CN109606383A

CN109606383A - Method and apparatus for generating model

Info

Publication number: CN109606383A
Application number: CN201811636798.3A
Authority: CN
Inventors: 张连川
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Baidu Online Network Technology Beijing Co Ltd; Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2018-12-29
Filing date: 2018-12-29
Publication date: 2019-04-12
Anticipated expiration: 2038-12-29
Also published as: CN109606383B

Abstract

Embodiment of the disclosure discloses the method and apparatus for generating model.One specific embodiment of this method includes: to be based on initial model using nitrification enhancement, executed following training step, to learn the generation of acceleration: from target velocity set, choosing target velocity；From acceleration set, acceleration is chosen；Target vehicle is determined in the state of being travelled according to selected acceleration, whether target vehicle meets predetermined traveling smoothness condition；Meet traveling smoothness condition in response to determining, establishes the corresponding relationship between selected target velocity and selected acceleration；Determine whether to meet preset end training condition；Terminate training condition in response to determining to meet, generates the running model at least one corresponding relationship that characterization is established.The embodiment controls the traveling of vehicle using the model that nitrification enhancement obtains, to enrich the control mode of vehicle.

Description

Method and apparatus for generating model

Technical field

Embodiment of the disclosure is related to field of computer technology, and in particular to the method and apparatus for generating model.

Background technique

Current closed-loop automatic control technology is often based on feedback to reduce uncertainty.Engineering in practice, usually Using proportional plus integral plus derivative controller, to realize above-mentioned adjusting control.Using the feedforward control system of proportional plus integral plus derivative controller System compensates, to reduce the deviation of system.

For example, when controlling vehicle, the prior art generallys use proportional plus integral plus derivative controller, to realize to vehicle Control.

Summary of the invention

The present disclosure proposes the method and apparatus for generating model, and the method and apparatus for generating information.

In a first aspect, embodiment of the disclosure provides a kind of method for generating model, this method comprises: obtaining mesh Mark sets of speeds and acceleration set, wherein the acceleration in acceleration set is used to indicate target vehicle and reaches target velocity Acceleration to be had；Using nitrification enhancement, it is based on initial model, following training step is executed, to learn acceleration It generates: from target velocity set, choosing target velocity；From acceleration set, acceleration is chosen；Determine target vehicle by In the state of being travelled according to selected acceleration, whether target vehicle meets predetermined traveling smoothness condition；Response Meet traveling smoothness condition in determining, establishes the corresponding relationship between selected target velocity and selected acceleration；Really It is fixed whether to meet preset end training condition；Terminate training condition in response to determining to meet, generates characterization and established at least The running model of one corresponding relationship.

In some embodiments, this method further include: be unsatisfactory for terminating training condition in response to determining, adjust initial model Model parameter training step is continued to execute using model parameter initial model adjusted.

In some embodiments, determine target vehicle in the state of being travelled according to selected acceleration, target Whether vehicle meets predetermined traveling smoothness condition, comprising: determines that target vehicle is being emulated according to selected acceleration In the state of being travelled in environment, whether target vehicle meets predetermined traveling smoothness condition.

In some embodiments, determine target vehicle in the state of being travelled according to selected acceleration, target Whether vehicle meets predetermined traveling smoothness condition, comprising: target vehicle is determined, according to selected acceleration in reality In the state of being travelled in driving process, whether target vehicle meets predetermined traveling smoothness condition.

In some embodiments, the acceleration one in the target velocity in target velocity set and acceleration set is a pair of It answers；And for the target velocity in target velocity set, the corresponding acceleration of the target velocity is to obtain as follows : determine that target vehicle reaches the time of the target velocity from initial velocity；Determine the difference of the target velocity and initial velocity； By the ratio of identified difference and identified time, it is determined as the corresponding acceleration of the target velocity.

In some embodiments, it is pre- to include at least one of the following: that the average speed of target vehicle is less than for traveling smoothness condition If threshold speed；The rate of acceleration change of target vehicle is less than preset rate of acceleration change threshold value.

Second aspect, embodiment of the disclosure provide a kind of for generating the device of model, which includes: first to obtain Unit is taken, is configured to obtain target velocity set and acceleration set, wherein the acceleration in acceleration set is used to indicate Target vehicle reaches target velocity acceleration to be had；Training unit is configured to using nitrification enhancement, based on initial Model executes following training step, to learn the generation of acceleration: from target velocity set, choosing target velocity；From acceleration In degree set, acceleration is chosen；Determine target vehicle in the state of being travelled according to selected acceleration, target vehicle Whether predetermined traveling smoothness condition is met；Meet traveling smoothness condition in response to determining, establishes selected target speed Corresponding relationship between degree and selected acceleration；Determine whether to meet preset end training condition；It is full in response to determining Foot terminates training condition, generates the running model at least one corresponding relationship that characterization is established.

In some embodiments, device further include: adjustment unit is configured in response to determine that being unsatisfactory for end trains Condition adjusts the model parameter of initial model, using model parameter initial model adjusted, continues to execute training step.

In some embodiments, training unit includes: the first determining module, is configured to determine target vehicle according to selected In the state that the acceleration taken is travelled in simulated environment, whether target vehicle meets the predetermined smooth item of traveling Part.

In some embodiments, training unit includes: the second determining module, is configured to determine target vehicle, according to institute For the acceleration of selection in the state of being travelled during actual travel, it is flat whether target vehicle meets predetermined traveling Sliding condition.

The third aspect, embodiment of the disclosure provide a kind of method for generating information, this method comprises: obtaining mesh Mark the initial velocity and target velocity of vehicle；Initial velocity and target velocity are input to running model trained in advance, obtained Acceleration, wherein running model is obtained according to the method training of any embodiment in such as above-mentioned method for being used to generate model , acceleration is used to indicate target vehicle and reaches target velocity acceleration to be had；According to acceleration, generation is used to indicate mesh Mark the instruction of vehicle driving.

Fourth aspect, embodiment of the disclosure provide a kind of for generating the device of information, which includes: second to obtain Unit is taken, is configured to obtain the initial velocity and target velocity of target vehicle；Input unit, be configured to initial velocity and Target velocity is input to running model trained in advance, obtains acceleration, wherein running model is according to such as above-mentioned for generating The method training of any embodiment obtains in the method for model, and acceleration, which is used to indicate target vehicle and reaches target velocity, waits having Some acceleration；Generation unit is configured to generate the instruction for being used to indicate target vehicle traveling according to acceleration.

5th aspect, embodiment of the disclosure provide a kind of electronic equipment, comprising: one or more processors；Storage Device is stored thereon with one or more programs, when said one or multiple programs are executed by said one or multiple processors, So that the method that the one or more processors realize any embodiment in the method as above-mentioned for generating model.

6th aspect, embodiment of the disclosure provide a kind of computer-readable medium, are stored thereon with computer program, The method of any embodiment in the method as above-mentioned for generating model is realized when the program is executed by processor.

The method and apparatus for generating model that embodiment of the disclosure provides, by obtaining target velocity set and adding Sets of speeds, wherein the acceleration in acceleration set is used to indicate target vehicle and reaches target velocity acceleration to be had, Then, using nitrification enhancement, it is based on initial model, following training step is executed, to learn the generation of acceleration: from target In sets of speeds, target velocity is chosen；From acceleration set, acceleration is chosen；Determine target vehicle according to selected In the state that acceleration is travelled, whether target vehicle meets predetermined traveling smoothness condition；Meet in response to determining Smoothness condition is travelled, the corresponding relationship between selected target velocity and selected acceleration is established；Determine whether to meet Preset end training condition；In response to determining that satisfaction terminates training condition, generates at least one that characterization is established and correspond to pass The running model of system can be realized the control to vehicle, enrich as a result, without being based on proportional plus integral plus derivative controller The mode of vehicle is controlled, in addition, helping by the model obtained based on nitrification enhancement training to control the traveling of vehicle In the precision for improving vehicle control, the safety of vehicle driving is helped to improve.

Detailed description of the invention

By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the disclosure is other Feature, objects and advantages will become more apparent upon:

Fig. 1 is that one embodiment of the disclosure can be applied to exemplary system architecture figure therein；

Fig. 2 is the flow chart according to one embodiment of the method for generating model of the disclosure；

Fig. 3 is the schematic diagram according to an application scenarios of the method for generating model of the disclosure；

Fig. 4 is the flow chart according to another embodiment of the method for generating model of the disclosure；

Fig. 5 is the structural schematic diagram according to one embodiment of the device for generating model of the disclosure；

Fig. 6 is the flow chart according to one embodiment of the method for generating information of the disclosure；

Fig. 7 is the structural schematic diagram according to one embodiment of the device for generating information of the disclosure；

Fig. 8 is adapted for the structural schematic diagram for the computer system for realizing the electronic equipment of embodiment of the disclosure.

Specific embodiment

The disclosure is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to Convenient for description, part relevant to related invention is illustrated only in attached drawing.

It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the disclosure can phase Mutually combination.The disclosure is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

Fig. 1 is shown can the method for generating model using embodiment of the disclosure or the dress for generating model It sets, alternatively, the exemplary system architecture 100 for generating the method for information or the embodiment of the device for generating information.

As shown in Figure 1, system architecture 100 may include terminal device 101,102, server 103, network 104 and vehicle 105.Network 104 between terminal device 101,102, server 103 and vehicle 105 to provide the medium of communication link.Net Network 104 may include various connection types, such as wired, wireless communication link or fiber optic cables etc..

Terminal device 101,102, server 103, vehicle 105 can be by the interactions of network 104, to receive or send data (such as the signal for being used to indicate the movement of vehicle) etc..Various telecommunication customer ends can be installed on terminal device 101,102 to answer With, such as equipment control application, image processing class application, web browser applications, the application of shopping class, searching class apply, be instant Means of communication, mailbox client, social platform software etc..

Terminal device 101,102 can be hardware, be also possible to software.It, can be with when terminal device 101,102 is hardware It is various electronic equipments, including but not limited to smart phone, tablet computer, pocket computer on knee and desktop computer etc. Deng.When terminal device 101,102 is software, may be mounted in above-mentioned cited electronic equipment.It may be implemented into more A software or software module (such as providing the software of Distributed Services or software module), also may be implemented into single software Or software module.It is not specifically limited herein.

Server 103 can be to provide the server of various services, such as to the backstage that the movement of vehicle 105 is controlled Server.Background server can carry out the data (such as target velocity set and acceleration set) received calculating etc. Reason, and processing result (such as the running model obtained based on target velocity set and the training of acceleration set) is fed back into vehicle 105。

It should be noted that server can be hardware, it is also possible to software.When server is hardware, may be implemented At the distributed server cluster that multiple servers form, individual server also may be implemented into.It, can when server is software To be implemented as multiple softwares or software module (such as providing the software of Distributed Services or software module), also may be implemented At single software or software module.It is not specifically limited herein.

Vehicle 105 can be the various vehicles that can be controlled.For example, vehicle 105 can by terminal device 101,102, Alternatively, the instruction that server 103 is sent is controlled；The controller or software that vehicle 105 itself can also be installed in are controlled System.As an example, vehicle 105 can include but is not limited to following any one: automobile, car, bus, automatic driving vehicle Etc..Vehicle 105 can be travelled after getting instruction according to the instruction of instruction.

It should be noted that can be by server 103 for generating the method for model provided by embodiment of the disclosure It executes, can also be executed by terminal device 101,102；Correspondingly, it can be set for generating the device of model in server 103 In, it also can be set in terminal device 101,102.In addition, for generating the side of information provided by embodiment of the disclosure Method can be executed by server 103, can also be executed, can also be executed by vehicle 105 by terminal device 101,102；Correspondingly, Device for generating information can be set in server 103, also can be set in terminal device 101,102, can be with It is set in vehicle 105.

It should be understood that the number of terminal device, network, server and vehicle in Fig. 1 is only schematical.According to reality It now needs, can have any number of terminal device, network, server and vehicle.For example, when the method for being used to generate information When the electronic equipment of operation thereon does not need to carry out data transmission with other electronic equipments, which can only include using Electronic equipment thereon is run in the method for generating model.

With continued reference to Fig. 2, the process of one embodiment of the method for generating model according to the disclosure is shown 200.The method for being used to generate model, comprising the following steps:

Step 201, target velocity set and acceleration set are obtained.

In the present embodiment, (such as server shown in FIG. 1, terminal are set the executing subject for generating the method for model Standby or vehicle) it can perhaps radio connection from other electronic equipments or local obtains mesh by wired connection mode Mark sets of speeds and acceleration set.

In the present embodiment, the acceleration in above-mentioned acceleration set, which is used to indicate target vehicle and reaches target velocity, waits having Some acceleration.Target velocity can be the speed that expectation reaches, alternatively, the speed that vehicle is to be achieved.

In practice, deviation usually there will be between the actual speed and target velocity of vehicle.For example, if for controlling vehicle Device or driver, it is expected that vehicle is travelled with 20 kilometer per hours of speed, then, in the case, due to The influence of the factors such as the resistance of vehicle, it will usually so that above-mentioned speed (the i.e. target speed that the actual speed of vehicle and expectation reach Degree) " 20 kilometer per hours " there are deviations.For example, the actual speed of vehicle may be less than 20 kilometer per hours.

Herein, above-mentioned target velocity set can be determined by technical staff.For example, a speed can be determined first It spends range (for example, 0 kilometer per hour to 10 kilometer per hours), then, above-mentioned velocity interval is divided into preset quantity (example Such as 300) set of the endpoint value in each section is determined as target velocity set by speed interval.Similar, above-mentioned acceleration Degree set can also be determined by technical staff.For example, can determine an acceleration range (for example, 0km/h first²It arrives 1km/h²), then, above-mentioned acceleration range is divided into preset quantity (such as 300) speed interval, by each section The set of endpoint value is determined as acceleration set.

In some optional implementations of the present embodiment, above-mentioned executing subject can in the following way, Lai Zhihang The step 201: the target velocity set and acceleration set in target vehicle driving process are obtained.

It is appreciated that when above-mentioned executing subject using this optional implementation to execute step 201 when, due to target speed Degree set and acceleration set are the real data in the driving process of target vehicle, thus, the target in target velocity set Between acceleration in velocity and acceleration set, it will usually which there are certain connections, it is possible thereby to reduce the instruction of running model Practice the time, improves the accuracy for the running model that training obtains.

Target velocity and acceleration set in some optional implementations of the present embodiment, in target velocity set In acceleration correspond.For the target velocity in target velocity set, the corresponding acceleration of the target velocity be can be Above-mentioned executing subject or the electronic equipment communicated to connect with above-mentioned executing subject, obtain as follows:

Firstly, determining that target vehicle reaches the time of the target velocity from initial velocity.Wherein, initial velocity can be mesh Vehicle is marked before receiving the instruction of acceleration or deceleration, actual speed possessed by target vehicle.It, can be with for each instruction It is corresponding with an initial velocity.

Then, it is determined that the difference of the target velocity and initial velocity.

Finally, the ratio of identified difference and identified time is determined as the corresponding acceleration of the target velocity.

Step 202, using nitrification enhancement, it is based on initial model, following training step is executed, to learn acceleration It generates.

In the present embodiment, above-mentioned executing subject can use nitrification enhancement, be based on initial model, execute following instruction White silk step (including step 2021- step 2026), to learn the generation of acceleration.

Herein, above-mentioned initial model can be indiscipline, or the model of preset condition is not up to after training.As Example, above-mentioned initial model can be Q table.Wherein, Q table can regard a bivariate table, the element value of Q table as It can be used for measuring under current state, take the superiority and inferiority degree of each behavior.Herein, above-mentioned initial model is also possible to have There is the model of deep neural network structure.

In the present embodiment, above-mentioned training step includes the following steps:

Step 2021, from target velocity set, target velocity is chosen.

In the present embodiment, above-mentioned executing subject can choose target from the target velocity set that step 201 is got Speed.

Herein, target velocity can be chosen in various manners, from acquired target velocity set.Such as with Machine is chosen, alternatively, choosing in particular order.

Step 2022, from acceleration set, acceleration is chosen.

In the present embodiment, above-mentioned executing subject can be chosen and accelerate from the acceleration set that step 201 is got Degree.

Step 2023, target vehicle is determined in the state of being travelled according to selected acceleration, and target vehicle is It is no to meet predetermined traveling smoothness condition.

In the present embodiment, above-mentioned executing subject can determine that target vehicle is travelled according to selected acceleration In the state of, whether target vehicle meets predetermined traveling smoothness condition.

Herein, above-mentioned target vehicle can be various vehicles, for example, automobile, car, bus, subway etc..It is above-mentioned Traveling smoothness condition can be predetermined for determine vehicle exercise whether smooth condition.For example, above-mentioned traveling is flat Sliding condition may include that the actual speed of target vehicle is less than pre-set velocity threshold value.

In some optional implementations of the present embodiment, traveling smoothness condition includes at least one of the following: target carriage Average speed be less than preset threshold speed；The rate of acceleration change of target vehicle is less than preset rate of acceleration change threshold Value.

Herein, the average speed of above-mentioned target vehicle can be target vehicle in preset duration (such as 1 minute) The average value of actual speed is also possible to before receiving acceleration or deceleration instruction, indicates to instructing according to acceleration or deceleration Acceleration traveling reach target velocity after, the average value of the actual speed in this period.

Above-mentioned rate of acceleration change can characterize the situation of change of acceleration of the vehicle within the unit time.Can by The difference of the acceleration at the endpoint moment of period, obtains divided by the duration of the period.

Step 2024, meet traveling smoothness condition in response to determining, establish selected target velocity and add with selected Corresponding relationship between speed.

In the present embodiment, in the case where determination meets above-mentioned traveling smoothness condition, above-mentioned executing subject be can establish Corresponding relationship between selected target velocity and selected acceleration.

Herein, the award of established corresponding relationship can be determined in the case where meeting above-mentioned traveling smoothness condition It is worth (reward), target is up to obtained total reward value, Lai Xunlian running model can pass through in the training process Each target velocity in above-mentioned target velocity set is determined, with the conversion between each acceleration in above-mentioned acceleration set Probability, the corresponding relationship between Lai Jianli target velocity and acceleration.

Step 2025, it is determined whether meet preset end training condition.

In the present embodiment, above-mentioned executing subject may determine whether to meet preset end training condition.Wherein, above-mentioned Terminating training condition can be the predetermined condition for being used to terminate above-mentioned training step of technical staff.For example, above-mentioned end Training condition can include but is not limited at least one of following: frequency of training meets or exceeds preset times；Training time reaches It or is more than preset duration.

Step 2026, terminate training condition in response to determining to meet, generate at least one corresponding relationship that characterization is established Running model.

In the present embodiment, in the case where determination meets above-mentioned end training condition, above-mentioned executing subject be can be generated Characterize the running model at least one corresponding relationship established.

In some optional implementations of the present embodiment, it is unsatisfactory for terminating training condition in response to determining, it is above-mentioned to hold Row main body can also adjust the model parameter of initial model, using model parameter initial model adjusted, continue to execute above-mentioned Training step.

It is appreciated that the process for executing training step is to adjust the process of each probability in Q table.For the first time or When executing above-mentioned steps 2022 before person several times, it can use greedy algorithm, from acceleration set, choose acceleration；And with Step 2022 be performed the increase of number, the selected corresponding maximum of target velocity can be chosen from acceleration set The acceleration of probability.

In some optional implementations of the present embodiment, for determining target vehicle according to selected acceleration In the state of being travelled, whether target vehicle meets predetermined traveling smoothness condition, the step for, above-mentioned executing subject It can execute in the following way: determine the shape that target vehicle is travelled in simulated environment according to selected acceleration Under state, whether target vehicle meets predetermined traveling smoothness condition.

Herein, above-mentioned simulated environment can be used for the running environment of simulating vehicle.For example, above-mentioned simulated environment can be with It is the running environment of the vehicle simulated by simulation software (such as CareMaker, carsim etc.).

It is appreciated that since in the training process, the traveling of vehicle usually there will be certain risk, therefore, it is possible to It was trained to primarily determine whether target vehicle meets predetermined traveling smoothness condition to reduce based on simulated environment Trial and error cost in journey reduces risk.

In some optional implementations of the present embodiment, for determining target vehicle according to selected acceleration In the state of being travelled, whether target vehicle meets predetermined traveling smoothness condition, the step for, above-mentioned executing subject It can also execute in the following way: determine target vehicle, be carried out during actual travel according to selected acceleration In the state of traveling, whether target vehicle meets predetermined traveling smoothness condition.

It is appreciated that determine whether target vehicle meets predetermined traveling smoothness condition during actual travel, It will obtain more accurately as a result, to improve the control precision of vehicle.

In addition, the driving model that training obtains by the way of the present embodiment does not learn the control of accelerator and brake directly, But learn the control of acceleration, training the control strategy come in this way has good migration characteristic, is not limited to a certain The vehicle of vehicle or a certain brand.

Furthermore the present embodiment can not depend on the traditional control algorithms such as proportional integral differential completely, to obtain acceleration, from And enrich the control mode of vehicle.The method of the present embodiment, be in and run at a low speed in vehicle (such as with less than 10 kms per small When speed traveling) state, alternatively, low speed climbing state when driving, compared with the prior art for, can more accurately control The traveling of vehicle processed.

Optionally, after generating above-mentioned running model, above-mentioned executing subject can also in cruise control on-line optimization The parameter of pi controller uses above-mentioned running model to go online optimization proportional integration control in a manner of data-driven The parameter of device processed makes cruise control reach desired performance.

With continued reference to the signal that Fig. 3, Fig. 3 are according to the application scenarios of the method for generating model of the present embodiment Figure.In the application scenarios of Fig. 3, server 301 obtains target velocity set 3001 and acceleration set 3002 first.Wherein, Acceleration in acceleration set 3001 is used to indicate target vehicle and reaches target velocity acceleration to be had.Then, it services Device 301 uses nitrification enhancement, is based on initial model, following training step is executed, to learn the generation of acceleration: from target In sets of speeds, target velocity is chosen；From acceleration set, acceleration is chosen；Determine target vehicle according to selected In the state that acceleration is travelled, whether target vehicle meets predetermined traveling smoothness condition；Meet in response to determining Smoothness condition is travelled, is established between initial velocity, selected target velocity and the selected acceleration three of target vehicle Corresponding relationship；Determine whether to meet preset end training condition；Terminate training condition in response to determining to meet, generates characterization The running model 3003 at least one corresponding relationship established.

The method provided by the above embodiment of the disclosure, by obtaining target velocity set and acceleration set, wherein add Acceleration in sets of speeds is used to indicate target vehicle and reaches target velocity acceleration to be had, then, using extensive chemical Algorithm is practised, initial model is based on, executes following training step, to learn the generation of acceleration: from target velocity set, choosing Target velocity；From acceleration set, acceleration is chosen；Determine that target vehicle is being travelled according to selected acceleration Under state, whether target vehicle meets predetermined traveling smoothness condition；Meet traveling smoothness condition in response to determining, establishes Corresponding relationship between selected target velocity and selected acceleration；Determine whether to meet preset end training item Part；Terminate training condition in response to determining to meet, generates the running model at least one corresponding relationship that characterization is established, thus It is not necessarily based on proportional plus integral plus derivative controller, the control to vehicle can be realized, enriches the mode of control vehicle as a result, this Outside, the essence of vehicle control is helped to improve to control the traveling of vehicle by the model obtained based on nitrification enhancement training Degree, helps to improve the safety of vehicle driving.

With further reference to Fig. 4, it illustrates the processes 400 of another embodiment of the method for generating model.The use In the process 400 for the method for generating model, comprising the following steps:

Step 401, using nitrification enhancement, it is based on initial model, following training step is executed, to learn acceleration It generates.

In the present embodiment, (such as server shown in FIG. 1, terminal are set the executing subject for generating the method for model Standby or vehicle) nitrification enhancement can be used, it is based on initial model, following training step is executed, to learn acceleration It generates.Wherein, above-mentioned training step includes the following steps, i.e. step 4001 to step 4013.

Step 4001, target velocity set and acceleration set are obtained.Later, step 4002 is executed.

In the present embodiment, above-mentioned executing subject can be electric from other by wired connection mode or radio connection Sub- equipment, or local acquisition target velocity set and acceleration set.

Herein, above-mentioned target velocity set can be determined by technical staff.Similar, above-mentioned acceleration set It can be determined by technical staff.

Step 4002, from target velocity set, target velocity is chosen, from acceleration set, chooses acceleration.It Afterwards, step 4003 is executed.

In the present embodiment, above-mentioned executing subject can choose mesh from the target velocity set that step 4001 is got Speed is marked, from the acceleration set that step 4001 is got, chooses acceleration.

Step 4003, in the state of determining that target vehicle is travelled in simulated environment according to selected acceleration, Whether target vehicle meets predetermined first traveling smoothness condition.Later, step 4004 is executed.

In the present embodiment, above-mentioned executing subject can determine target vehicle according to selected acceleration in simulated environment In travelled in the state of, target vehicle whether meet it is predetermined first traveling smoothness condition.

It is appreciated that simulated environment can be based on, to primarily determine it is flat whether target vehicle meets predetermined traveling Sliding condition reduces risk to reduce the trial and error cost in training process.

Step 4004, meet the first traveling smoothness condition in response to determining, establish selected target velocity with it is selected Acceleration between corresponding relationship.Later, step 4005 is executed.

In the present embodiment, in the case where determining that meeting first travels smoothness condition, above-mentioned executing subject can also be built Found the corresponding relationship between selected target velocity and selected acceleration.Wherein, the first traveling smoothness condition can be It is predetermined for determine vehicle in simulated environment exercise whether smooth condition.For example, the first traveling smoothness condition Include at least one of the following: that the average speed of target vehicle is less than preset threshold speed；The rate of acceleration change of target vehicle Less than preset rate of acceleration change threshold value.

Step 4005, it is determined whether meeting preset first terminates training condition.Later, if so, thening follow the steps 4006； If it is not, thening follow the steps 4002.

In the present embodiment, above-mentioned executing subject may also determine whether that meeting preset first terminates training condition.Its In, it is predetermined for terminating the condition of the training in simulated environment that the first end training condition can be technical staff. For example, above-mentioned first end training condition can include but is not limited to it is at least one of following: frequency of training meets or exceeds default Number；Training time meets or exceeds preset duration.

Step 4006, the target velocity set and acceleration set in target vehicle driving process are obtained.Later, step is executed Rapid 4007.

In the present embodiment, above-mentioned executing subject can also obtain target velocity set in target vehicle driving process and Acceleration set.

It is appreciated that the target velocity set and acceleration set in the step 4006 are in the driving process of target vehicle Real data, thus, between the acceleration in the target velocity in target velocity set and acceleration set, it will usually exist Certain connection improves the accuracy for the running model that training obtains it is possible thereby to reduce the training time of running model.

Herein, target velocity set and acceleration set acquired in the step 4006 can also be that driver is driving It is collected during above-mentioned target vehicle, thus, it is possible to learn the generation of acceleration in driving procedure, help to lead to It crosses the model that final training obtains and obtains more accurate acceleration.

Step 4007, from the target velocity set and acceleration set in target vehicle driving process, mesh is chosen respectively Velocity and acceleration is marked, in the state of determining target vehicle according to selected acceleration traveling, it is pre- whether target vehicle meets The the second traveling smoothness condition first determined.Later, step 4008 is executed.

In the present embodiment, above-mentioned executing subject can also from target vehicle driving process target velocity set and plus In sets of speeds, target velocity and acceleration are chosen respectively, determine the state that target vehicle is travelled according to selected acceleration Under, whether target vehicle meets predetermined second traveling smoothness condition.Wherein, the second traveling smoothness condition can be in advance Whether determining is used to determine enforcement of the vehicle in the generating process using the data study acceleration obtained in driving process Smooth condition.For example, the second traveling smoothness condition includes at least one of the following: the average speed of target vehicle less than preset Threshold speed；The rate of acceleration change of target vehicle is less than preset rate of acceleration change threshold value.

Step 4008, meet the second traveling smoothness condition in response to determining, establish selected target velocity with it is selected Acceleration between corresponding relationship.Later, step 4009 is executed.

In the present embodiment, above-mentioned executing subject, which may also respond to determine, meets the second traveling smoothness condition, establishes institute Corresponding relationship between the target velocity of selection and selected acceleration.

Step 4009, it is determined whether meeting preset second terminates training condition.Later, if so, thening follow the steps 4010； If it is not, thening follow the steps 4007.

In the present embodiment, above-mentioned executing subject may determine whether that meeting preset second terminates training condition.Wherein, It is predetermined for terminating to be trained based on the second traveling smoothness condition that second end training condition can be technical staff Condition.For example, above-mentioned second end training condition can include but is not limited to it is at least one of following: frequency of training reaches or surpasses Cross preset times；Training time meets or exceeds preset duration.

Step 4010, from the target velocity set and acceleration set in target vehicle driving process, mesh is chosen respectively Velocity and acceleration is marked, determines target vehicle, the state travelled during actual travel according to selected acceleration Under, whether target vehicle meets predetermined third traveling smoothness condition.Later, step 4011 is executed.

In the present embodiment, above-mentioned executing subject can be from the target velocity set and acceleration in target vehicle driving process In degree set, target velocity and acceleration are chosen respectively, target vehicle is determined, according to selected acceleration in actual travel mistake In the state of being travelled in journey, whether target vehicle meets predetermined third traveling smoothness condition.Wherein, third travels Smoothness condition can be predetermined for determining vehicle in the life using the data study acceleration obtained in driving process At in the process exercise whether smooth condition.For example, third traveling smoothness condition includes at least one of the following: target vehicle Average speed is less than preset threshold speed；The rate of acceleration change of target vehicle is less than preset rate of acceleration change threshold value.

Step 4011, meet third traveling smoothness condition in response to determining, establish selected target velocity with it is selected Acceleration between corresponding relationship.Later, step 4012 is executed.

In the present embodiment, above-mentioned executing subject, which may also respond to determine, meets third traveling smoothness condition, establishes institute Corresponding relationship between the target velocity of selection and selected acceleration.

Step 4012, it is determined whether meeting preset third terminates training condition.Later, if so, thening follow the steps 4013； If it is not, thening follow the steps 4010.

In the present embodiment, above-mentioned executing subject may also determine whether that meeting preset third terminates training condition.Its In, third terminate training condition can be technical staff it is predetermined for terminate based on third traveling smoothness condition instruct Experienced condition.For example, above-mentioned third terminate training condition can include but is not limited to it is at least one of following: frequency of training reach or More than preset times；Training time meets or exceeds preset duration etc..

It should be noted that above-mentioned first traveling smoothness condition, the second traveling smoothness condition and third travel smoothness condition In first, second, and third be used only as distinguish traveling smoothness condition, do not constitute to traveling smoothness condition particular determination, on Stating the first traveling smoothness condition, the second traveling smoothness condition and third traveling smoothness condition can be the identical smooth item of traveling Part is also possible to different traveling smoothness conditions.Above-mentioned first end training condition, the second end training condition and third terminate First, second, and third in training condition is used only as distinguishing end training condition, does not constitute to the spy for terminating training condition Different to limit, above-mentioned first end training condition, the second end training condition and third terminate training condition and can be identical knot Beam training condition is also possible to different end training conditions.It is not limited thereto.

Step 4013, the running model at least one corresponding relationship that characterization is established is generated.

In the present embodiment, the traveling mould at least one corresponding relationship that characterization is established can be generated in above-mentioned executing subject Type.

It is appreciated that the process for executing training step is to adjust the process of each probability in Q table.For the first time or When choosing target velocity and acceleration before person respectively from acceleration set several times, greedy algorithm can be used, from acceleration In set, acceleration is chosen；And as step 2022 is performed the increase of number, it can be from acceleration set, selected by selection The acceleration of the corresponding maximum probability of the target velocity taken.Furthermore, it is possible in various manners, from acquired target velocity collection In conjunction, target velocity is chosen.Such as randomly select, alternatively, choosing in particular order.

In some usage scenarios, the target velocity in the target vehicle driving process of above-mentioned steps 4007 and step 4010 Set and acceleration set, can be identical respectively, can also be different.For example, in target vehicle driving process in step 4007 Target velocity set and acceleration set can be the data generated during driver drives above-mentioned target vehicle, and walk Target velocity set in target vehicle driving process and acceleration set in rapid 4010 can be and given birth to automatically by target vehicle At data (i.e. the data that generate in actual measurement training process), thus, it is possible to first pass through simulation training, then existed using driver The training data training generated in driving procedure, finally gets on the bus and surveys training, can solve the demand that intensified learning needs trial and error, With the time and cost for reducing real vehicle training.

Herein, the specific embodiment of the step in the present embodiment can refer to the corresponding embodiment of Fig. 2 or optional Implementation, the description above with respect to the corresponding embodiment of Fig. 2 are readily applicable to the present embodiment, and details are not described herein.

Figure 4, it is seen that the method for generating model compared with the corresponding embodiment of Fig. 2, in the present embodiment Process 400 highlight in simulated environment the step of training obtains driving model, so as to primarily determine that target vehicle is It is no to meet predetermined traveling smoothness condition, to reduce the trial and error cost in training process, reduce risk.

With further reference to Fig. 5, as the realization to method shown in above-mentioned each figure, present disclose provides one kind for generating mould One embodiment of the device of type, the Installation practice is corresponding with embodiment of the method shown in Fig. 2, except following documented special Sign is outer, which can also include feature identical or corresponding with embodiment of the method shown in Fig. 2.The device specifically may be used To be applied in various electronic equipments.

As shown in figure 5, the device 500 for generating model of the present embodiment includes: that first acquisition unit 501 and training are single Member 502.Wherein, first acquisition unit 501 is configured to obtain target velocity set and acceleration set, wherein acceleration collection Acceleration in conjunction is used to indicate target vehicle and reaches target velocity acceleration to be had；Training unit 502 is configured to adopt With nitrification enhancement, it is based on initial model, following training step is executed, to learn the generation of acceleration: from target velocity collection In conjunction, target velocity is chosen；From acceleration set, acceleration is chosen；Determine target vehicle according to selected acceleration In the state of being travelled, whether target vehicle meets predetermined traveling smoothness condition；In response to determining that meeting traveling puts down Sliding condition is established corresponding between initial velocity, selected target velocity and the selected acceleration three of target vehicle Relationship；Determine whether to meet preset end training condition；Terminate training condition in response to determining to meet, generates characterization and established At least one corresponding relationship running model.

It in the present embodiment, can be by wired connection side for generating the first acquisition unit 501 of the device 500 of model Perhaps radio connection obtains target velocity set and acceleration set from other electronic equipments or locally to formula.

In the present embodiment, above-mentioned training unit 502 can use nitrification enhancement, be based on initial model, execute such as Lower training step, to learn the generation of acceleration: from target velocity set, choosing target velocity；From acceleration set, choosing Take acceleration；Target vehicle is determined in the state of being travelled according to selected acceleration, it is pre- whether target vehicle meets First determining traveling smoothness condition；Meet traveling smoothness condition in response to determining, establishes the initial velocity, selected of target vehicle Target velocity and selected acceleration three between corresponding relationship；Determine whether to meet preset end training condition； Terminate training condition in response to determining to meet, generates the running model at least one corresponding relationship that characterization is established.

In some optional implementations of the present embodiment, the device 500 further include: adjustment unit (not shown) It is configured in response to determine and is unsatisfactory for terminating training condition, adjust the model parameter of initial model, adjusted using model parameter Initial model afterwards, continues to execute training step.

In some optional implementations of the present embodiment, training unit 502 include: the first determining module (in figure not Show) in the state of be configured to determine target vehicle and travelled in simulated environment according to selected acceleration, target Whether vehicle meets predetermined traveling smoothness condition.

In some optional implementations of the present embodiment, training unit 502 include: the second determining module (in figure not Show) it is configured to determine target vehicle, according to selected acceleration in the state of being travelled during actual travel, Whether target vehicle meets predetermined traveling smoothness condition.

In some optional implementations of the present embodiment, first acquisition unit 501 includes: to obtain module (in figure not Show) it is configured to obtain target velocity set and acceleration set in target vehicle driving process.

Target velocity and acceleration set in some optional implementations of the present embodiment, in target velocity set In acceleration correspond；And for the target velocity in target velocity set, the corresponding acceleration of the target velocity is It obtains as follows: determining that target vehicle reaches the time of the target velocity from initial velocity；Determine the target velocity With the difference of initial velocity；By the ratio of identified difference and identified time, be determined as the target velocity it is corresponding plus Speed.

The device provided by the above embodiment of the disclosure obtains target velocity set by first acquisition unit 501 and adds Sets of speeds, wherein the acceleration in acceleration set is used to indicate target vehicle and reaches target velocity acceleration to be had, Later, training unit 502 uses nitrification enhancement, is based on initial model, following training step is executed, to learn acceleration It generates: from target velocity set, choosing target velocity；From acceleration set, acceleration is chosen；Determine target vehicle by In the state of being travelled according to selected acceleration, whether target vehicle meets predetermined traveling smoothness condition；Response Meet traveling smoothness condition in determining, establishes initial velocity, selected target velocity and the selected acceleration of target vehicle Spend the corresponding relationship between three；Determine whether to meet preset end training condition；Terminate training item in response to determining to meet Part generates the running model at least one corresponding relationship that characterization is established, without being based on proportional plus integral plus derivative controller, The control to vehicle can be realized, enrich the mode of control vehicle as a result, in addition, by based on nitrification enhancement training Obtained model helps to improve the precision of vehicle control to control the traveling of vehicle, helps to improve the safety of vehicle driving Property.

Fig. 6 is turned next to, it illustrates the streams of one embodiment of the method according to the disclosure for generating information Journey 600.The method for being used to generate information, comprising the following steps:

Step 601, the initial velocity and target velocity of target vehicle are obtained.

In the present embodiment, (such as server shown in FIG. 1, terminal are set the executing subject for generating the method for model Standby or vehicle) it can perhaps radio connection from other electronic equipments or local obtains mesh by wired connection mode Mark the initial velocity and target velocity of vehicle.

Herein, above-mentioned target vehicle can be but not limited to automobile, bus etc., and the target vehicle and Fig. 2 are implemented Vehicle described in example can be same vehicle, be also possible to different vehicle.

Above-mentioned initial velocity can be target vehicle before receiving the instruction of acceleration or deceleration, and target vehicle is had Actual speed.For each instruction, an initial velocity can be corresponding with.

Above-mentioned target velocity can be the speed that expectation reaches, alternatively, the speed that target vehicle is to be achieved.

Step 602, initial velocity and target velocity are input to running model trained in advance, obtain acceleration.

In the present embodiment, above-mentioned executing subject can input the initial velocity and target velocity that step 601 is got To running model trained in advance, acceleration is obtained.Wherein, running model is according to such as above-mentioned method for generating model What the method training of middle any embodiment obtained, acceleration is used to indicate target vehicle and reaches target velocity acceleration to be had Degree.Acceleration can serve to indicate that target vehicle reaches target velocity acceleration to be had.

It is appreciated that due to the running model in the embodiment of Fig. 2, to use the obtained model of nitrification enhancement, Thus its can to characterize the corresponding relationship between initial velocity, target velocity and acceleration,

Step 603, according to acceleration, the instruction for being used to indicate target vehicle traveling is generated.

In the present embodiment, above-mentioned executing subject can generate the finger for being used to indicate target vehicle traveling according to acceleration It enables.

As an example, above-mentioned executing subject can directly generate it is comprising acceleration, be used to indicate target vehicle traveling Instruction；The throttle of target vehicle or the scale value of brake can also be determined to according to obtained acceleration, to generate comprising upper State scale value, be used to indicate target vehicle traveling instruction.

The method provided by the above embodiment of the disclosure, by obtaining the initial velocity and target velocity of target vehicle, so Afterwards, initial velocity and target velocity are input to in advance trained running model, obtain acceleration, wherein running model be by According to as the method training of any embodiment in the above-mentioned method for generating model obtains, acceleration is used to indicate target vehicle Reach target velocity acceleration to be had, later, according to acceleration, generates the instruction for being used to indicate target vehicle traveling, from Without be based on proportional plus integral plus derivative controller, realize the control to vehicle, enrich as a result, control vehicle mode, this Outside, the essence of vehicle control is helped to improve to control the traveling of vehicle by the model obtained based on nitrification enhancement training Degree, helps to improve the safety of vehicle driving.

Referring next to Fig. 7, as the realization to method shown in above-mentioned each figure, present disclose provides one kind for generating One embodiment of the device of information, the Installation practice is corresponding with embodiment of the method shown in fig. 6, documented by following Outside feature, which can also include feature identical or corresponding with embodiment of the method shown in fig. 6.The device is specific It can be applied in various electronic equipments.

As shown in fig. 7, the present embodiment includes: that second acquisition unit 701 is configured to for generating the device 700 of information Obtain the initial velocity and target velocity of target vehicle；Input unit 702 is configured to input initial velocity and target velocity To running model trained in advance, acceleration is obtained, wherein running model is in the method according to such as above-mentioned for generating model What the method training of any embodiment obtained, acceleration is used to indicate target vehicle and reaches target velocity acceleration to be had； Generation unit 703 is configured to generate the instruction for being used to indicate target vehicle traveling according to acceleration.

It in the present embodiment, can be by wired connection side for generating the second acquisition unit 701 of the device 700 of information Formula perhaps radio connection from other electronic equipments or the local initial velocity and target velocity for obtaining target vehicle.

In the present embodiment, initial velocity and mesh that above-mentioned input unit 702 can get second acquisition unit 701 Mark speed is input to running model trained in advance, obtains acceleration.Wherein, running model is according to such as above-mentioned for generating The method training of any embodiment obtains in the method for model, and acceleration, which is used to indicate target vehicle and reaches target velocity, waits having Some acceleration.Acceleration can serve to indicate that target vehicle reaches target velocity acceleration to be had.

In the present embodiment, the acceleration that above-mentioned generation unit 703 can be obtained according to input unit 702, generation are used for Indicate the instruction of target vehicle traveling.

The device provided by the above embodiment of the disclosure obtains the initial speed of target vehicle by second acquisition unit 701 Initial velocity and target velocity are input to running model trained in advance by degree and target velocity, later, input unit 702, are obtained To acceleration, wherein running model is that method according to such as above-mentioned for generating any embodiment in the method for model is trained It arrives, acceleration is used to indicate target vehicle and reaches target velocity acceleration to be had, finally, generation unit 703 is according to adding Speed generates the instruction for being used to indicate target vehicle traveling, without being based on proportional plus integral plus derivative controller, realizes to vehicle Control, the mode of control vehicle is enriched as a result, in addition, coming by based on the obtained model of nitrification enhancement training The traveling for controlling vehicle, helps to improve the precision of vehicle control, helps to improve the safety of vehicle driving.

Below with reference to Fig. 8, it illustrates the computer systems for the electronic equipment for being suitable for being used to realize embodiment of the disclosure 800 structural schematic diagram.Electronic equipment shown in Fig. 8 is only an example, should not function to embodiment of the disclosure and Use scope brings any restrictions.

As shown in figure 8, computer system 800 includes central processing unit (CPU) 801, it can be read-only according to being stored in Program in memory (ROM) 802 or be loaded into the program in random access storage device (RAM) 803 from storage section 808 and Execute various movements appropriate and processing.In RAM 803, also it is stored with system 800 and operates required various programs and data. CPU 801, ROM 802 and RAM 803 are connected with each other by bus 804.Input/output (I/O) interface 805 is also connected to always Line 804.

I/O interface 805 is connected to lower component: the importation 806 including keyboard, mouse etc.；It is penetrated including such as cathode The output par, c 807 of spool (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.；Storage section 808 including hard disk etc.； And the communications portion 809 of the network interface card including LAN card, modem etc..Communications portion 809 via such as because The network of spy's net executes communication process.Driver 810 is also connected to I/O interface 805 as needed.Detachable media 811, such as Disk, CD, magneto-optic disk, semiconductor memory etc. are mounted on as needed on driver 810, in order to read from thereon Computer program be mounted into storage section 808 as needed.

Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium On computer program, which includes the program code for method shown in execution flow chart.In such reality It applies in example, which can be downloaded and installed from network by communications portion 809, and/or from detachable media 811 are mounted.When the computer program is executed by central processing unit (CPU) 801, limited in execution disclosed method Above-mentioned function.

It should be noted that computer-readable medium described in the disclosure can be computer-readable signal media or meter Calculation machine readable storage medium storing program for executing either the two any combination.Computer readable storage medium for example can be --- but not Be limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or any above combination.Meter The more specific example of calculation machine readable storage medium storing program for executing can include but is not limited to: have the electrical connection, just of one or more conducting wires Taking formula computer disk, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only storage Device (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device, Or above-mentioned any appropriate combination.In the disclosure, computer readable storage medium can be it is any include or storage journey The tangible medium of sequence, the program can be commanded execution system, device or device use or in connection.And at this In open, computer-readable signal media may include in a base band or as the data-signal that carrier wave a part is propagated, Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for By the use of instruction execution system, device or device or program in connection.Include on computer-readable medium Program code can transmit with any suitable medium, including but not limited to: wireless, electric wire, optical cable, RF etc. are above-mentioned Any appropriate combination.

The calculating of the operation for executing the disclosure can be write with one or more programming languages or combinations thereof Machine program code, described program design language include object-oriented programming language-such as Python, Java, Smalltalk, C++ further include conventional procedural programming language-such as " C " language or similar program design language Speech.Program code can be executed fully on the user computer, partly be executed on the user computer, as an independence Software package execute, part on the user computer part execute on the remote computer or completely in remote computer or It is executed on server.In situations involving remote computers, remote computer can pass through the network of any kind --- packet It includes local area network (LAN) or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as benefit It is connected with ISP by internet).

Flow chart and block diagram in attached drawing are illustrated according to the system of the various embodiments of the disclosure, method and computer journey The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation A part of one module, program segment or code of table, a part of the module, program segment or code include one or more use The executable instruction of the logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually It can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it to infuse Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction Combination realize.

Being described in unit involved in embodiment of the disclosure can be realized by way of software, can also be passed through The mode of hardware is realized.Described unit also can be set in the processor, for example, can be described as: a kind of processor Including first acquisition unit and training unit.In another example can be described as: a kind of processor includes second acquisition unit, input Unit and generation unit.Wherein, the title of these units does not constitute the restriction to the unit itself, example under certain conditions Such as, first acquisition unit is also described as " obtaining the unit of target velocity set and acceleration set ".

As on the other hand, the disclosure additionally provides a kind of computer-readable medium, which can be Included in electronic equipment described in above-described embodiment；It is also possible to individualism, and without in the supplying electronic equipment. Above-mentioned computer-readable medium carries one or more program, when said one or multiple programs are held by the electronic equipment When row, so that the electronic equipment: obtaining target velocity set and acceleration set, wherein the acceleration in acceleration set is used Reach target velocity acceleration to be had in instruction target vehicle；Using nitrification enhancement, it is based on initial model, is executed such as Lower training step, to learn the generation of acceleration: from target velocity set, choosing target velocity；From acceleration set, choosing Take acceleration；Target vehicle is determined in the state of being travelled according to selected acceleration, it is pre- whether target vehicle meets First determining traveling smoothness condition；Meet traveling smoothness condition in response to determining, establish selected target velocity with it is selected Acceleration between corresponding relationship；Determine whether to meet preset end training condition；Terminate training in response to determining to meet Condition generates the running model at least one corresponding relationship that characterization is established.Alternatively, making the electronic equipment: obtaining target The initial velocity and target velocity of vehicle；Initial velocity and target velocity are input to running model trained in advance, added Speed；According to acceleration, the instruction for being used to indicate target vehicle traveling is generated.

Above description is only the preferred embodiment of the disclosure and the explanation to institute's application technology principle.Those skilled in the art Member is it should be appreciated that invention scope involved in the disclosure, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic Scheme, while should also cover in the case where not departing from foregoing invention design, it is carried out by above-mentioned technical characteristic or its equivalent feature Any combination and the other technical solutions formed.Such as features described above has similar function with (but being not limited to) disclosed in the disclosure Can technical characteristic replaced mutually and the technical solution that is formed.

Claims

1. a kind of method for generating model, comprising:

Obtain target velocity set and acceleration set, wherein the acceleration in the acceleration set is used to indicate target carriage Reach target velocity acceleration to be had；

Using nitrification enhancement, it is based on initial model, executes following training step, to learn the generation of acceleration: from described In target velocity set, target velocity is chosen；From the acceleration set, acceleration is chosen；Determine that the target vehicle exists In the state of being travelled according to selected acceleration, whether the target vehicle meets the predetermined smooth item of traveling Part；Meet the traveling smoothness condition in response to determination, establishes between selected target velocity and selected acceleration Corresponding relationship；Determine whether to meet preset end training condition；Meet the end training condition in response to determining, generates table Levy the running model at least one corresponding relationship established.

2. according to the method described in claim 1, wherein, the method also includes:

It is unsatisfactory for the end training condition in response to determination, the model parameter of initial model is adjusted, is adjusted using model parameter Initial model afterwards continues to execute the training step.

3. according to the method described in claim 1, wherein, the determination target vehicle according to selected acceleration into In the state that every trade is sailed, whether the target vehicle meets predetermined traveling smoothness condition, comprising:

In the state of determining that the target vehicle is travelled in simulated environment according to selected acceleration, the target carriage Whether meet predetermined traveling smoothness condition.

4. according to the method described in claim 1, wherein, the determination target vehicle according to selected acceleration into In the state that every trade is sailed, whether the target vehicle meets predetermined traveling smoothness condition, comprising:

Determine the target vehicle, it is described according to selected acceleration in the state of being travelled during actual travel Whether target vehicle meets predetermined traveling smoothness condition.

5. target velocity and the acceleration collection according to the method described in claim 1, wherein, in the target velocity set Acceleration in conjunction corresponds；And

For the target velocity in the target velocity set, the corresponding acceleration of the target velocity is to obtain as follows :

Determine that target vehicle reaches the time of the target velocity from initial velocity；

Determine the difference of the target velocity and initial velocity；

By the ratio of identified difference and identified time, it is determined as the corresponding acceleration of the target velocity.

6. method described in one of -5 according to claim 1, wherein the traveling smoothness condition includes at least one of the following:

The average speed of the target vehicle is less than preset threshold speed；

The rate of acceleration change of the target vehicle is less than preset rate of acceleration change threshold value.

7. a kind of method for generating model, comprising:

Obtain the initial velocity and target velocity of target vehicle；

The initial velocity and the target velocity are input to running model trained in advance, obtain acceleration, wherein described Running model is obtaining according to the method training as described in one of claim 1-6, and the acceleration is used to indicate the mesh Mark vehicle reaches target velocity acceleration to be had；

According to the acceleration, the instruction for being used to indicate the target vehicle traveling is generated.

8. a kind of for generating the device of model, comprising:

First acquisition unit is configured to obtain target velocity set and acceleration set, wherein in the acceleration set Acceleration is used to indicate target vehicle and reaches target velocity acceleration to be had；

Training unit is configured to be based on initial model using nitrification enhancement, executed following training step, be added with study The generation of speed: from the target velocity set, target velocity is chosen；From the acceleration set, acceleration is chosen；Really In the state of being travelled according to selected acceleration, whether the target vehicle meets in advance really the fixed target vehicle Fixed traveling smoothness condition；Meet the traveling smoothness condition in response to determination, establish selected target velocity with it is selected Acceleration between corresponding relationship；Determine whether to meet preset end training condition；Meet the end in response to determination Training condition generates the running model at least one corresponding relationship that characterization is established.

9. device according to claim 8, wherein described device further include:

Adjustment unit is configured in response to determination and is unsatisfactory for the end training condition, adjusts the model parameter of initial model, Using model parameter initial model adjusted, the training step is continued to execute.

10. device according to claim 8, wherein the training unit includes:

First determining module is configured to determine the target vehicle and is gone in simulated environment according to selected acceleration In the state of sailing, whether the target vehicle meets predetermined traveling smoothness condition.

11. device according to claim 8, wherein the training unit includes:

Second determining module is configured to determine the target vehicle, according to selected acceleration during actual travel In the state of being travelled, whether the target vehicle meets predetermined traveling smoothness condition.

12. device according to claim 8, wherein target velocity and the acceleration in the target velocity set Acceleration in set corresponds；And

Determine the difference of the target velocity and initial velocity；

13. the device according to one of claim 8-12, wherein the traveling smoothness condition includes at least one of the following:

The average speed of the target vehicle is less than preset threshold speed；

14. a kind of for generating the device of model, comprising:

Second acquisition unit is configured to obtain the initial velocity and target velocity of target vehicle；

Input unit is configured to for the initial velocity and the target velocity being input to running model trained in advance, obtains To acceleration, wherein the running model be according to as described in one of claim 1-6 method training obtain, it is described plus Speed is used to indicate the target vehicle and reaches target velocity acceleration to be had；

Generation unit is configured to generate the instruction for being used to indicate the target vehicle traveling according to the acceleration.

15. a kind of electronic equipment, comprising:

One or more processors；

Storage device is stored thereon with one or more programs,

When one or more of programs are executed by one or more of processors, so that one or more of processors are real The now method as described in any in claim 1-7.

16. a kind of computer-readable medium, is stored thereon with computer program, wherein real when described program is executed by processor The now method as described in any in claim 1-7.