阅读说明：本技术 骑行路线生成方法、装置、设备及存储介质 (Riding route generation method, device, equipment and storage medium ) 是由 余玉霞 于 2021-08-26 设计创作，主要内容包括：本发明涉及人工智能领域,公开了一种骑行路线生成方法、装置、设备及存储介质,用于提高骑行路线生成的准确率。所述骑行路线生成方法包括：调用卷积神经网络模型分别对多个虚拟骑行路线进行道路环境识别,得到每一虚拟骑行路线对应的目标环境要素；对每一虚拟骑行路线对应的目标环境要素进行数据处理,得到离散化道路数据；调用离散选择模型对离散化道路数据进行偏好分析,得到每一虚拟骑行路线对应的候选概率值；对每一虚拟骑行路线对应的候选概率值进行比较,并将目标概率值对应的虚拟骑行路线作为目标骑行路线,目标概率值为多个候选概率值中的最大值。此外,本发明还涉及区块链技术,目标骑行路线可存储于区块链节点中。(The invention relates to the field of artificial intelligence, and discloses a riding route generation method, a riding route generation device, riding route generation equipment and a riding route generation storage medium, which are used for improving the accuracy of riding route generation. The riding route generation method comprises the following steps: calling a convolutional neural network model to respectively identify the road environment of the virtual riding routes to obtain a target environment element corresponding to each virtual riding route; carrying out data processing on the target environment elements corresponding to each virtual riding route to obtain discretized road data; calling a discrete selection model to perform preference analysis on the discretization road data to obtain a candidate probability value corresponding to each virtual riding route; and comparing the candidate probability values corresponding to each virtual riding route, and taking the virtual riding route corresponding to the target probability value as the target riding route, wherein the target probability value is the maximum value in the candidate probability values. In addition, the invention also relates to a block chain technology, and the target riding route can be stored in the block chain node.)

1. A riding route generation method is characterized by comprising the following steps:

the method comprises the steps of obtaining initial data of a target user, preprocessing the initial data to obtain target data, wherein the target data are used for indicating a plurality of virtual riding routes;

calling a preset convolutional neural network model to respectively identify the road environment of the virtual riding routes to obtain a target environment element corresponding to each virtual riding route;

performing data processing on the target environment elements corresponding to each virtual riding route to obtain discretization road data corresponding to each virtual riding route;

calling a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route;

and comparing the candidate probability values corresponding to each virtual riding route, and taking the virtual riding route corresponding to the target probability value as the target riding route, wherein the target probability value is the maximum value in the candidate probability values.

2. The riding route generation method according to claim 1, wherein the obtaining initial data of a target user and preprocessing the initial data to obtain target data, the target data being used for indicating a plurality of virtual riding routes, comprises:

inquiring initial data of a target user from a preset database;

and carrying out data cleaning on the initial data to obtain target data, wherein the target data is used for indicating a plurality of virtual riding routes.

3. The riding route generation method according to claim 1, wherein the step of calling a preset convolutional neural network model to perform road environment recognition on the plurality of virtual riding routes respectively to obtain a target environment element corresponding to each virtual riding route comprises:

calling a preset street view map to match with the street view images of the virtual riding routes to obtain the street view image corresponding to each virtual riding route;

carrying out convolution operation on the street view image corresponding to each virtual riding route through a convolution layer in a preset convolution neural network model to obtain a target road environment corresponding to each virtual riding route;

and performing feature extraction on the target road environment corresponding to each virtual riding route through a full connection layer in a preset convolutional neural network model to obtain a target environment element corresponding to each virtual riding route.

4. The riding route generation method according to claim 1, wherein the step of performing data processing on the target environment elements corresponding to each virtual riding route to obtain discretized road data corresponding to each virtual riding route comprises:

classifying the target environment elements corresponding to each virtual riding route to obtain a plurality of target environment element classes;

acquiring the characteristic quantity of each target environment element class, and constructing a decision tree according to the characteristic quantity;

and carrying out data division on the multiple target environment element classes based on the decision tree, and carrying out data discretization on the multiple target environment element classes after data division to obtain discretization road data corresponding to each virtual riding route.

5. The riding route generation method according to claim 1, wherein the calling a preset discrete selection model to perform preference analysis on discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route includes:

obtaining the weight corresponding to the discretization road data corresponding to each virtual riding route, and obtaining the weight coefficient of the discretization road data corresponding to each virtual riding route;

carrying out utility calculation on the discretization road data corresponding to each virtual riding route through a utility function in a preset discrete selection model based on the weight coefficient to obtain a target utility value corresponding to each virtual riding route;

and calculating a route probability value of the target utility value corresponding to each virtual riding route through a probability function in a preset discrete selection model, and determining a candidate probability value corresponding to each virtual riding route.

6. The riding route generation method according to claim 1, wherein the comparing candidate probability values corresponding to each virtual riding route and taking the virtual riding route corresponding to a target probability value as the target riding route, wherein the target probability value is a maximum value of the candidate probability values, and comprises:

comparing candidate probability values corresponding to each virtual riding route to obtain target sequences corresponding to the candidate probability values;

and acquiring a maximum value in the target sequence based on the target sequence corresponding to the candidate probability values, and taking a virtual riding route corresponding to the maximum value in the candidate probability values as a target riding route.

7. The riding route generation method according to any one of claims 1 to 6, wherein after comparing the candidate probability values corresponding to each virtual riding route and taking the virtual riding route corresponding to a target probability value as a target riding route, the target probability value being a maximum value of the candidate probability values, the riding route generation method further comprises:

analyzing the route environment reasonability of the target riding route to obtain the target environment reasonability;

grading the target environment reasonableness to obtain a transformation priority;

and generating a road improvement evaluation index of the target riding route according to the transformation priority.

8. A riding route generating device, characterized by comprising:

the system comprises an acquisition module, a pre-processing module and a display module, wherein the acquisition module is used for acquiring initial data of a target user and preprocessing the initial data to obtain target data, and the target data is used for indicating a plurality of virtual riding routes;

the identification module is used for calling a preset convolutional neural network model to respectively identify the road environment of the virtual riding routes to obtain a target environment element corresponding to each virtual riding route;

the processing module is used for carrying out data processing on the target environment elements corresponding to each virtual riding route to obtain discretization road data corresponding to each virtual riding route;

the analysis module is used for calling a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route;

and the generating module is used for comparing the candidate probability values corresponding to each virtual riding route and taking the virtual riding route corresponding to the target probability value as the target riding route, wherein the target probability value is the maximum value in the candidate probability values.

9. A riding route generating device, characterized in that the riding route generating device comprises: a memory and at least one processor, the memory having instructions stored therein;

the at least one processor invokes the instructions in the memory to cause the cycling route generating device to perform the cycling route generating method according to any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, wherein the instructions, when executed by a processor, implement the riding route generation method of any of claims 1-7.

Technical Field

The invention relates to the field of image processing, in particular to a riding route generating method, a riding route generating device, riding route generating equipment and a storage medium.

Background

Under the condition that low-carbon sustainable development is more and more deeply mastered, the severity that urban life depends on car traveling is gradually emphasized, and the concept of green traveling is more and more deeply mastered. Thus, the traveling mode represented by the bicycle traveling starts returning to the field of vision of people. In fact, in recent years, green systems and public bicycle systems are built in a plurality of cities, and the practices have good social responses and reflect the urgent need of people for green travel.

The existing scheme mainly comprises the steps of obtaining an actual path of a rider, analyzing the correlation between the actual path and road attributes, or selecting the path which is most willing to ride from virtual paths to determine the influence of environmental elements on the path preference, but the accuracy of the current riding path prediction is low.

Disclosure of Invention

The invention provides a riding route generation method, a riding route generation device, riding route generation equipment and a riding route generation storage medium, which are used for improving the accuracy of riding route generation.

The invention provides a riding route generation method in a first aspect, which comprises the following steps: the method comprises the steps of obtaining initial data of a target user, preprocessing the initial data to obtain target data, wherein the target data are used for indicating a plurality of virtual riding routes; calling a preset convolutional neural network model to respectively identify the road environment of the virtual riding routes to obtain a target environment element corresponding to each virtual riding route; performing data processing on the target environment elements corresponding to each virtual riding route to obtain discretization road data corresponding to each virtual riding route; calling a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route; and comparing the candidate probability values corresponding to each virtual riding route, and taking the virtual riding route corresponding to the target probability value as the target riding route, wherein the target probability value is the maximum value in the candidate probability values.

Optionally, in a first implementation manner of the first aspect of the present invention, the obtaining initial data of a target user and preprocessing the initial data to obtain target data, where the target data is used to indicate multiple virtual riding routes, includes: inquiring initial data of a target user from a preset database; and carrying out data cleaning on the initial data to obtain target data, wherein the target data is used for indicating a plurality of virtual riding routes.

Optionally, in a second implementation manner of the first aspect of the present invention, the invoking a preset convolutional neural network model to respectively perform road environment recognition on the multiple virtual riding routes to obtain a target environment element corresponding to each virtual riding route includes: calling a preset street view map to match with the street view images of the virtual riding routes to obtain the street view image corresponding to each virtual riding route; carrying out convolution operation on the street view image corresponding to each virtual riding route through a convolution layer in a preset convolution neural network model to obtain a target road environment corresponding to each virtual riding route; and performing feature extraction on the target road environment corresponding to each virtual riding route through a full connection layer in a preset convolutional neural network model to obtain a target environment element corresponding to each virtual riding route.

Optionally, in a third implementation manner of the first aspect of the present invention, the performing data processing on the target environment element corresponding to each virtual riding route to obtain discretization road data corresponding to each virtual riding route includes: classifying the target environment elements corresponding to each virtual riding route to obtain a plurality of target environment element classes; acquiring the characteristic quantity of each target environment element class, and constructing a decision tree according to the characteristic quantity; and carrying out data division on the multiple target environment element classes based on the decision tree, and carrying out data discretization on the multiple target environment element classes after data division to obtain discretization road data corresponding to each virtual riding route.

Optionally, in a fourth implementation manner of the first aspect of the present invention, the invoking a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route includes: obtaining the weight corresponding to the discretization road data corresponding to each virtual riding route, and obtaining the weight coefficient of the discretization road data corresponding to each virtual riding route; carrying out utility calculation on the discretization road data corresponding to each virtual riding route through a utility function in a preset discrete selection model based on the weight coefficient to obtain a target utility value corresponding to each virtual riding route; and calculating a route probability value of the target utility value corresponding to each virtual riding route through a probability function in a preset discrete selection model, and determining a candidate probability value corresponding to each virtual riding route.

Optionally, in a fifth implementation manner of the first aspect of the present invention, the comparing the candidate probability values corresponding to each virtual riding route, and taking the virtual riding route corresponding to a target probability value as the target riding route, where the target probability value is a maximum value of the candidate probability values, includes: comparing candidate probability values corresponding to each virtual riding route to obtain target sequences corresponding to the candidate probability values; and acquiring a maximum value in the target sequence based on the target sequence corresponding to the candidate probability values, and taking a virtual riding route corresponding to the maximum value in the candidate probability values as a target riding route.

Optionally, in a sixth implementation manner of the first aspect of the present invention, after the comparing the candidate probability values corresponding to each virtual riding route and taking the virtual riding route corresponding to a target probability value as a target riding route, where the target probability value is a maximum value among the candidate probability values, the riding route generating method further includes: analyzing the route environment reasonability of the target riding route to obtain the target environment reasonability; grading the target environment reasonableness to obtain a transformation priority; and generating a road improvement evaluation index of the target riding route according to the transformation priority.

A second aspect of the present invention provides a riding route generating device, including: the system comprises an acquisition module, a pre-processing module and a display module, wherein the acquisition module is used for acquiring initial data of a target user and preprocessing the initial data to obtain target data, and the target data is used for indicating a plurality of virtual riding routes; the identification module is used for calling a preset convolutional neural network model to respectively identify the road environment of the virtual riding routes to obtain a target environment element corresponding to each virtual riding route; the processing module is used for carrying out data processing on the target environment elements corresponding to each virtual riding route to obtain discretization road data corresponding to each virtual riding route; the analysis module is used for calling a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route; and the generating module is used for comparing the candidate probability values corresponding to each virtual riding route and taking the virtual riding route corresponding to the target probability value as the target riding route, wherein the target probability value is the maximum value in the candidate probability values.

Optionally, in a first implementation manner of the second aspect of the present invention, the obtaining module is specifically configured to: inquiring initial data of a target user from a preset database; and carrying out data cleaning on the initial data to obtain target data, wherein the target data is used for indicating a plurality of virtual riding routes.

Optionally, in a second implementation manner of the second aspect of the present invention, the identification module is specifically configured to: calling a preset street view map to match with the street view images of the virtual riding routes to obtain the street view image corresponding to each virtual riding route; carrying out convolution operation on the street view image corresponding to each virtual riding route through a convolution layer in a preset convolution neural network model to obtain a target road environment corresponding to each virtual riding route; and performing feature extraction on the target road environment corresponding to each virtual riding route through a full connection layer in a preset convolutional neural network model to obtain a target environment element corresponding to each virtual riding route.

Optionally, in a third implementation manner of the second aspect of the present invention, the processing module is specifically configured to: classifying the target environment elements corresponding to each virtual riding route to obtain a plurality of target environment element classes; acquiring the characteristic quantity of each target environment element class, and constructing a decision tree according to the characteristic quantity; and carrying out data division on the multiple target environment element classes based on the decision tree, and carrying out data discretization on the multiple target environment element classes after data division to obtain discretization road data corresponding to each virtual riding route.

Optionally, in a fourth implementation manner of the second aspect of the present invention, the analysis module is specifically configured to: obtaining the weight corresponding to the discretization road data corresponding to each virtual riding route, and obtaining the weight coefficient of the discretization road data corresponding to each virtual riding route; carrying out utility calculation on the discretization road data corresponding to each virtual riding route through a utility function in a preset discrete selection model based on the weight coefficient to obtain a target utility value corresponding to each virtual riding route; and calculating a route probability value of the target utility value corresponding to each virtual riding route through a probability function in a preset discrete selection model, and determining a candidate probability value corresponding to each virtual riding route.

Optionally, in a fifth implementation manner of the second aspect of the present invention, the generating module is specifically configured to: comparing candidate probability values corresponding to each virtual riding route to obtain target sequences corresponding to the candidate probability values; and acquiring a maximum value in the target sequence based on the target sequence corresponding to the candidate probability values, and taking a virtual riding route corresponding to the maximum value in the candidate probability values as a target riding route.

Optionally, in a sixth implementation manner of the second aspect of the present invention, the riding route generating device further includes: the reasonability analysis module is used for carrying out route environment reasonability analysis on the target riding route to obtain target environment reasonability; grading the target environment reasonableness to obtain a transformation priority; and generating a road improvement evaluation index of the target riding route according to the transformation priority.

A third aspect of the present invention provides a riding route generating apparatus, including: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the cycling route generating device to perform the cycling route generating method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which when run on a computer, cause the computer to execute the above-described riding route generation method.

According to the technical scheme, target data are obtained by preprocessing initial data, and the target data comprise a plurality of virtual riding routes; road environment recognition is carried out on the virtual riding routes through a preset convolutional neural network model respectively to obtain discretization road data corresponding to each virtual riding route, the convolutional neural network model can effectively improve the accuracy of the road environment recognition, preference analysis is carried out on the discretization road data corresponding to each virtual riding route through a preset discrete selection model to obtain a target probability value corresponding to each virtual riding route, finally, the virtual riding route with the maximum target probability value is used as the target riding route, and the accuracy of the generation of the riding route is improved.

Drawings

FIG. 1 is a schematic diagram of an embodiment of a riding route generation method in an embodiment of the invention;

FIG. 2 is a schematic diagram of another embodiment of a riding route generation method in the embodiment of the invention;

FIG. 3 is a schematic diagram of one embodiment of a riding route generating device in the embodiment of the invention;

fig. 4 is a schematic view of another embodiment of the riding route generating device in the embodiment of the invention;

fig. 5 is a schematic diagram of an embodiment of the riding route generating device in the embodiment of the invention.

Detailed Description

The embodiment of the invention provides a riding route generation method, a riding route generation device, equipment and a storage medium, which are used for improving the accuracy of riding route generation. The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For convenience of understanding, a specific flow of the embodiment of the present invention is described below, and referring to fig. 1, a first embodiment of a riding route generating method in the embodiment of the present invention includes:

101. the method comprises the steps of obtaining initial data of a target user, preprocessing the initial data to obtain target data, wherein the target data are used for indicating a plurality of virtual riding routes;

it is to be understood that the executing subject of the present invention may be a riding route generating device, and may also be a terminal or a server, which is not limited herein. The embodiment of the present invention is described by taking a server as an execution subject. The server may be an independent server, or may be a cloud server that provides basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), and a big data and artificial intelligence platform.

The embodiment of the invention can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result. The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.

Specifically, the server inquires initial data of a target user from a preset database, wherein the initial data are main influence elements of a riding environment determined by combining questionnaire survey and existing research, the server realizes design and generates virtual route selection scenes through visualization operation of the main influence elements, and the server preprocesses the initial data to obtain target data, wherein the target data comprise a plurality of virtual riding routes.

102. Calling a preset convolutional neural network model to respectively identify the road environment of the virtual riding routes to obtain a target environment element corresponding to each virtual riding route;

specifically, the server identifies the bicycle lanes and the riding environments in the virtual riding routes through a preset Convolutional Neural Network (CNN), and it should be noted that before identifying the road environments of the virtual riding routes, a preset street view map needs to be input into each virtual riding route, a street view image of each virtual riding route is obtained through the preset street view map, and the street view image is used for identifying the road environments of the virtual riding routes, so that the friendliness and the existing problems of the bicycle riding environments in the urban area are quantified, and thus the target environment elements corresponding to each virtual riding route are identified. The preset convolutional neural network model is processed to obtain the spatial distribution of a plurality of virtual riding routes, and some characteristics set manually are added, such as: and the weights are shared, so that the convolutional neural network model is easy to train, a deeper network structure can be made, and a better recognition effect is achieved.

103. Performing data processing on the target environment elements corresponding to each virtual riding route to obtain discretization road data corresponding to each virtual riding route;

specifically, the server carries out classification processing, carries out data discretization on a plurality of target environment element types respectively, collects data such as vehicle positions, vehicle postures and road conditions through a plurality of data acquisition devices, plans an optimal riding route for a target, recommends an optimal riding gear, generates a riding track, finds riding partners and the like according to the vehicle positions, the vehicle postures and the road conditions.

104. Calling a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route;

specifically, the specific process of the server performing the preference analysis includes: the server sets a plurality of alternative items of a target user, the alternative items correspond to certain utilities respectively, the utilities are formed by adding a fixed part and a random part, the fixed utilities can be explained by certain observable elements, the random part represents the unobserved utilities and the influence of errors, the server carries out preference calculation on the target user based on the utilities to obtain probability values corresponding to the selection of the target user on a plurality of virtual riding routes.

105. And comparing the candidate probability values corresponding to each virtual riding route, and taking the virtual riding route corresponding to the target probability value as the target riding route, wherein the target probability value is the maximum value in the candidate probability values.

Specifically, when the server compares, the candidate probability values are ranked according to the candidate probability value of each virtual riding route, and the server takes the virtual riding route corresponding to the maximum value in the candidate probability values as the target riding route.

Further, the server stores the target riding route in a blockchain database, which is not limited herein.

In the embodiment of the invention, the preset convolutional neural network model is used for respectively identifying the road environment of the virtual riding routes to obtain the discretization road data corresponding to each virtual riding route, and the convolutional neural network model is used for identifying the road environment of the virtual riding routes, so that the accuracy of identifying the road environment can be effectively improved.

Referring to fig. 2, a second embodiment of the riding route generating method according to the embodiment of the present invention includes:

201. the method comprises the steps of obtaining initial data of a target user, preprocessing the initial data to obtain target data, wherein the target data are used for indicating a plurality of virtual riding routes;

specifically, the server queries initial data of a target user from a preset database; and the server performs data cleaning on the initial data to obtain target data, wherein the target data is used for indicating a plurality of virtual riding routes.

The server inquires initial data of a target user from a preset database, the preset database comprises the initial data of a plurality of users, the server combines data of field investigation, and the server collects narrative preference of the target user on a virtual riding route. And the server performs data cleaning on the initial data to obtain target data. Due to investigation, coding and logging errors in the initial data, some invalid and missing values in the initial data, appropriate processing needs to be given, the data processing including: estimating, deleting the whole instance, deleting the variables and deleting the variables in pairs, wherein the server checks whether the initial data meets the requirements according to the reasonable value range and the mutual relation of the preset variables, and if the initial data exceeds the normal range, is logically unreasonable or is inconsistent with the error data, the server deletes the error data in the initial data to obtain target data, wherein the target data is used for indicating a plurality of virtual riding routes. The server inputs the target data into a preset street view map, and generates a plurality of virtual riding routes corresponding to the target user through the preset street view map, namely the preset street view map generates a plurality of virtual riding routes by analyzing the target data of the virtual routes of the target user, wherein the street view map is a real-view map service and provides 360-degree panoramic images (namely street view images) of cities, streets or other environments for the user.

202. Calling a preset street view map to match with street view images of a plurality of virtual riding routes to obtain the street view image corresponding to each virtual riding route;

specifically, before the server identifies the road environment of the virtual riding routes, a preset street view map needs to be input into each virtual riding route, and a street view image of each virtual riding route is acquired through the preset street view map and is used for identifying the road environment of the virtual riding routes.

203. Carrying out convolution operation on the street view image corresponding to each virtual riding route through a convolution layer in a preset convolution neural network model to obtain a target road environment corresponding to each virtual riding route;

specifically, the server respectively carries out image recognition on each virtual riding route through a preset convolutional neural network model to obtain a target road environment corresponding to each virtual riding route; and the server performs environment analysis on the target road environment corresponding to each virtual riding route through a preset convolutional neural network model to obtain target environment elements corresponding to each virtual riding route.

The server inquires the position information of each virtual riding route from a preset street view map, and the server acquires real-time images of each virtual route according to the position information, such as: clicking a certain position in a street view map to trigger and input street view inquiry position information, in this embodiment, a street view image corresponding to preset street view inquiry position information may be pre-stored, and the server performs image recognition on each virtual riding route through a preset convolutional neural network model, that is, the server recognizes the street view image of each virtual riding route obtained from the preset street view map to obtain a target road environment corresponding to each virtual riding route, wherein the target road environment includes: motor vehicle traffic, bicycle lane type, isolation measures, motor vehicle roadside parking, bicycle lane width, road shading, road landscape, approach parks, approach rivers, the number of traffic lights, and the like.

204. Extracting the characteristics of the target road environment corresponding to each virtual riding route through a full connection layer in a preset convolutional neural network model to obtain the target environment elements corresponding to each virtual riding route;

the method comprises the steps that a server calculates the optimal threshold value of a target road environment, then the server divides pixels, then a preset environment element database is accessed for similarity matching, the highest similarity is obtained through matching and serves as a target environment element, the target environment element is specifically an environment element such as a motor vehicle flow value, a bicycle lane type, an approach river, the number of traffic lights and the like, and the server scores data of a plurality of target environment elements, so that data of each virtual riding route are discretized, and subsequent data prediction is facilitated.

205. Performing data processing on the target environment elements corresponding to each virtual riding route to obtain discretization road data corresponding to each virtual riding route;

optionally, the server classifies the target environment elements corresponding to each virtual riding route to obtain a plurality of target environment element classes; the server acquires the characteristic quantity of each target environment element class and constructs a decision tree according to the characteristic quantity; the server divides data of the multiple target environment element classes based on the decision tree, and performs data discretization on the multiple target environment element classes after data division to obtain discretization road data corresponding to each virtual riding route.

The server performs data discretization on the multiple target environment element types respectively to obtain discretization road data corresponding to each virtual riding route. The server collects the characteristic quantity of each virtual riding route corresponding to the target environment element, the server establishes an experiment sample according to the characteristic quantity, the server establishes a decision tree according to the experiment sample, and the decision tree learning is carried out, and the method specifically comprises the following steps: the server takes the whole experiment sample as a root node, analyzes the variation of the single characteristic quantity, queries the variation item with the maximum variation as a segmentation criterion, and then sequentially branches according to the maximum variation condition until a plurality of target environment element classes are obtained. The server instantiates parameters of a hidden Markov model according to the constructed decision tree, and performs data discretization on a plurality of target environment element classes according to the hidden Markov model, wherein the data discretization means that the server performs data partitioning on the plurality of target environment element classes according to each virtual riding route, and performs data scoring on the target environment element classes corresponding to each divided virtual riding route, so that the target environment element classes of each virtual riding route are discretized, and discretization road data corresponding to each virtual riding route is obtained.

206. Calling a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route;

specifically, the server obtains the weight corresponding to the discretization road data corresponding to each virtual riding route, and obtains the weight coefficient of the discretization road data corresponding to each virtual riding route; the server performs utility calculation on the discretization road data corresponding to each virtual riding route through a utility function in a preset discrete selection model based on the weight coefficient to obtain a target utility value corresponding to each virtual riding route; and the server calculates the route probability value of the target utility value corresponding to each virtual riding route through a probability function in a preset discrete selection model, and determines the candidate probability value corresponding to each virtual riding route.

The server obtains a preference Model of a target user for a bicycle travel environment through a discrete selection Model, and then obtains higher goodness-of-fit and parameter estimation conforming to reality, the server estimates the influence of riding environment elements on the selection of riding routes of people based on a general Mixed Logit Model, and a Model function is defined as follows:

V_ij＝a₁time_j+a₂vol_j+a₃vol2_j+a₄type1_j+a₅type2_j+a₆width_j+a₇green_j+...+a_nvar_j；

wherein, V_ijTarget utility value, V, available from Path j for target user i_cThreshold for route j, obtainable by training, P_ijA candidate probability value representing the rider selected route j, a variable a representing a weight coefficient of an environment element, time, vol1_j、width_jEtc. represent collected target environment elements.

207. And comparing the candidate probability values corresponding to each virtual riding route, and taking the virtual riding route corresponding to the target probability value as the target riding route, wherein the target probability value is the maximum value in the candidate probability values.

Specifically, the server compares candidate probability values corresponding to each virtual riding route to obtain target sequences corresponding to a plurality of candidate probability values; the server obtains the maximum value in the target sequence based on the target sequence corresponding to the candidate probability values, and takes the virtual riding route corresponding to the maximum value in the candidate probability values as the target riding route.

The server sets different weights for different attributes, a more accurate and efficient optimization algorithm is provided, reasonable evaluation of the riding environment of each virtual riding route and road element improvement suggestions can be finally obtained, further, the server identifies riding behaviors, bicycle lanes and the riding environment through a convolutional neural network model in deep learning, the friendliness degree and existing problems of the riding environment of the bicycles in urban areas are further quantized, the server discretizes the identified data, the server predicts and evaluates the data of the virtual routes through a discrete selection model, different weights are set for different elements for prediction, and the evaluation of the riding environment of the target riding route is more accurate.

Specifically, the route environment reasonability degree of the target riding route is analyzed to obtain the target environment reasonability degree; grading the target environment reasonableness to obtain a transformation priority; and generating a road improvement evaluation index of the target riding route according to the transformation priority.

The server sets different weights for different attributes, provides a more accurate and efficient optimization algorithm, and finally can obtain reasonable evaluation of the riding environment of each road and road element improvement suggestions. The server analyzes the reasonability degree of the route environment of the target riding route to obtain the reasonability degree of the target environment, and the method specifically comprises the following steps: the server equally divides the scoring result of the road into 5 levels according to the theoretical highest score and the lowest score, wherein the level 1 is the lowest score, the level 5 is the highest score, the server obtains a weight coefficient according to training, the larger the weight coefficient is, the larger the influence of the variable on the riding environment is, the more improvement and optimization are needed, so that road improvement suggestions are provided at the same time, namely the improvement levels of factors such as the flow of a motor vehicle, the type of a bicycle lane, the width of the bicycle lane, isolation measures and the like, the improvement priority can also be divided into 1-3 levels, the level 1 is the least preferential improvement, the level 3 is the most preferential improvement, the target environment reasonability is obtained, and the server generates the most reasonable road improvement evaluation index according to the target environment reasonability.

Further, the server stores the target riding route in a blockchain database, which is not limited herein.

In the embodiment of the invention, the identified data is discretized, the data of a plurality of virtual routes is predicted and evaluated through the discrete selection model, different weights are set for different elements to predict, and the evaluation of the riding environment of the target riding route is more accurate.

With reference to fig. 3, the riding route generating method in the embodiment of the present invention is described above, and a riding route generating device in the embodiment of the present invention is described below, where a first embodiment of the riding route generating device in the embodiment of the present invention includes:

an obtaining module 301, configured to obtain initial data of a target user, and pre-process the initial data to obtain target data, where the target data is used to indicate multiple virtual riding routes;

the identification module 302 is configured to call a preset convolutional neural network model to perform road environment identification on the multiple virtual riding routes respectively, so as to obtain a target environment element corresponding to each virtual riding route;

the processing module 303 is configured to perform data processing on the target environment element corresponding to each virtual riding route to obtain discretized road data corresponding to each virtual riding route;

the analysis module 304 is configured to invoke a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route;

the generating module 305 is configured to compare candidate probability values corresponding to each virtual riding route, and use the virtual riding route corresponding to a target probability value as a target riding route, where the target probability value is a maximum value among the multiple candidate probability values.

In the embodiment of the invention, target data are obtained by preprocessing initial data, wherein the target data comprise a plurality of virtual riding routes; road environment recognition is carried out on the virtual riding routes through a preset convolutional neural network model respectively to obtain discretization road data corresponding to each virtual riding route, the convolutional neural network model can effectively improve the accuracy of the road environment recognition, preference analysis is carried out on the discretization road data corresponding to each virtual riding route through a preset discrete selection model to obtain a target probability value corresponding to each virtual riding route, finally, the virtual riding route with the maximum target probability value is used as the target riding route, and the accuracy of the generation of the riding route is improved.

Referring to fig. 4, a second embodiment of the riding route generating device according to the embodiment of the present invention includes:

an obtaining module 301, configured to obtain initial data of a target user, and pre-process the initial data to obtain target data, where the target data is used to indicate multiple virtual riding routes;

the identification module 302 is configured to call a preset convolutional neural network model to perform road environment identification on the multiple virtual riding routes respectively, so as to obtain a target environment element corresponding to each virtual riding route;

the processing module 303 is configured to perform data processing on the target environment element corresponding to each virtual riding route to obtain discretized road data corresponding to each virtual riding route;

the analysis module 304 is configured to invoke a preset discrete selection model to perform preference analysis on the discretization road data corresponding to each virtual riding route to obtain a candidate probability value corresponding to each virtual riding route;

the generating module 305 is configured to compare candidate probability values corresponding to each virtual riding route, and use the virtual riding route corresponding to a target probability value as a target riding route, where the target probability value is a maximum value among the multiple candidate probability values.

Optionally, the obtaining module 301 is specifically configured to:

inquiring initial data of a target user from a preset database; and carrying out data cleaning on the initial data to obtain target data, wherein the target data is used for indicating a plurality of virtual riding routes.

Optionally, the identifying module 302 is specifically configured to:

calling a preset street view map to match with the street view images of the virtual riding routes to obtain the street view image corresponding to each virtual riding route; carrying out convolution operation on the street view image corresponding to each virtual riding route through a convolution layer in a preset convolution neural network model to obtain a target road environment corresponding to each virtual riding route; and performing feature extraction on the target road environment corresponding to each virtual riding route through a full connection layer in a preset convolutional neural network model to obtain a target environment element corresponding to each virtual riding route.

Optionally, the processing module 303 is specifically configured to:

classifying the target environment elements corresponding to each virtual riding route to obtain a plurality of target environment element classes; acquiring the characteristic quantity of each target environment element class, and constructing a decision tree according to the characteristic quantity; and carrying out data division on the multiple target environment element classes based on the decision tree, and carrying out data discretization on the multiple target environment element classes after data division to obtain discretization road data corresponding to each virtual riding route.

Optionally, the analysis module 304 is specifically configured to:

obtaining the weight corresponding to the discretization road data corresponding to each virtual riding route, and obtaining the weight coefficient of the discretization road data corresponding to each virtual riding route; carrying out utility calculation on the discretization road data corresponding to each virtual riding route through a utility function in a preset discrete selection model based on the weight coefficient to obtain a target utility value corresponding to each virtual riding route; and calculating a route probability value of the target utility value corresponding to each virtual riding route through a probability function in a preset discrete selection model, and determining a candidate probability value corresponding to each virtual riding route.

Optionally, the generating module 305 is specifically configured to:

comparing candidate probability values corresponding to each virtual riding route to obtain target sequences corresponding to the candidate probability values; and acquiring a maximum value in the target sequence based on the target sequence corresponding to the candidate probability values, and taking a virtual riding route corresponding to the maximum value in the candidate probability values as a target riding route.

Optionally, the riding route generating device further includes:

the reasonability analysis module 306 is used for carrying out route environment reasonability analysis on the target riding route to obtain target environment reasonability; grading the target environment reasonableness to obtain a transformation priority; and generating a road improvement evaluation index of the target riding route according to the transformation priority.

In the embodiment of the invention, target data are obtained by preprocessing initial data, wherein the target data comprise a plurality of virtual riding routes; road environment recognition is carried out on the virtual riding routes through a preset convolutional neural network model respectively to obtain discretization road data corresponding to each virtual riding route, the convolutional neural network model can effectively improve the accuracy of the road environment recognition, preference analysis is carried out on the discretization road data corresponding to each virtual riding route through a preset discrete selection model to obtain a target probability value corresponding to each virtual riding route, finally, the virtual riding route with the maximum target probability value is used as the target riding route, and the accuracy of the generation of the riding route is improved.

Fig. 3 and 4 describe the riding route generating device in the embodiment of the present invention in detail from the perspective of the modular functional entity, and the riding route generating device in the embodiment of the present invention is described in detail from the perspective of the hardware processing.

Fig. 5 is a schematic structural diagram of a riding route generating device provided by an embodiment of the present invention, where the riding route generating device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (CPUs) 510 (e.g., one or more processors) and a memory 520, one or more storage media 530 (e.g., one or more mass storage devices) storing applications 533 or data 532. Memory 520 and storage media 530 may be, among other things, transient or persistent storage. The program stored in the storage medium 530 may include one or more modules (not shown), each of which may include a series of instruction operations for the ride route generation apparatus 500. Still further, the processor 510 may be configured to communicate with the storage medium 530, and execute a series of instruction operations in the storage medium 530 on the riding route generation apparatus 500.

The ride route generation apparatus 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input-output interfaces 560, and/or one or more operating systems 531, such as Windows server, Mac OS X, Unix, Linux, FreeBSD, and so forth. Those skilled in the art will appreciate that the riding route generation device configuration shown in fig. 5 does not constitute a limitation of the riding route generation device, and may include more or less components than those shown, or combine certain components, or a different arrangement of components.

The invention further provides riding route generating equipment which comprises a memory and a processor, wherein computer readable instructions are stored in the memory, and when being executed by the processor, the computer readable instructions cause the processor to execute the steps of the riding route generating method in the embodiments.

The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, and which may also be a volatile computer-readable storage medium, having stored therein instructions, which, when run on a computer, cause the computer to perform the steps of the riding route generation method.

Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.

The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.