WO2021146906A1

WO2021146906A1 - Test scenario simulation method and apparatus, computer device, and storage medium

Info

Publication number: WO2021146906A1
Application number: PCT/CN2020/073476
Authority: WO
Inventors: 刘宇辰
Original assignee: 深圳元戎启行科技有限公司
Priority date: 2020-01-21
Filing date: 2020-01-21
Publication date: 2021-07-29
Also published as: CN113498511A

Abstract

A test scenario simulation method, comprising: acquiring feature information corresponding to each object in a test scenario; acquiring a field template, the field template comprising a plurality of candidate fields; on the basis of the feature information corresponding to each object, determining a target field of each object from the field template and, on the basis of the feature information corresponding to each object, assigning a value to the target field of each object; on the basis of the target field of each object after value assignment, generating target code corresponding to the target field of each object and, on the basis of the target code corresponding to each object, generating a simulation model corresponding to each object; and, on the basis of the simulation model corresponding to each object, establishing a simulation scenario corresponding to the test scenario.

Description

Test scene simulation method, device, computer equipment and storage medium

Technical field

This application relates to the field of automatic driving technology, and in particular to a test scenario simulation method, device, computer equipment, and storage medium.

Background technique

With the development of vehicle control technology, unmanned driving technology has emerged. Unmanned driving technology is to automatically plan the driving route of the unmanned vehicle based on the unmanned driving algorithm, and control the unmanned vehicle based on the driving route, so that the unmanned vehicle can reach the predetermined target location. The test and evaluation of unmanned driving algorithms is mainly based on the simulation of real test scenarios.

The existing simulation methods of real test scenes often classify objects that appear on the road in real life, such as cars, trucks, bicycles, pedestrians, etc., and write corresponding codes to simulate and simulate each object according to the function of each type of object. However, different types of objects have the same function, so there are a lot of repeated codes between different types of objects, resulting in low code generation efficiency.

Summary of the invention

The various embodiments provided in this application provide a test scenario simulation method, device, computer equipment, and storage medium. The technical solution is as follows:

A simulation method for test scenarios, including:

Obtain feature information corresponding to each object in the test scene;

Obtain a field template, the field template includes multiple candidate fields;

Determine the target field of each object from the field template according to the feature information corresponding to each object, and assign a value to the target field of each object according to the feature information corresponding to each object;

Generate the target code corresponding to the target field of each object according to the target field of each object after assignment, and generate the simulation model corresponding to each object according to the target code corresponding to each object;

The simulation scene corresponding to the test scene is established according to the simulation model corresponding to each object.

In one of the embodiments, before obtaining the field template, the method further includes:

Obtain historical simulation data, extract the fields of each historical simulation model from the historical simulation data, and obtain the field set corresponding to each historical simulation model; obtain the fields in the field set corresponding to each historical simulation model; calculate the repetition rate of each field; The fields whose repetition rate is greater than the preset threshold constitute a field template.

In one of the embodiments, the target field of each object is determined from the field template according to the characteristic information corresponding to each object, and the target field of each object is assigned according to the characteristic information corresponding to each object, including:

Determine the feature field and feature field value corresponding to each object according to the feature information corresponding to each object; match the candidate field in the field template with the feature field corresponding to each object; when the matching is successful, match the feature field corresponding to each object The successful candidate field is used as the target field of each object; the target field of each object is assigned according to the characteristic field value corresponding to each object.

In one of the embodiments, it further includes:

When the matching fails, the feature field that fails to match is used as the update candidate field;

Add the update candidate field to the field template.

In one of the embodiments, before determining the feature field and feature field value corresponding to each object according to the feature information corresponding to each object, the method further includes:

The feature information corresponding to each object is input into the pre-trained deep learning neural network, and the feature field and feature field value corresponding to each object are output.

In one of the embodiments, before inputting the feature information corresponding to each object into the pre-trained deep learning neural network, the method further includes:

Obtain object sample data; obtain the feature information, fields and field values of the sample objects according to the object sample data to train the deep learning neural network.

In one of the embodiments, acquiring the feature information, fields, and field values of the sample objects according to the object sample data to train the deep learning neural network includes:

Obtain the feature information of the object according to the object sample data, and input the feature information of the object into the deep learning neural network for unsupervised training; obtain the field and field value corresponding to the feature information of the sample object from the object sample data, and use the obtained sample The feature information of the object is used as the input data of the deep learning neural network, the obtained fields and field values are used as the expected output of the deep learning neural network, and the deep learning neural network is supervised training.

A test scenario simulation device, including:

The obtaining module is used to obtain characteristic information corresponding to each object in the test scene; obtain a field template, which includes multiple candidate fields;

The target field determination module is configured to determine the target field of each object from the field template according to the feature information corresponding to each object, and assign a value to the target field of each object according to the feature information corresponding to each object;

The simulation model generation module is used to generate the target code corresponding to the target field of each object according to the target field of each object after the assignment, and generate the simulation model corresponding to each object according to the target code corresponding to each object;

The simulation scene establishment module is used to establish a simulation scene corresponding to the test scene according to the simulation model corresponding to each object.

In one aspect, a computer device is provided. The computer device includes a processor and a memory, and computer-readable instructions are stored in the memory. When the computer-readable instructions are executed by the processor, the processor Perform the following steps:

Obtain feature information corresponding to each object in the test scene;

Obtain a field template, the field template includes multiple candidate fields;

In one aspect, one or more non-volatile storage media storing computer-readable instructions are provided. When the computer-readable instructions are executed by one or more processors, the one or more processors perform the following step:

Obtain feature information corresponding to each object in the test scene;

Obtain a field template, the field template includes multiple candidate fields;

The above-mentioned test scene simulation method, device, computer equipment and storage medium obtain the characteristic information corresponding to each object in the test scene; obtain a field template, which includes multiple candidate fields; determine from the field template according to the characteristic information corresponding to each object The target field of each object is assigned to the target field of each object according to the characteristic information corresponding to each object; the target code corresponding to the target field of each object is generated according to the target field of each object after the assignment, and the target code corresponding to each object is generated The simulation model corresponding to each object; the simulation scene corresponding to the test scene is established according to the simulation model corresponding to each object. In this way, the field template includes multiple candidate fields, and each candidate field has a corresponding code. Since the feature information corresponding to the object can indicate the function corresponding to the object, the function corresponding to the object can be realized by assigning the value of the target field, and the target code is automatically generated according to the target field, and then the simulation model corresponding to the object is obtained without changing the code, and Redundant codes are reduced, code maintenance is reduced, and code generation efficiency is improved.

Description of the drawings

In order to better describe and illustrate the embodiments and/or examples of those applications disclosed herein, one or more drawings may be referred to. The additional details or examples used to describe the drawings should not be considered as limiting the scope of any of the disclosed inventions, the currently described embodiments and/or examples, and the best mode of these applications currently understood.

Fig. 1 is a schematic diagram of an application environment of a test scenario simulation method in an embodiment.

Fig. 2 is a schematic flowchart of a test scenario simulation method in an embodiment.

FIG. 3 is a flow diagram of the steps of determining the target field of each object from the field template according to the characteristic information corresponding to each object, and assigning the target field of each object according to the characteristic information corresponding to each object in an embodiment.

Fig. 4 is a schematic flowchart of a test scenario simulation method in another embodiment.

Figure 5 is a structural block diagram of a test scenario simulation device in an embodiment

Fig. 6 is a schematic diagram of the internal structure of the server in an embodiment.

Detailed ways

In order to make the purpose, technical solutions, and advantages of this application clearer and clearer, the following further describes the application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application, and are not used to limit the present application.

It can be understood that the terms "first", "second", etc. used in the embodiments of the present application can be used herein to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from another element. For example, without departing from the scope of the present application, the first control may be referred to as the second control, and both the first control and the second control are controls, but they are not the same control.

Fig. 1 is an application environment diagram of a test scenario simulation method in an embodiment. As shown in FIG. 1, the application environment includes a terminal 102 and a server 104. The terminal 102 may be a desktop terminal or a mobile terminal, and the mobile terminal may be at least one of a mobile phone, a tablet computer, and a notebook computer. The server 104 may be a single server or a server cluster, and the terminal 102 and the server 104 communicate through a network.

Specifically, the server 104 obtains the characteristic information corresponding to each object in the test scene from the terminal 102. The server 104 obtains a field template, which includes multiple candidate fields, determines the target field of each object from the field template according to the feature information corresponding to each object, and assigns the target field of each object according to the feature information corresponding to each object. The server 104 generates the target code corresponding to the target field of each object according to the assigned target field of each object, generates the simulation model corresponding to each object according to the target code corresponding to each object, and establishes the simulation corresponding to the test scene according to the simulation model corresponding to each object Scenes.

It can be understood that the above application scenario is only an example, and does not constitute a limitation on the data processing method of this application. For example, the data processing method provided by this application may also be executed in a terminal.

Fig. 2 is a flowchart of a test scenario simulation method in an embodiment. As shown in Fig. 2, a test scenario simulation method, which is applied to the server in Fig. 1 as an example, is described, which specifically includes:

S202: Acquire feature information corresponding to each object in the test scene.

Among them, the test scenario is a collection of surrounding objects and environmental states during a certain period of time during the operation of the autonomous vehicle. Self-driving cars driving in different test scenarios can test the robustness of the self-driving algorithm. The test scene can be a real scene or a virtual extreme scene. Test scenes include, but are not limited to, school scenes, dense pedestrian scenes, rural scenes, tunnel scenes, cross scenes (such as T-shaped cross scenes, cross scenes, and roundabout scenes). The extreme scene may be an extreme weather scene, an extreme disaster scene, etc., or a scene including objects with abnormal behaviors.

Objects are substances that exhibit corresponding characteristics or perform corresponding operations according to the course of time or the course of events. Objects include, but are not limited to, objects that can react to the surroundings according to the road at a certain moment and state, and objects that change or do not change regularly. Objects that react to the surroundings according to the road at a certain moment and state include but are not limited to pedestrians, bicycles, small vehicles, large vehicles, etc. Objects that change or do not change regularly include, but are not limited to, roadblocks, traffic lights, traffic signs, etc.

Feature information is information used to describe the features and attributes of an object. The feature information includes, but is not limited to, the description of the behavior of the object, the description of the position of the object, the description of the shape of the object, and so on. For example, the characteristic information of a pedestrian crossing signal light can be: coordinates (x1, y1, z1), height 2m, red light flashes at 1HZ for 15s and then switches to green light, and green light flashes at 1HZ for 15s and then switches to red light. The characteristic information of the bicycle may be: coordinates (x2, y2, z2), the body length is 1.7m, the width is 0.7m, the height is 1.0m, and the speed is 16km/h.

Specifically, the server may obtain the characteristic information of each object in the test scene from the terminal. For example, when the test scene is a real scene, the server can receive sensor data uploaded by the sensors of an autonomous vehicle, and determine the characteristic information of each object in the real scene based on the sensor data. When the test scene is an extreme scene, the user can input the characteristic information of each object in the extreme scene at the terminal, and the terminal sends the characteristic information of each object in the extreme scene to the server.

In one embodiment, the self-driving car carries various sensors, such as lidar, millimeter wave radar, and cameras. The sensor data can include pictures collected by a camera. After receiving the pictures, the server recognizes the pictures, and the shape and structure of the objects in the pictures can be recognized. For example, the surrounding vehicles, pedestrians, road markings, traffic sign text, traffic lights, etc. can be identified based on the picture. Sensor data can also include data collected by lidar. The data collected by the lidar can assist the pictures collected by the camera, so as to obtain the distance of each object from the lidar and the movement speed of each object. Combining the high-precision map and the data collected by various sensors can determine the characteristic information of each object in the real scene.

S204: Obtain a field template, where the field template includes multiple candidate fields.

Among them, the field template is a template composed of multiple candidate fields. The field value corresponding to each candidate field is empty by default. Candidate fields are keywords used to determine a certain function of an object and its expression, such as ID (Identity Document), Size, Velocity, etc. According to the different description dimensions and methods required by specific functions, the content that needs to be filled in the candidate fields is also different. The corresponding function can be realized according to the candidate field filled in the corresponding field value, and the object with the corresponding function can be simulated.

Specifically, the server stores a field template, the field template includes a plurality of preset candidate fields, and each candidate field provides a blank default value. Each candidate field is optional, so the field template is highly extensible.

In an embodiment, the candidate fields in the field template can be set according to the drive test data. Self-driving cars carry various sensors, such as lidar, millimeter-wave radar, and cameras. Therefore, during the road test of an autonomous vehicle, various sensor data can be obtained through sensors to form the road test data. According to the drive test data, multiple fields can be determined, and the multiple fields are used as candidate fields to form a field template.

In one embodiment, the candidate fields in the field template can be set according to historical simulation data. The historical simulation data includes multiple historical simulation models. According to each historical simulation model, the field set corresponding to each historical simulation model can be determined. The fields included in each field set can be used as candidate fields to form a field template, which improves the applicability and comprehensiveness of the field template. It is also possible to count the repetition rate of the fields in all the field sets, and sort all the fields according to the repetition rate from large to small. Filter out fields with lower repetition rate, that is, filter out fields with low usage rate, and use fields with higher repetition rate as candidate fields to form a field template, which can save server resources.

In an embodiment, the corresponding field template can be set according to the scene type, thereby improving the specificity of the field template. Since the feature attributes of objects in multiple scenes of the same type are more similar, and the feature attributes of objects in different types of scenes are less similar, the corresponding field template can be set according to the scene type. For example, a school-type scene corresponds to a field template, and the field template can have a school as a label. When the test scene is the school end scene of elementary school A, a field template labeled with the school can be obtained to simulate the test scene.

S206: Determine the target field of each object from the field template according to the feature information corresponding to each object, and assign a value to the target field of each object according to the feature information corresponding to each object.

Specifically, the server may determine the feature field and feature field value corresponding to each object according to the feature information corresponding to each object. Furthermore, the server can search for candidate fields that are the same as the feature field in the field template according to the feature field, use the found candidate field as the target field of the object corresponding to the feature field, and assign a value to the target field according to the feature field value corresponding to the feature field. . For example, when the characteristic information of the crosswalk signal light is the coordinates (x1, y1, z1), the red light flashes at 1HZ for 15s and then switches to green light, and the green light flashes at 1HZ for 15s and then switches to red. The characteristic fields are coordinates, color, and switching frequency. Coordinate, color, and switching frequency fields exist in the field template. Therefore, the target fields of the crosswalk signal light are: coordinates, color, and switching frequency, where the field values corresponding to the coordinates are (x1, y1, z1), the field values corresponding to the colors are red, green, and the field value corresponding to the switching frequency is 1. .

In an embodiment, when a candidate field matching the characteristic field cannot be found in the field template, the characteristic field may be added to the field template as a new candidate field.

In an embodiment, the feature field and feature field value corresponding to the feature information of the object can be determined according to a pre-trained deep learning neural network. The feature information of the object is input into the pre-trained deep learning neural network, and the pre-trained deep learning neural network outputs the feature field and feature field value corresponding to the feature information.

In one embodiment, when the test scene is a real scene, the characteristic field and characteristic field value corresponding to the object can be determined through the drive test data.

S208: Generate a target code corresponding to the target field of each object according to the assigned target field of each object, and generate a simulation model corresponding to each object according to the target code corresponding to each object.

Specifically, the target field follows the definition of the format specification. After assigning a value to the target field corresponding to the object according to the characteristic information of the object, the corresponding target code can be generated according to the format specification of the target field, and the simulation model of the object can be generated by running the target code.

In one embodiment, since the simulation models corresponding to each object are based on the same field template, the simulation models can be managed conveniently and efficiently. Because based on the same field template, the underlying implementation of each simulation model is exactly the same, so a completely unified data structure can be used for reading, calling, assigning and saving. The server can read each target field corresponding to the object in turn, eliminate all fields with blank default values, and automatically generate codes for the target fields that have been assigned according to the format specifications corresponding to the target fields, and then obtain different simulation models.

In one embodiment, the code generation tool can be used to automatically generate code according to the format specification, that is, the corresponding target code is automatically generated according to the target field. The code generation tool has the characteristics of cross-platform, cross-language, and portability, and does not require users to write, which effectively improves the efficiency of code generation.

In an embodiment, the target field corresponding to each object may include a name field. The name field can be used to identify the object corresponding to each simulation model, so as to distinguish the simulation models corresponding to different objects.

S210: Establish a simulation scene corresponding to the test scene according to the simulation model corresponding to each object.

Among them, the simulation scene is a real-time multi-dimensional virtual world that reflects the changes and interactions of objects created by simulation, which can be used to test and evaluate automatic driving algorithms. Testing autonomous vehicles through simulation scenarios can provide methods and theoretical guidance for on-site testing of autonomous vehicles, and can improve the safety of on-site testing.

Specifically, the server may establish a corresponding simulation scene according to the simulation model. For example, the simulation scene may include simulation models corresponding to static objects such as lane lines, traffic signs, and sidewalks, and simulation models corresponding to dynamic objects such as bicycles and pedestrians. Further, the autonomous driving algorithm can plan the driving route for the autonomous car according to the simulation scene, and control the autonomous driving car to drive in the simulation scene. According to the actual driving route of the autonomous vehicle, the reliability of the autonomous driving algorithm can be evaluated.

The above test scenario simulation method is to obtain the feature information corresponding to each object in the test scenario; obtain the field template, which includes multiple candidate fields; determine the target field of each object from the field template according to the feature information corresponding to each object, and determine the target field of each object from the field template according to the The feature information corresponding to the object is assigned to the target field of each object; the target code corresponding to the target field of each object is generated according to the target field of each object after the assignment, and the simulation model corresponding to each object is generated according to the target code corresponding to each object; The simulation model corresponding to each object establishes the simulation scene corresponding to the test scene. In this way, the field template includes multiple candidate fields, and each candidate field has a corresponding code. Since the feature information corresponding to the object can indicate the function corresponding to the object, the function corresponding to the object can be realized by assigning the value of the target field, and the target code is automatically generated according to the target field, and then the simulation model corresponding to the object is obtained without changing the code, and Redundant codes are reduced, code maintenance is reduced, and code generation efficiency is improved.

In one embodiment, before S202, the test scenario simulation method further includes: obtaining historical simulation data, extracting the fields of each historical simulation model from the historical simulation data, and obtaining the field set corresponding to each historical simulation model; obtaining each historical simulation model separately Corresponding fields in the field set; count the repetition rate of each field; compose a field template according to the fields whose repetition rate is greater than a preset threshold.

Among them, the repetition rate refers to the ratio of the number of repetitions of a field to the total number of field sets. The preset threshold is set in advance and can be set according to actual needs.

Specifically, the server may obtain historical simulation data from the database, or may obtain historical simulation data from other servers. The historical simulation data includes multiple historical simulation scenarios, and each historical simulation scenario includes multiple historical simulation models. Since the historical simulation model is generated based on at least one field, the server can obtain the field set corresponding to the historical simulation model. The field set corresponding to each historical simulation model includes at least one field. Each historical simulation model can include the same field or different fields. For example, the field set a corresponding to the historical simulation model A includes field 1, field 2, field 3, and field 4. The field set b corresponding to the historical simulation model B includes field 1, field 2, field 5, field 6, and field 7. The server can calculate the repetition rate of each field, and compare the relationship between the repetition rate of each field and the preset threshold. When the repetition rate of a field is greater than the preset threshold, the field is used as a candidate field of the field template. All fields with a repetition rate greater than a preset threshold are used as candidate fields to form a field template. ,

In the above embodiment, by obtaining historical simulation data, extracting the fields of each historical simulation model from the historical simulation data to obtain the field set corresponding to each historical simulation model; respectively obtaining the fields in the field set corresponding to each historical simulation model; The repetition rate of the field; the field template is composed according to the fields whose repetition rate is greater than the preset threshold. The repetition rate of a field reflects the importance of the field, and fields with higher importance are combined into a field template, which improves the concentration of the field template. In addition, filtering out less important fields can also save server storage resources.

As shown in Figure 3, in one embodiment, S206 includes:

S302: Determine the feature field and feature field value corresponding to each object according to the feature information corresponding to each object.

S304: Match the candidate field in the field template with the feature field corresponding to each object.

S306: When the matching is successful, the candidate field that successfully matches the feature field corresponding to each object is used as the target field of each object.

S308: Assign a value to the target field of each object according to the feature field value corresponding to each object.

Specifically, the server can determine the feature field and feature field value corresponding to each object according to the feature information of each object. When the characteristic field is obtained, the candidate field that is the same as the characteristic field can be searched in the field template. When the candidate field corresponding to the characteristic field is found, the characteristic field is matched successfully. The candidate field that is successfully matched is used as the target field corresponding to the feature field, that is, the candidate field is used as the target field corresponding to the object. One object corresponds to at least one feature field. When the feature fields corresponding to the object are matched successfully, the target field corresponding to the at least one feature field can be determined from the field template. The target field corresponding to the characteristic field can be assigned a value according to the characteristic field value corresponding to the characteristic field.

In an embodiment, S206 further includes: when the matching fails, the feature field that fails to match is used as the update candidate field; and the update candidate field is added to the field template.

Specifically, when a candidate field that is the same as the feature field cannot be found in the field template, the feature field fails to match. The feature field that fails to match is used as a new candidate field and added to the field template to obtain an updated field template. Furthermore, the object including the characteristic field can determine the corresponding target field from the updated field template, and generate the corresponding simulation model according to the target field. In addition, the new candidate field will not affect the candidate field in the field template before the update, and the code corresponding to the new candidate field will not affect the code corresponding to the candidate field in the field template before the update.

In one embodiment, S302 includes: inputting feature information corresponding to each object into a pre-trained deep learning neural network, and outputting feature fields and feature field values corresponding to each object.

Specifically, the feature field and feature field value corresponding to the feature information of the object can be determined according to a pre-trained deep learning neural network. For the pre-trained deep learning neural network, the feature information of the object has been set as the input variable of the deep learning neural network, and the feature field and feature field value have been set as the output variable of the deep learning neural network. Therefore, when the acquired feature information of the object is input into the pre-trained deep learning neural network through the input device, the pre-trained deep learning neural network will calculate and output the feature field and feature field value of the corresponding object.

In the above embodiment, through the application of the deep learning neural network, the feature field and the feature field value of the object can be predicted when the feature information of the object is obtained, thereby providing data support for establishing a simulation model.

In one embodiment, before inputting the feature information corresponding to each object into a pre-trained deep learning neural network, and outputting the feature field and feature field value corresponding to each object, it includes: obtaining object sample data; obtaining sample object according to the object sample data The feature information, fields and field values of the deep learning neural network are trained.

Specifically, in the database, multiple object sample data are stored, and the object sample data contains feature information, fields, and field values of multiple objects. In the deep learning neural network, the input variables and output variables of the deep learning neural network are set in advance, the feature information is set as the input variables of the deep learning neural network, and the fields and field values are set as the output variables of the deep learning neural network. . Therefore, the feature information, fields and field values of multiple objects are respectively used as the input and expected output of the deep learning neural network, and the deep learning neural network will be trained based on multiple sets of data. The trained deep learning neural network can predict the output data based on the input data when it is used. Training the deep learning neural network can make the obtained prediction results more accurate.

In one embodiment, acquiring the feature information, fields, and field values of the sample object according to the object sample data to train the deep learning neural network includes: acquiring the feature information of the object according to the object sample data, and inputting the feature information of the object into the deep learning Unsupervised training in the neural network; obtain the fields and field values corresponding to the feature information of the sample object from the object sample data, use the obtained feature information of the sample object as the input data of the deep learning neural network, and use the obtained fields and fields The value is used as the expected output of the deep learning neural network for supervised training of the deep learning neural network.

Specifically, the feature information value of the object is input into the input variable corresponding to the deep learning neural network for unsupervised training. After unsupervised training, the deep learning neural network is trained again. When conducting supervised training, the input variables and expected output of the deep learning neural network are provided intact. For example, the feature information of the object is used as the input variable of the deep learning neural network, the corresponding field and field value are used as the expected output of the deep learning neural network, and the deep learning neural network is supervised training.

The feature information of multiple objects and the corresponding fields and field values are stored in the object sample data, that is, multiple sets of object data are stored in the object sample data. The data type in each group of object data includes the feature information of each object, and the field And field value. However, not every set of object data is complete. There may be missing fields and field values in some object data. This part of the object data lacking fields and field values can be used for unsupervised training of deep learning neural networks to avoid The waste of data. Since unsupervised training first trains the feature extraction capabilities of deep learning neural networks, supervised training after unsupervised training can also improve the training effect of deep learning neural networks and improve the predictive output of deep learning neural networks after training. The accuracy of the data.

In one embodiment, the unsupervised training adopts a bottom-up training method. A single layer of neurons can be constructed layer by layer, including an input layer, an output layer, and multiple hidden layers. The input layer is at the bottom, and the input variable of the input layer is the feature information of the object. The output layer is at the top, and the output variables of the output layer are the feature field and feature field value of the object. The hidden layer is located between the input layer and the output layer, and the number of hidden layers can be set according to actual needs. Unsupervised training is to train the output layer layer by layer from the input layer. Each layer can use wake-sleep algorithm for parameter tuning, and only adjust one layer at a time, adjusting layer by layer. In unsupervised training, no expected output is needed. The purpose of unsupervised training is not to predict the output, but to perceive the input. After unsupervised training of deep learning neural network, supervised training is carried out. Supervised training uses a top-down training method. On the basis of obtaining the parameters corresponding to the neurons in each layer through unsupervised training, a classifier is added to the output layer, and then through the supervised learning of the complete object data, the gradient descent method is used to fine-tune the parameters corresponding to the neurons in each layer.

In a specific embodiment, as shown in FIG. 4, a test scenario simulation method is provided, which specifically includes the following steps:

S402: Obtain historical simulation data, extract fields of each historical simulation model from the historical simulation data, and obtain a field set corresponding to each historical simulation model.

S404: Obtain respectively the fields in the field set corresponding to each historical simulation model.

S406: Count the repetition rate of each field.

S408: Form a field template according to fields whose repetition rate is greater than a preset threshold.

S410: Obtain object sample data.

S412: Obtain feature information of the object according to the object sample data, and input the feature information of the object into the deep learning neural network for unsupervised training.

S416: Obtain the fields and field values corresponding to the feature information of the sample object from the object sample data, use the obtained feature information of the sample object as the input data of the deep learning neural network, and use the obtained fields and field values as the deep learning neural network The expected output of the deep learning neural network for supervised training.

S418: Acquire feature information corresponding to each object in the test scene.

S420: Input the feature information corresponding to each object into the pre-trained deep learning neural network, and output the feature field and feature field value corresponding to each object.

S422: Match the candidate field in the field template with the feature field corresponding to each object.

S424: When the matching is successful, the candidate field that successfully matches the feature field corresponding to each object is used as the target field of each object.

S426: Assign a value to the target field of each object according to the feature field value corresponding to each object.

S428: Generate a target code corresponding to the target field of each object according to the assigned target field of each object, and generate a simulation model corresponding to each object according to the target code corresponding to each object.

S430: Establish a simulation scene corresponding to the test scene according to the simulation model corresponding to each object.

At this stage, multiple objects in the simulation test scene are managed by multiple containers. Each container stores a certain type of specific object, which increases the pre-defined memory usage, increases the number of cycles during processing and the difficulty of maintenance. New objects need to be modified after loading the data storage part of the code. Or, use a container to store pointers to the base class to manage, each of which points to an inherited class. When simulating objects in the test scene, frequent base class-inherited class type conversions are required, which will face problems such as pointer out-of-bounds, memory leaks, and system crashes caused by incorrect type conversions at runtime.

However, the test scenario simulation method of this application is used to uniformly use a container for storing field templates for management. Each object to be simulated is an instance that conforms to the field template, and different functions are realized by assigning values to different fields. Unused fields will be left blank by default. In this way, even if a new field is added to the field template, there is no need to change the data storage code, and it will not affect the correctness of the previous code. At the same time, the fields are stored in the form of variables instead of pointers, which effectively reduces the chance of system errors.

It should be understood that although the various steps in the above flowchart are displayed in sequence as indicated by the arrows, these steps are not necessarily performed in sequence in the order indicated by the arrows. Unless specifically stated in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least part of the steps in the above flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The order of execution is not necessarily performed sequentially, but may be performed alternately or alternately with other steps or at least a part of sub-steps or stages of other steps.

Fig. 5 is a structural block diagram of a test scenario simulation device in an embodiment. As shown in FIG. 5, a test scenario simulation device includes an acquisition module 502, a target field determination module 504, a simulation model generation module 506, and a simulation scenario establishment module 508. in:

The obtaining module 502 is configured to obtain feature information corresponding to each object in the test scene; obtain a field template, which includes a plurality of candidate fields.

The target field determination module 504 is configured to determine the target field of each object from the field template according to the feature information corresponding to each object, and assign a value to the target field of each object according to the feature information corresponding to each object.

The simulation model generation module 506 is configured to generate the target code corresponding to the target field of each object according to the assigned target field of each object, and generate the simulation model corresponding to each object according to the target code corresponding to each object.

The simulation scene establishing module 508 is used to establish a simulation scene corresponding to the test scene according to the simulation model corresponding to each object.

In one embodiment, the obtaining module 502 is also used to obtain historical simulation data, extract the fields of each historical simulation model from the historical simulation data, and obtain the field set corresponding to each historical simulation model; respectively obtain the field set corresponding to each historical simulation model Count the repetition rate of each field; compose a field template according to the fields whose repetition rate is greater than a preset threshold.

In one embodiment, the target field determination module 504 is further configured to determine the feature field and feature field value corresponding to each object according to the feature information corresponding to each object; match the candidate field in the field template with the feature field corresponding to each object; When the matching is successful, the candidate field that successfully matches the feature field corresponding to each object is used as the target field of each object; the target field of each object is assigned according to the value of the feature field corresponding to each object.

In one embodiment, the target field determining module 504 is further configured to use the feature field that failed to match as an update candidate field when the matching fails; add the update candidate field to the field template.

In an embodiment, the target field determination module 504 is further configured to input the feature information corresponding to each object into a pre-trained deep learning neural network, and output the feature field and feature field value corresponding to each object.

In an embodiment, the target field determination module 504 is also used to obtain object sample data; obtain feature information, fields, and field values of the sample objects according to the object sample data to train the deep learning neural network.

In one embodiment, the target field determination module 504 is also used to obtain the feature information of the object according to the sample data of the object, and input the feature information of the object into the deep learning neural network for unsupervised training; The field and field value corresponding to the feature information of the sample object are used as the input data of the deep learning neural network, and the obtained field and field value are used as the expected output of the deep learning neural network. Supervise training.

The above-mentioned test scene simulation device obtains the characteristic information corresponding to each object in the test scene; obtains a field template, which includes multiple candidate fields; determines the target field of each object from the field template according to the characteristic information corresponding to each object, and determines the target field of each object from the field template according to the The feature information corresponding to the object is assigned to the target field of each object; the target code corresponding to the target field of each object is generated according to the target field of each object after the assignment, and the simulation model corresponding to each object is generated according to the target code corresponding to each object; The simulation model corresponding to each object establishes the simulation scene corresponding to the test scene. In this way, the field template includes multiple candidate fields, and each candidate field has a corresponding code. Since the feature information corresponding to the object can indicate the function corresponding to the object, the function corresponding to the object can be realized by assigning the value of the target field, and the target code is automatically generated according to the target field, and then the simulation model corresponding to the object is obtained without changing the code, and Redundant codes are reduced, code maintenance is reduced, and code generation efficiency is improved.

For the specific limitation of the test scenario simulation device, please refer to the above limitation of the test scenario simulation method, which will not be repeated here. Each module in the above test scenario simulation device can be implemented in whole or in part by software, hardware, and a combination thereof. The above-mentioned modules may be embedded in the form of hardware or independent of the processor in the computer equipment, or may be stored in the memory of the computer equipment in the form of software, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

In some embodiments, a computer device is provided. The computer device may be the server 104 in FIG. 1, and its internal structure diagram may be as shown in FIG. 6. The computer equipment includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program is executed by the processor to realize a test scenario simulation method.

Those skilled in the art can understand that the structure shown in FIG. 6 is only a block diagram of part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may Including more or fewer parts than shown in the figure, or combining some parts, or having a different arrangement of parts.

In an embodiment, the test scenario simulation device provided in this application may be implemented in the form of a computer program, and the computer program may run on the computer device as shown in FIG. 6. The memory of the computer device can store various program modules that make up the test scenario simulation device, such as the acquisition module, the target field determination module, the simulation model generation module, and the simulation scenario establishment module shown in FIG. 5. The computer program composed of each program module causes the processor to execute the steps in the test scenario simulation method of each embodiment of the present application described in this specification.

For example, the computer device shown in FIG. 6 may execute step S202 through the acquisition module in the test scenario simulation apparatus shown in FIG. 5. The computer device may execute step S206 through the target field determination module. The computer device may execute step S208 through the simulation model generation module. The computer device may establish and execute step S210 through a simulation scenario.

In one embodiment, a computer device is provided, which includes a memory and a processor, and the memory stores a computer program. When the computer program is executed by the processor, the processor causes the processor to execute the steps of the test scenario simulation method. Here, the steps of the test scenario simulation method may be the steps in the test scenario simulation method of each of the foregoing embodiments.

In one embodiment, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the above-mentioned test scenario simulation method. Here, the steps of the test scenario simulation method may be the steps in the test scenario simulation method of each of the foregoing embodiments.

A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage. In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, or optical storage. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not a limitation, RAM may be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc.

The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be It is considered as the range described in this specification.

The above-mentioned embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the scope of the invention patent. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can be made, and these all fall within the protection scope of this application. Therefore, the scope of protection of the patent of this application shall be subject to the appended claims.

Claims

A test scenario simulation method, which is characterized in that it includes:

Obtain feature information corresponding to each object in the test scene;

Obtaining a field template, where the field template includes a plurality of candidate fields;

Determine the target field of each object from the field template according to the feature information corresponding to each object, and assign a value to the target field of each object according to the feature information corresponding to each object;

Generating a target code corresponding to the target field of each object according to the assigned target field of each object, and generating a simulation model corresponding to each object according to the target code corresponding to each object;

The simulation scene corresponding to the test scene is established according to the simulation model corresponding to each object.
The method according to claim 1, wherein before said obtaining the field template, the method further comprises:

Acquiring historical simulation data, extracting the fields of each historical simulation model from the historical simulation data, and obtaining a field set corresponding to each historical simulation model;

Respectively obtain the fields in the field set corresponding to each historical simulation model;

Count the repetition rate of each field;

The field template is formed according to the fields whose repetition rate is greater than the preset threshold.
The method according to claim 1, wherein the target field of each object is determined from the field template according to the characteristic information corresponding to each object, and the target field of each object is determined according to the characteristic information corresponding to each object. The target field is assigned, including:

Determine the feature field and feature field value corresponding to each object according to the feature information corresponding to each object;

Matching the candidate field in the field template with the feature field corresponding to each object;

When the matching is successful, use the candidate field that successfully matches the feature field corresponding to each object as the target field of each object;

Assign a value to the target field of each object according to the characteristic field value corresponding to each object.
The method according to claim 3, wherein the method further comprises:

When the matching fails, the feature field that fails to match is used as the update candidate field;

The update candidate field is added to the field template.
The method according to claim 3, characterized in that, before determining the characteristic field and the characteristic field value corresponding to each object according to the characteristic information corresponding to each object, the method further comprises:

The feature information corresponding to each object is input into a pre-trained deep learning neural network, and the feature field and feature field value corresponding to each object are output.
The method according to claim 5, characterized in that, before inputting the characteristic information corresponding to each object into a pre-trained deep learning neural network, the method further comprises:

Obtain object sample data;

The feature information, fields, and field values of the sample objects are acquired according to the object sample data to train the deep learning neural network.
The method according to claim 6, wherein the acquiring feature information, fields, and field values of the sample objects according to the object sample data to train the deep learning neural network comprises:

Acquiring feature information of the object according to the object sample data, and inputting the feature information of the object into the deep learning neural network for unsupervised training;

Obtain the fields and field values corresponding to the feature information of the sample object from the object sample data, use the obtained feature information of the sample object as the input data of the deep learning neural network, and use the obtained fields and field values as the input data of the deep learning neural network. The expected output of the deep learning neural network performs supervised training on the deep learning neural network.
A test scenario simulation device, which is characterized in that it comprises:

An obtaining module, used to obtain feature information corresponding to each object in the test scene; obtain a field template, the field template including a plurality of candidate fields;

The target field determination module is configured to determine the target field of each object from the field template according to the characteristic information corresponding to each object, and assign a value to the target field of each object according to the characteristic information corresponding to each object;

A simulation model generation module, configured to generate a target code corresponding to the target field of each object according to the assigned target field of each object, and generate a simulation model corresponding to each object according to the target code corresponding to each object;

The simulation scene establishment module is used to establish a simulation scene corresponding to the test scene according to the simulation model corresponding to each object.
8. The device according to claim 8, wherein the acquisition module is further configured to acquire historical simulation data, extract fields of each historical simulation model from the historical simulation data, and obtain a set of fields corresponding to each historical simulation model; The fields in the field set corresponding to each historical simulation model are respectively obtained; the repetition rate of each field is counted; the field template is formed according to the fields with the repetition rate greater than a preset threshold.
The device according to claim 8, wherein the target field determination module is further configured to determine the feature field and feature field value corresponding to each object according to the feature information corresponding to each object; Match the feature field corresponding to each object; when the matching is successful, use the candidate field that successfully matches the feature field corresponding to each object as the target field of each object; according to the feature field corresponding to each object The value is assigned to the target field of each object.
The device according to claim 10, wherein the target field determining module is further configured to, when the matching fails, use the feature field of the matching failure as an update candidate field; and add the update candidate field to the field template .
The device according to claim 10, wherein the target field determination module is further configured to input feature information corresponding to each object into a pre-trained deep learning neural network, and output feature fields and features corresponding to each object Field value.
The device according to claim 12, wherein the target field determination module is further used to obtain object sample data; according to the object sample data to obtain feature information, fields, and field values of the sample object, perform a deep learning neural network Training.
The device according to claim 12, wherein the target field determination module is further configured to obtain feature information of the object according to the object sample data, and input the feature information of the object into the deep learning neural network Perform unsupervised training; obtain fields and field values corresponding to the feature information of the sample object from the object sample data, and use the obtained feature information of the sample object as the input data of the deep learning neural network, and the obtained Fields and field values are used as expected outputs of the deep learning neural network, and supervised training is performed on the deep learning neural network.
A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when the computer program is executed by the processor.
A computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the method according to any one of claims 1 to 7 when the computer program is executed by a processor.