first weight value can be weightedly summed, and then the weighted average of each first classification label is used as the predicted probability of the vehicle to be identified.

[0129] Specifically, since the correlation between multiple prediction results obtained by using multiple first classification models is too high, the generalization ability of the entire vehicle identification model will be reduced. Therefore, each first classification model is assigned a predetermined weight between 0 and 1, and the model results with correlations lower than the predetermined weight in the output results of each first classification model can be fused by an integration method. Specifically, if the correlation between the prediction results of two first classification models is higher than 0.8, both prediction results will be discarded, and other first classification models will continue to output prediction results. Then, a weighted score will be given to each first classification model based on the accuracy of these prediction results (the first classification model with a high accuracy rate will be assigned a relatively large score ratio). Finally, the prediction results of all first classification models are summed up by weighted average to obtain the predicted probability that each car is an online car-hailing vehicle.

[0130] In an embodiment of the present application, a vehicle identification method is also provided, which is performed based on the interaction between the terminal 101 and the server 102. The method can be applied to the relevant software or computer equipment in the terminal 101 and the server 102, but is not limited to the above form. The method mainly includes the following steps:

[0131] Initiate a vehicle identification request to the server;

[0132] receiving the predicted probability sent by the server, so that the type of the vehicle to be identified is loaded and displayed according to the predicted probability;

[0133] The predicted probability is obtained by the server performing the following steps:

[0134] Retrieving input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters;

[0135] Inputting the input parameters into a plurality of first classification models respectively, and obtaining a plurality of first classification labels of the vehicles to be identified through the plurality of first classification models;

[0136] A predicted probability of a vehicle to be identified is obtained according to the multiple first classification labels.

[0137] based on Figure 2 The contents of the vehicle identification method embodiment shown in the figure can be understood from the contents of the vehicle identification method embodiment, so it will not be repeated here. The functions specifically implemented by the method embodiment are similar to those of the vehicle identification method embodiment. Figure 2 The vehicle identification method embodiment shown is the same as that of the embodiment of the present invention, and the beneficial effects achieved are the same as those of the embodiment of the present invention. Figure 2 The beneficial effects achieved by the vehicle identification method embodiment shown are also the same.

[0138] Based on the above configuration method, the embodiment of the present application further provides a vehicle identification system, including:

[0139] A receiving module, used for receiving a vehicle identification request initiated by a terminal;

[0140] The retrieval module is used to retrieve the input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters;

[0141] A classification module, used for inputting the input parameters into a plurality of first classification models respectively, and obtaining a plurality of first classification labels of the vehicle to be identified through the plurality of first classification models;

[0142] A prediction module, used to obtain a prediction probability of a vehicle to be identified according to a plurality of first classification labels;

[0143] The identification module is used to send the predicted probability to the terminal so that the terminal determines the type of the vehicle to be identified based on the predicted probability.

[0144] based on Figure 2 The contents of the vehicle identification method embodiment shown in the figure can be used to understand the contents of the vehicle identification system embodiment, so it will not be repeated here. Figure 2 The vehicle identification method embodiment shown is the same as that of the embodiment of the present invention, and the beneficial effects achieved are the same as those of the embodiment of the present invention. Figure 2 The beneficial effects achieved by the vehicle identification method embodiment shown are also the same.

[0145] Reference Figure 6 , the embodiment of the present application further discloses a computer device, including:

[0146] at least one processor 610;

[0147] At least one memory 620, used to store at least one program;

[0148] When at least one program is executed by at least one processor 610, the at least one processor 610 implements the following Figure 2 An embodiment of a vehicle identification method is shown.

[0149] It is understandable that if Figure 2 The contents of the vehicle identification method embodiment shown in the figure are applicable to the computer device embodiment. The functions specifically implemented by the computer device embodiment are similar to those in the embodiment shown in the figure. Figure 2 The vehicle identification method embodiment shown is the same as that shown in FIG. Figure 2 The beneficial effects achieved by the vehicle identification method embodiment shown are also the same.

[0150] The present application also discloses a computer-readable storage medium in which a program executable by a processor is stored. When the program executable by the processor is executed by the processor, it is used to implement the following Figure 2 An embodiment of a vehicle identification method is shown.

[0151] It is understandable that if Figure 2 The contents of the vehicle identification method embodiment shown in the figure are applicable to the computer-readable storage medium embodiment. The functions specifically implemented by the computer-readable storage medium embodiment are similar to those in the embodiment shown in the figure. Figure 2 The vehicle identification method embodiment shown is the same as that shown in FIG. Figure 2 The beneficial effects achieved by the vehicle identification method embodiment shown are also the same.

[0152] In some selectable embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the application is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part described as a larger operation is performed independently.

[0153] In addition, although the present application is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and / or features can be integrated into a single physical device and / or software module, or one or more functions and / or features can be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present application. More specifically, in view of the properties, functions, and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional techniques of the engineer. Therefore, those skilled in the art can implement the present application set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present application, which is determined by the full scope of the attached claims and their equivalents.

[0154] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0155] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0156] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or processing in another suitable manner as necessary, and then stored in a computer memory.

[0157] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0158] In the above description of this specification, the description with reference to the terms "one embodiment / example", "another embodiment / example" or "certain embodiments / examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0159] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present application, and that the scope of the present application is defined by the claims and their equivalents.

[0160] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present application.

[0161] In the description of this specification, the description with reference to the terms "one embodiment", "another embodiment" or "certain embodiments" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example.

Claims

1. A vehicle identification method, characterized in that: The method is applied to a server, and the method comprises: Receiving a vehicle identification request initiated by a terminal; Retrieving input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters; Inputting the input parameters into a plurality of pre-trained first classification models respectively to obtain a plurality of first classification labels of the vehicle to be identified; Obtaining a predicted probability of the to-be-identified vehicle according to the plurality of the first classification labels; Sending the predicted probability to the terminal, so that the terminal determines whether the vehicle to be identified is an online-hailing vehicle according to the predicted probability; The step of retrieving input parameters of the vehicle to be identified according to the vehicle identification request includes: According to the vehicle identification request, the WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters of the vehicle to be identified are respectively obtained; Standardizing the numerical features of the WIFI connection parameter, the vehicle configuration parameter, the historical travel parameter, and the step data parameter to obtain a plurality of first standardized parameters; Encoding the category features in the WIFI connection parameter, the vehicle configuration parameter, the historical travel parameter, and the step data parameter to obtain a plurality of first encoded parameters; The input parameter is obtained according to a plurality of the first standardized parameters and a plurality of the first encoded parameters.

2. The vehicle identification method according to claim 1, characterized in that: The step of obtaining the input parameter according to the plurality of the first standardized parameters and the plurality of the first encoded parameters comprises: Regularizing the plurality of first standardized parameters and the plurality of first encoded parameters to obtain a plurality of first weights; Screening out a plurality of the first weights that are less than a first threshold value to obtain a plurality of second weights; Delete multiple first standardized parameters and multiple first encoded parameters corresponding to multiple second weights, so as to obtain the input parameters.

3. The vehicle identification method according to claim 1, characterized in that: The obtaining of the WIFI connection parameter comprises: Acquire multiple historical WIFI names of the vehicle to be identified and multiple POI categories corresponding to the multiple historical WIFI names; Training a second classification model based on the plurality of historical WIFI names and the plurality of POI categories; The WIFI name of the vehicle to be identified is input into the trained second classification model to obtain the WIFI connection parameters of the vehicle to be identified.

4. The vehicle identification method according to claim 1, characterized in that: The training of the first classification model includes: Acquire multiple training results when training the first classification model, and obtain a first ratio of positive numbers and a second ratio of negative numbers in the multiple training results, wherein the sum of the first ratio and the second ratio is 1; According to the first ratio and the second ratio, a loss value compensation is given to the output result of the first classification model.

5. The vehicle identification method according to claim 1, characterized in that: The step of obtaining the predicted probability of the to-be-identified vehicle according to the plurality of the first classification labels includes: Determine a first weight of each of the first classification models according to an accuracy parameter obtained when pre-training the first classification model; wherein the first weight of the first classification model with a high accuracy parameter is greater than the first weight of the first classification model with a low accuracy parameter; The predicted probability of the vehicle to be identified is obtained according to the first weights of the plurality of the first classification models and the corresponding plurality of first classification labels.

6. A vehicle identification method, characterized in that: The method is applied to a terminal, and the method includes: Initiate a vehicle identification request to the server; Receiving the predicted probability sent by the server; wherein the predicted probability is obtained according to any one of the methods of claims 1 to 5; Loading and displaying whether the vehicle to be identified is an online-hailing vehicle according to the predicted probability; The predicted probability is obtained by the server executing the following steps: Retrieving input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters; Inputting the input parameters into a plurality of pre-trained first classification models respectively to obtain a plurality of first classification labels of the vehicle to be identified; The predicted probability of the vehicle to be identified is obtained according to the plurality of the first classification labels.

7. A vehicle identification system, characterized in that: include: A receiving module, used for receiving a vehicle identification request initiated by a terminal; A retrieval module, used to retrieve input parameters of the vehicle to be identified according to the vehicle identification request; wherein the input parameters include WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters; A classification module, used for inputting the input parameters into a plurality of pre-trained first classification models respectively to obtain a plurality of first classification labels of the vehicle to be identified; A prediction module, used for obtaining a prediction probability of the vehicle to be identified according to the plurality of the first classification labels; An identification module, used to send the predicted probability to the terminal, so that the terminal determines whether the vehicle to be identified is an online-hailing vehicle according to the predicted probability; The retrieval module is used to retrieve the input parameters of the vehicle to be identified according to the vehicle identification request, including: According to the vehicle identification request, the WIFI connection parameters, vehicle configuration parameters, historical travel parameters and step data parameters of the vehicle to be identified are respectively obtained; Standardizing the numerical features of the WIFI connection parameter, the vehicle configuration parameter, the historical travel parameter, and the step data parameter to obtain a plurality of first standardized parameters; Encoding the category features in the WIFI connection parameter, the vehicle configuration parameter, the historical travel parameter, and the step data parameter to obtain a plurality of first encoded parameters; The input parameter is obtained according to a plurality of the first standardized parameters and a plurality of the first encoded parameters.

8. A computer device, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the vehicle identification method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a program executable by a processor, characterized in that: The program executable by the processor is used to implement the vehicle identification method according to any one of claims 1 to 6 when executed by the processor.

Images