CN115230718A - Control method and system for unmanned sweeper and readable storage medium - Google Patents
Control method and system for unmanned sweeper and readable storage medium Download PDFInfo
- Publication number
- CN115230718A CN115230718A CN202111395420.0A CN202111395420A CN115230718A CN 115230718 A CN115230718 A CN 115230718A CN 202111395420 A CN202111395420 A CN 202111395420A CN 115230718 A CN115230718 A CN 115230718A
- Authority
- CN
- China
- Prior art keywords
- sweeper
- hardware
- vehicle
- sample
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W2050/0001—Details of the control system
- B60W2050/0019—Control system elements or transfer functions
- B60W2050/0028—Mathematical models, e.g. for simulation
- B60W2050/0031—Mathematical model of the vehicle
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W2300/00—Indexing codes relating to the type of vehicle
- B60W2300/34—Compact city vehicles
Abstract
The application provides a control method and system of an unmanned sweeper and a readable storage medium. The control method of the unmanned sweeper comprises the following steps: acquiring hardware delay time and vehicle operation parameter data of the current sweeper in the driving process; inputting the vehicle operation parameter data into a trained sweeper model to output initial driving data after hardware of the current sweeper executes operation, wherein the trained sweeper model is obtained by training a sample set of the current sweeper, and the sample set comprises a hardware delay time length sample and a vehicle operation parameter data sample; and generating a hardware control instruction according to the hardware delay time and the initially determined running data, and controlling the hardware of the current sweeper to execute operation in advance of the hardware delay time. Therefore, the accuracy of hardware control of the current sweeper is improved.
Description
Technical Field
The invention relates to the technical field of unmanned vehicles, in particular to a control method and a control system of an unmanned sweeper and a readable storage medium.
Background
At present, the development current situation of the unmanned vehicle and the development demand of intelligent sanitation are integrated, and the unmanned sanitation vehicle has wide application prospect in the sanitation industry. The unmanned sweeper can replace manpower to carry out sweeping work, and the efficiency is greatly improved. The unmanned driving is mainly realized by the cooperation of a task decision module, an environment perception module, a motion planning module and a vehicle platform subsystem. The motion planning module can generate a control signal according to the current state of the vehicle, environmental information, task requirements and the constraint of the vehicle dynamic model, and controls the motion of the actual vehicle by controlling an accelerator, a brake and a steering wheel corner. In this process, it is important to select a vehicle model that matches the actual operating conditions.
Disclosure of Invention
The application provides a control method, a control system and a readable storage medium of an unmanned sweeper, and the method is beneficial to improving the accuracy of hardware control of the existing sweeper.
The application provides a control method of an unmanned sweeper, which comprises the following steps:
acquiring hardware delay time and vehicle operation parameter data of the current sweeper in the driving process;
inputting the vehicle operation parameter data into a trained sweeper model to output initial driving data after hardware of the current sweeper executes operation, wherein the trained sweeper model is obtained by training a sample set of the current sweeper, and the sample set comprises a hardware delay time length sample and a vehicle operation parameter data sample;
and generating a hardware control instruction according to the hardware delay time and the initially determined running data, and controlling the hardware of the current sweeper to carry out operation in advance of the hardware delay time.
Further, after the inputting the vehicle operation parameter data into the trained sweeper model to output the initial driving data after the hardware of the current sweeper performs the operation, the method further comprises:
training the trained sweeper model according to the vehicle operation parameter data, the hardware delay time and the initial driving data to obtain a sweeper improved model;
inputting the vehicle operation parameter data into the sweeper improvement model to output online driving data after the hardware executes operation;
generating a hardware control instruction according to the hardware delay time and the initially determined driving data, and controlling the hardware of the current sweeper to execute operation in advance of the hardware delay time, wherein the method comprises the following steps:
and generating a hardware control instruction according to the hardware delay time and the initial driving data or the online driving data, and controlling the hardware of the current sweeper to execute operation in advance of the hardware delay time.
Further, the trained sweeper model comprises a first trained sweeper model; the vehicle operation parameter data comprises a vehicle total weight, a first hardware control signal and a vehicle speed, the first hardware control signal is used for controlling first hardware in the hardware, and the first hardware comprises an accelerator and/or a brake;
the inputting of the vehicle operation parameter data into a trained sweeper model to output the initial driving data after the hardware of the current sweeper executes the operation comprises:
inputting the total vehicle weight, the first hardware control signal and the vehicle speed to the first trained sweeper model to output vehicle acceleration in the preliminary travel data after the first hardware executes the operation.
Further, before inputting the vehicle operation parameter data into the trained sweeper model to output the initial driving data after the hardware of the current sweeper performs the operation, the method further comprises:
obtaining a first hardware delay time length sample and a vehicle operation parameter data sample of each sweeper in a sweeper set, wherein the vehicle operation parameter data sample comprises a weight change interval sample, a first hardware control signal sample and a vehicle speed sample;
determining a sample set according to the first hardware delay time length sample, the weight change interval sample, the first hardware control signal sample and the vehicle speed sample;
and training the sweeper model according to the sample set to obtain the first trained sweeper model, and determining the trained sweeper model according to the first trained sweeper model.
Further, the trained sweeper model comprises a second trained sweeper model; the vehicle operation parameter data comprises a vehicle total weight, a second hardware control signal and a front wheel rotation angle, wherein the second hardware control signal is used for controlling second hardware in the hardware, and the second hardware comprises a steering wheel;
the inputting the vehicle operation parameter data into the trained sweeper model to output the initial driving data after the hardware of the current sweeper is executed comprises:
and inputting the total vehicle weight, the second hardware control signal and the front wheel steering angle into the second trained sweeper model to output the front wheel steering angle in the initial driving data after the steering wheel is operated.
Further, before inputting the vehicle operation parameter data into the trained sweeper model to output the initial driving data after the hardware of the current sweeper performs the operation, the method further comprises:
obtaining a second hardware delay time length sample of each sweeper in a sweeper set and a vehicle operation parameter data sample, wherein the vehicle operation parameter data sample comprises a weight change interval sample and a second hardware control signal sample;
determining a sample set according to the weight change interval sample and the second hardware delay time sample;
and training the sweeper model according to the sample set to obtain the second trained sweeper model, and determining the trained sweeper model according to the second trained sweeper model.
Further, the method further comprises:
and discretizing the total weight of historical changes of all the sweeper trucks in the sweeper truck set by adopting equidistant value to obtain a weight change interval sample of each sweeper truck.
Further, the vehicle speed sample comprises a vehicle common speed sample;
and/or the presence of a gas in the gas,
the obtaining mode of the first hardware control signal sample comprises the following steps: discretizing throttle opening and/or brake pressure of the brake in the first hardware control signal by adopting equidistant value obtaining to obtain throttle opening samples and/or brake pressure samples;
and/or the presence of a gas in the gas,
the second hardware controls the manner of obtaining the signal samples, including: and discretizing the steering wheel corner in the second hardware control signal by adopting equidistant value to obtain steering wheel corner samples.
The present application provides a control system for an unmanned sweeper vehicle comprising one or more processors configured to implement the method of any one of the above.
The present application provides a readable storage medium having stored thereon a program which, when executed by a processor, implements a method as described in any one of the above.
In some embodiments, in the control method of the unmanned sweeper, the hardware delay time obtained in the current sweeper running process is known, and vehicle running parameter data is input into a trained sweeper model to output initial running data after hardware of the current sweeper performs operation. The trained sweeper model is obtained by training the hardware delay time length sample and the vehicle operation parameter data sample, the trained sweeper model considers the hardware delay time length sample, better accords with the self hardware condition of the actual current sweeper, outputs the initial running data more accords with the reality, controls the hardware of the current sweeper to execute operation in advance of the hardware delay time length, and is more beneficial to improving the accuracy of hardware control of the current sweeper.
Drawings
FIG. 1 is a schematic flow chart illustrating an exemplary method of controlling an unmanned sweeping vehicle according to the present disclosure;
FIG. 2 is a schematic flow chart illustrating one embodiment of the manner in which the trained sweeper model is determined at step 120 in the method for controlling the unmanned sweeper shown in FIG. 1;
FIG. 3 is a schematic flow diagram illustrating another embodiment of the manner in which the trained sweeper model is determined at step 120 in the method of controlling the unmanned sweeper shown in FIG. 1;
FIG. 4 is a schematic flow chart diagram illustrating another embodiment of the manner in which the trained sweeper model is determined at step 120 in the method for controlling the unmanned sweeper shown in FIG. 1;
FIG. 5 is a schematic flow chart diagram illustrating another embodiment of the method of controlling the unmanned sweeper vehicle of FIG. 1 after step 120;
FIG. 6 is a block diagram illustrating the modules of an embodiment of a control system for an unmanned sweeper provided herein.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with one or more embodiments of the present specification. Rather, they are merely examples of apparatus and methods consistent with certain aspects of one or more embodiments of the specification, as detailed in the claims which follow.
It should be noted that: in other embodiments, the steps of the corresponding methods are not necessarily performed in the order shown and described herein. In some other embodiments, the methods may include more or fewer steps than those described herein. Moreover, a single step described in this specification may be broken down into multiple steps for description in other embodiments; multiple steps described in this specification may be combined into a single step in other embodiments.
Fig. 1 is a schematic flowchart illustrating a control method of an unmanned sweeping vehicle according to an embodiment of the present disclosure. Referring to fig. 1, the method for controlling the unmanned sweeping vehicle of the present application includes steps 110 to 130.
And step 110, acquiring hardware delay time and vehicle operation parameter data of the current sweeper in the driving process.
The hardware delay time length may refer to a time length that a response result lags behind an operation after the current hardware of the sweeper is operated. There are various embodiments for obtaining the hardware delay time of the current sweeper during the driving process in step 110, and in some embodiments, the hardware delay time of the current sweeper during the driving process is obtained in real time. The data obtained in this way are more timely, and the accuracy of the input of the trained current sweeper model is improved. In another embodiment, the hardware delay time of the current sweeper during driving is periodically acquired.
And 120, inputting vehicle operation parameter data into a trained current sweeper model to output initial driving data after hardware of the current sweeper executes operation, wherein the trained sweeper model is obtained by training a sample set of the sweeper, and the sample set comprises a hardware delay time length sample and a vehicle operation parameter data sample.
The vehicle operation parameter data sample can refer to historical vehicle operation parameter data of the current sweeper. Wherein the historical vehicle operating parameter data may include vehicle operating parameter data acquired offline.
The predetermined driving data may refer to driving data that the current sweeper may achieve after the hardware of the current sweeper performs an operation, and the driving data may include acceleration data and/or front wheel steering data, for example.
In step 120, the trained sweeper model is obtained by off-line training. In some embodiments, a sweeper model may be trained based on offline operational data samples for each sweeper in the set of sweepers, resulting in a trained sweeper model, wherein the offline operational data samples include vehicle operational parameter data samples, including weight change interval samples. The weight change interval sample can reflect the change of the weight of each sweeper in the sweeper set during driving. The weight contains dead weight and load, and the motor sweeper can be equipped with the water tank that is used for holding water, and the load contains the total weight of water tank at least, and when there was water in the water tank, the load contained the total weight of water tank and water at least. Because the weight of water tank can slowly descend at the in-process of motor sweeper operation, this change process can exert an influence to motor sweeper final control, consequently the weight of water tank is considered in this application, can effectively deal with the quality change problem of motor sweeper in the course of the work to the hardware of more accurate control motor sweeper. The specific determination mode of the trained sweeper model has various embodiments, which are described as follows:
FIG. 2 is a flowchart illustrating an embodiment of the manner in which the trained sweeper model is determined at step 120 in the method of controlling the unmanned sweeper shown in FIG. 1. In the embodiment shown in fig. 2, step 210 to step 230 are further included before step 120 in the control method of the unmanned sweeper truck of the present application.
Wherein, the sweeper set can comprise more than two sweepers. The first hardware delay for each sweeper may be different and there may also be losses in the first hardware of each sweeper during long term use, which may include throttle and/or brake. The first hardware delay time sample can include but is not limited to an accelerator delay time, a brake delay time, a delay time for switching the accelerator to the brake and/or a delay time for switching the brake to the accelerator, so that the first hardware delay time can be mastered, and the first hardware can be operated in advance according to the first hardware delay time before the acceleration required to be reached, so as to improve the accuracy of the control of the current sweeper. The first hardware control signal may reflect data that the first hardware of the vehicle is controlled. The first hardware control signal may include an accelerator opening magnitude and/or a brake pressure magnitude of the brake.
The vehicle speed samples may include vehicle common speed samples, and the common speed samples may refer to speed samples in which the frequency of use in the speed interval is higher than the frequency of use in other speeds. Therefore, the actual use condition of the vehicle can be more accurately determined by taking values in the vehicle common speed sample.
In some examples, there are various embodiments for obtaining samples of weight change intervals in samples of vehicle operating parameter data. In some embodiments, the historical change total weight of each sweeper truck in the sweeper truck set is discretized by adopting an equidistant value, and a weight change interval sample of each sweeper truck is obtained. The equidistance value is more favorable to the weight change interval sample of getting like this, and the value is more effective, more accords with the actual weight change interval for the fitting more accords with actual demand, and the fitting is more lifelike. The longer the distance of the equidistant value is, the more sparse the collected weight change interval samples are, the smaller the calculated amount is, and the lower the calculation precision is; and the shorter the distance of equidistance value is, the denser the weight change interval sample of collection is, the more the calculated amount is, and the calculation accuracy is higher. The sample of the weight change interval of each sweeper can be determined according to the requirement of a user for calculation accuracy, which is not limited herein. In other embodiments, random values may be used, and are not limited herein.
In some embodiments, the obtaining of the first hardware control signal sample is similar to the obtaining of the weight change interval sample, and compared with the obtaining of the weight change interval sample, the obtaining of the first hardware control signal sample may include an equidistant value, and discretizing the first hardware control signal of each vehicle in the sweeper set to obtain the first hardware control signal sample of each sweeper. Furthermore, the accelerator opening and/or the brake pressure of the brake in the first hardware control signal are discretized by adopting equidistant values, and each accelerator opening sample and/or brake pressure sample of the brake are obtained. The first hardware control signal sample may include a brake pressure sample and an accelerator opening sample of a brake. Specifically, the equidistance value is taken, more than half of data is taken from the upper limit and the lower limit of the throttle opening size and is used as a throttle opening sample in the first hardware control signal sample. And taking more than half of the data within the upper limit and the lower limit of the brake pressure magnitude as brake pressure samples of the brake in the first hardware control signal samples. Therefore, offline operation data in the first hardware control signal sample can be fully utilized, so that the trained sweeper model is more practical.
And 230, training the sweeper model according to the sample set to obtain a first trained sweeper model, and determining the trained sweeper model according to the first trained sweeper model. In some embodiments, this step 230 further includes determining the first trained sweeper model as the trained sweeper model. Therefore, samples related to the first hardware can be trained to obtain a trained sweeper model, so that the first hardware can be conveniently and accurately operated at a later stage. The trained sweeper models may be solidified in the controller of each sweeper in the set of sweepers before each sweeper leaves the factory. Therefore, the model can be put into use without online learning, and the time cost is low.
In connection with FIG. 2 above, the trained sweeper model includes a first trained sweeper model; the vehicle operating parameter data includes a total vehicle weight, a first hardware control signal, and a vehicle speed, the first hardware control signal being for controlling a first one of the hardware. In some embodiments, this step 120 further includes inputting the gross vehicle weight, the first hardware control signal, and the vehicle speed to a first trained sweeper model to output vehicle acceleration in the tentative driving data after the first hardware has performed the operation. Therefore, the first trained sweeper model can be used for processing data related to the vehicle acceleration, the calculation efficiency is higher, and the vehicle acceleration is obtained more efficiently. For example, the hardware delay of the current sweeper may be obtained in real time, at a known amount, such as 0.3 seconds, while the current sweeper is in use. After determining the vehicle acceleration in the initial driving data according to the step 120, the following step 130 is subsequently executed to obtain a hardware control command, which requires controlling the accelerator or the pedal 0.3 second before the vehicle acceleration is reached, so that the current sweeper avoids hardware delay and reaches the vehicle acceleration in time. Of course, the 0.3 second is merely an example, and the specific hardware delay is determined according to practical situations and is not limited herein.
FIG. 3 is a schematic flow chart illustrating another example of the manner in which the trained sweeper model is determined in step 120 of the method for controlling the unmanned sweeper shown in FIG. 1. The embodiment of fig. 3 is similar to the embodiment shown in fig. 2, and compared with the embodiment shown in fig. 2, in the embodiment shown in fig. 3, steps 310 to 330 are further included before step 120 in the control method of the unmanned sweeping vehicle of the present application.
Wherein the second hardware delay of each sweeper truck may be different, and the second hardware of each sweeper truck may be lossy during long term use, and the second hardware may include a steering wheel angle. The second hardware delay time sample can include but is not limited to a steering wheel delay time, so that the second hardware delay time can be mastered, and the second hardware can be operated in advance according to the second hardware delay time before the front wheel steering angle which needs to be reached, so as to improve the accuracy of the control of the current sweeper. The second hardware control signal may reflect data that the first hardware of the vehicle is controlled. The first hardware control signal may include a front wheel steering angle.
In some embodiments, the obtaining manner of the second hardware control signal sample is similar to the obtaining manner of the weight change interval sample, and compared with the obtaining manner of the weight change interval sample, the obtaining process of the second hardware control signal sample may include an equidistant value, and discretizing the second hardware control signal of each vehicle in the sweeper set to obtain the second hardware control signal sample of each sweeper. And further, discretizing the steering wheel rotation angle in the second hardware control signal by adopting an equidistance value to obtain each steering wheel rotation angle sample. Wherein the second hardware control signal samples may comprise steering wheel angle samples. Specifically, the equidistance value is taken, more than half of data are taken from the upper limit and the lower limit of the steering wheel angle size and serve as the steering wheel angle sample in the second hardware control signal sample, so that the offline operation data in the second hardware control signal sample can be fully utilized, and the trained sweeper model is more practical.
In some embodiments of this step 320, a driving data collection experiment is performed on each sweeper truck in the set of sweeper trucks according to the weight change interval sample and the second hardware delay duration sample, and an offline operation data sample is collected; the data samples are run off-line to form a sample set. The model includes, but is not limited to, a neural network.
In conjunction with FIG. 3 above, the trained sweeper model includes a second trained sweeper model; the vehicle operation parameter data comprises a vehicle total weight, a second hardware control signal and a front wheel rotation angle, wherein the second hardware control signal is used for controlling second hardware in the hardware. In some embodiments, step 120 further includes inputting the gross vehicle weight, the second hardware control signal, and the front wheel steering angle into a second trained sweeper model to output the front wheel steering angle in the initial driving data after the steering wheel is operated. Therefore, the second trained sweeper model can be used for processing data related to the front wheel steering angle, the calculated data amount is small, and the vehicle acceleration can be obtained quickly.
FIG. 4 is a schematic flow chart illustrating another example of the manner in which the trained sweeper model is determined at step 120 in the method for controlling the unmanned sweeper shown in FIG. 1. The embodiment of fig. 4 is similar to the embodiment shown in fig. 2 and 3, and compared with the embodiment shown in fig. 2 and 3, in the embodiment shown in fig. 4, steps 410 to 430 are further included before step 120 in the control method of the unmanned sweeper truck of the present application.
And step 430, training the sweeper model according to the sample set to obtain a third trained sweeper model, and determining the trained sweeper model according to the third trained sweeper model. In some embodiments, this step 430 further includes determining a third trained sweeper model as the trained sweeper model. Compared with the embodiment shown in fig. 2 and 3, in the embodiment of the present application, the weight change interval sample, the first hardware control signal sample, the vehicle speed sample, and the second hardware control signal sample are used, training is performed after the sample set is determined, the calculation is performed only once, and the calculated data amount is small, so that data related to the first hardware and the second hardware can be obtained through a trained sweeper model at a later stage, the first hardware and the second hardware can be conveniently controlled, and the control efficiency is improved.
And step 130, generating a hardware control instruction according to the hardware delay time and the initially determined running data, and controlling the hardware of the current sweeper to advance the hardware delay time to execute the operation. The hardware control command can be used for executing the command after the operation on the hardware of the current sweeper.
In this embodiment, during the current driving process of the sweeper truck, the acquired hardware delay time is known, and vehicle operation parameter data is input into the trained sweeper truck model to output initial driving data after the hardware of the current sweeper truck executes the operation. The trained sweeper model is obtained by training a hardware delay time sample and a vehicle operation parameter data sample, the hardware delay time sample is considered by the trained sweeper model, and the trained sweeper model better accords with the actual hardware condition of the current sweeper, meanwhile, the hardware delay time is obtained in the driving process of the current sweeper, and the actual hardware delay data is obtained, so that the actual hardware delay data more accords with the actual driving process of the current sweeper is obtained in time, the trained sweeper model can output the initial driving data more accords with the actual according to the actual hardware delay data, the hardware of the current sweeper is controlled to carry out operation in advance of the hardware delay time, and the accuracy of hardware control of the current sweeper is better improved.
Fig. 5 is a schematic flow chart illustrating another embodiment of a control method of the unmanned sweeping vehicle shown in fig. 1. The embodiment of fig. 5 is similar to the embodiment shown in fig. 1, and compared with the embodiment shown in fig. 1, the embodiment shown in fig. 5 further includes steps 130 to 150 after step 120.
And step 130, training the trained sweeper model according to the vehicle operation parameter data, the hardware delay duration and the initially determined driving data to obtain an improved sweeper model.
In some embodiments, this step 130 further includes adding the vehicle operating parameter data, the hardware delay time, and the preliminary travel data to the sample set to obtain an updated sample set; and inputting the updated sample into the trained sweeper model to train the trained sweeper model on line, so as to obtain an improved sweeper model. Therefore, the trained sweeper model can be used on line to obtain the initial driving data, the sweeper obtained through on-line training is improved to obtain the on-line driving data, and the initial driving data is compared with the on-line driving data, so that the accuracy of the trained sweeper model can be mastered. Here, the trained sweeper model is trained on line, and model parameters of the trained sweeper model can be generated through a nonlinear algorithm.
This step 130 trains the trained sweeper model according to the vehicle operating parameter data, the hardware delay duration, and the preliminary travel data to obtain an improved sweeper model, which may be referred to as an online learning process.
And 140, inputting the vehicle operation parameter data into the sweeper improved model to output the online driving data after the hardware executes the operation.
The vehicle operation parameter data may refer to data of operation parameters of the current sweeper during driving, and may also be referred to as online data. This step 140 updates the improved model of the sweeper using the online data to output the online driving data after the hardware has performed the operation. This allows for modification of the trained sweeper model and updating of model parameters of the trained sweeper model, such as hardware delay periods.
And 150, generating a hardware control instruction according to the hardware delay time and the initial driving data or the online driving data, and controlling the hardware of the current sweeper to execute operation in advance of the hardware delay time.
In some embodiments, the step 150 further includes generating a hardware control command according to the hardware delay duration and the online driving data when the initial driving data is different from the online driving data, and controlling the hardware of the current sweeper to advance the hardware delay duration to perform the operation. Therefore, the improved model of the sweeper can be used, and the actual situation of the sweeper is better met. When the initial running data is the same as the on-line running data, a hardware control instruction can be generated according to the data selected from the initial running data and the on-line running data and the hardware delay time length, and the hardware of the current sweeper is controlled to carry out operation in advance of the hardware delay time length. Because the result of the improved model of the sweeper is the same as that of the improved model of the sweeper, any data of the initial driving data and the online driving data can be used, and the actual driving condition of the current sweeper is better met. Compared with the prior art that the trained sweeper model needs to be updated offline, manual intervention inspection is needed, the workload is large, the trained sweeper model is modified and updated in the driving process of the sweeper, manual intervention is not needed, the workload efficiency is high, large-scale and automatic completion can be realized, and meanwhile, the change of the weight of the current sweeper and the change of hardware delay time can be effectively responded.
FIG. 6 is a block diagram illustrating one embodiment of a control system 500 for an unmanned sweeper vehicle as provided herein. The control system 500 for the unmanned sweeper includes one or more processors 501 for implementing the control method for the unmanned sweeper as described above.
In some embodiments, the control system 500 of the unmanned sweeper vehicle may include a readable storage medium 509, and the readable storage medium 509 may store a program that may be invoked by the processor 501, and may include a non-volatile storage medium. In some embodiments, the control system 500 of the unmanned sweeper may include a memory 508 and an interface 507. In some embodiments, the control system 500 of the unmanned sweeper vehicle may also include other hardware depending on the application.
The readable storage medium 509 of the embodiment of the present application has stored thereon a program for implementing the control method of the unmanned sweeping vehicle as described above when the program is executed by the processor 501.
This application may take the form of a computer program product embodied on one or more readable storage media 509 (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having program code embodied therein. Readable storage media 509 includes permanent and non-permanent, removable and non-removable media, and information storage may be accomplished by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of readable storage media 509 include, but are not limited to: phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium, which may be used to store information that may be accessed by a computing device.
The above description is only a preferred embodiment of the present disclosure, and should not be taken as limiting the present disclosure, and any modifications, equivalents, improvements, etc. made within the spirit and principle of the present disclosure should be included in the scope of the present disclosure.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.
Claims (10)
1. A control method of an unmanned sweeper is characterized by comprising the following steps:
acquiring hardware delay time and vehicle operation parameter data of the current sweeper in the driving process;
inputting the vehicle operation parameter data into a trained sweeper model to output initial driving data after hardware of the current sweeper executes operation, wherein the trained sweeper model is obtained by training a sample set of the current sweeper, and the sample set comprises hardware delay time length samples and vehicle operation parameter data samples;
and generating a hardware control instruction according to the hardware delay time and the initially determined running data, and controlling the hardware of the current sweeper to carry out operation in advance of the hardware delay time.
2. The method of controlling an unmanned sweeper vehicle of claim 1, wherein after the inputting of the vehicle operating parameter data to a trained sweeper vehicle model to output the initial driving data after the hardware of the current sweeper vehicle has performed an operation, the method further comprises:
training the trained sweeper model according to the vehicle operation parameter data, the hardware delay time and the initial driving data to obtain a sweeper improved model;
inputting the vehicle operation parameter data into the sweeper improvement model to output online driving data after the hardware executes operation;
generating a hardware control instruction according to the hardware delay time and the initially determined running data, and controlling the hardware of the current sweeper to carry out operation in advance of the hardware delay time, wherein the operation comprises the following steps:
and generating a hardware control instruction according to the hardware delay time and the initial running data or the online running data, and controlling the hardware of the current sweeper to execute operation in advance of the hardware delay time.
3. The control method of the unmanned sweeper truck of claim 1, wherein the trained sweeper truck model includes a first trained sweeper truck model; the vehicle operation parameter data comprises a total vehicle weight, a first hardware control signal and a vehicle speed, wherein the first hardware control signal is used for controlling first hardware in the hardware, and the first hardware comprises an accelerator and/or a brake;
the inputting the vehicle operation parameter data into the trained sweeper model to output the initial driving data after the hardware of the current sweeper is executed comprises:
inputting the total vehicle weight, the first hardware control signal and the vehicle speed to the first trained sweeper model to output vehicle acceleration in the preliminary travel data after the first hardware executes the operation.
4. The method of controlling an unmanned sweeper vehicle of claim 3, wherein prior to inputting the vehicle operating parameter data to a trained sweeper model to output initial travel data after hardware of the current sweeper vehicle has performed an operation, the method further comprises:
obtaining a first hardware delay time length sample and a vehicle operation parameter data sample of each sweeper in a sweeper set, wherein the vehicle operation parameter data sample comprises a weight change interval sample, a first hardware control signal sample and a vehicle speed sample;
determining a sample set according to the first hardware delay duration sample, the weight change interval sample, the first hardware control signal sample and the vehicle speed sample;
and training the sweeper model according to the sample set to obtain the first trained sweeper model, and determining the trained sweeper model according to the first trained sweeper model.
5. The control method of the unmanned sweeping vehicle of any one of claims 1 to 3, wherein the trained sweeping vehicle model includes a second trained sweeping vehicle model; the vehicle operation parameter data comprises a vehicle total weight, a second hardware control signal and a front wheel rotation angle, wherein the second hardware control signal is used for controlling second hardware in the hardware, and the second hardware comprises a steering wheel;
the inputting of the vehicle operation parameter data into a trained sweeper model to output the initial driving data after the hardware of the current sweeper executes the operation comprises:
and inputting the total vehicle weight, the second hardware control signal and the front wheel steering angle into the second trained sweeper model to output the front wheel steering angle in the initial driving data after the steering wheel is operated.
6. The method of controlling an unmanned sweeper vehicle of claim 5, wherein before inputting the vehicle operation parameter data to a trained sweeper vehicle model to output the initial driving data after the hardware of the current sweeper vehicle has performed the operation, the method further comprises:
obtaining a second hardware delay time length sample of each sweeper in a sweeper set and a vehicle operation parameter data sample, wherein the vehicle operation parameter data sample comprises a weight change interval sample and a second hardware control signal sample;
determining a sample set according to the weight change interval sample and the second hardware delay time length sample;
and training the sweeper model according to the sample set to obtain the second trained sweeper model, and determining the trained sweeper model according to the second trained sweeper model.
7. The control method of the unmanned sweeping vehicle of claim 6, further comprising: discretizing the total weight of historical changes of all the sweeper trucks in a sweeper truck set by adopting equidistant value obtaining to obtain a weight change interval sample of each sweeper truck;
and/or the presence of a gas in the gas,
the second hardware controls the manner of obtaining the signal samples, including: and discretizing the steering wheel corner in the second hardware control signal by adopting equidistant value to obtain steering wheel corner samples.
8. The control method of the unmanned sweeping vehicle of claim 4, wherein the vehicle speed samples include vehicle common speed samples;
and/or the presence of a gas in the gas,
the method for obtaining the first hardware control signal sample includes: and discretizing the throttle opening and/or the brake pressure of the brake in the first hardware control signal by adopting an equidistant value to obtain samples of the throttle opening and/or the brake pressure of the brake.
9. A control system for an unmanned sweeper vehicle, comprising one or more processors configured to implement a method of controlling an unmanned sweeper vehicle according to any one of claims 1 to 8.
10. A readable storage medium, characterized in that a program is stored thereon, which when executed by a processor, implements the control method of the unmanned sweeping vehicle of any one of claims 1 to 8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111395420.0A CN115230718A (en) | 2021-11-23 | 2021-11-23 | Control method and system for unmanned sweeper and readable storage medium |
| PCT/CN2022/071110 WO2023092834A1 (en) | 2021-11-23 | 2022-01-10 | Method and system for controlling unmanned sweeper, and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111395420.0A CN115230718A (en) | 2021-11-23 | 2021-11-23 | Control method and system for unmanned sweeper and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115230718A true CN115230718A (en) | 2022-10-25 |
Family
ID=83666973
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111395420.0A Pending CN115230718A (en) | 2021-11-23 | 2021-11-23 | Control method and system for unmanned sweeper and readable storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115230718A (en) |
| WO (1) | WO2023092834A1 (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10293485B2 (en) * | 2017-03-30 | 2019-05-21 | Brain Corporation | Systems and methods for robotic path planning |
| CN109778750A (en) * | 2019-03-02 | 2019-05-21 | 华东交通大学 | Intelligent unattended drives falling leave cleaning machine device people |
| US11604472B2 (en) * | 2019-06-21 | 2023-03-14 | Weride Corp | Method for controlling vehicle speed for autonomous driving, electronic device, and computer-readable storage medium |
| CN111047111B (en) * | 2019-12-27 | 2022-07-12 | 广州赛特智能科技有限公司 | Method, device and medium for dynamically distributing routes of unmanned sweeper |
| CN111708372B (en) * | 2020-08-24 | 2020-12-29 | 广州赛特智能科技有限公司 | Self-adaptive safe driving method and system for bumpy road surface of unmanned sweeper |
-
2021
- 2021-11-23 CN CN202111395420.0A patent/CN115230718A/en active Pending
-
2022
- 2022-01-10 WO PCT/CN2022/071110 patent/WO2023092834A1/en unknown
Also Published As
| Publication number | Publication date |
|---|---|
| WO2023092834A1 (en) | 2023-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103381826B (en) | Based on the self-adapting cruise control method of approximate Policy iteration | |
| Xu et al. | An automated learning-based procedure for large-scale vehicle dynamics modeling on baidu apollo platform | |
| Gáspár et al. | Robust control design for active driver assistance systems | |
| Xu et al. | Accurate and smooth speed control for an autonomous vehicle | |
| WO2023102962A1 (en) | Method for training end-to-end autonomous driving strategy | |
| CN112305938B (en) | Control model open-loop simulation verification method, device, equipment and medium | |
| WO2018173933A1 (en) | Information processing device, travel data processing method, vehicle, and program recording medium | |
| CN115280322A (en) | Hidden state planning actor control using learning | |
| Joševski et al. | Tube-based MPC for the energy management of hybrid electric vehicles with non-parametric driving profile prediction | |
| CN115496201A (en) | Train accurate parking control method based on deep reinforcement learning | |
| Wi et al. | Model predictive longitudinal control for heavy-duty vehicle platoon using lead vehicle pedal information | |
| CN113015981A (en) | System and method for efficient, continuous and safe learning using first principles and constraints | |
| CN116108717A (en) | Traffic transportation equipment operation prediction method and device based on digital twin | |
| CN115230718A (en) | Control method and system for unmanned sweeper and readable storage medium | |
| Qiu | A design of automobile cruise control system based on fuzzy PID | |
| Guo et al. | Modeling, learning and prediction of longitudinal behaviors of human-driven vehicles by incorporating internal human DecisionMaking process using inverse model predictive control | |
| CN112158196B (en) | Automatic parking method and device | |
| CN115454082A (en) | Vehicle obstacle avoidance method and system, computer readable storage medium and electronic device | |
| Da Rocha et al. | Model predictive control of a heavy-duty truck based on Gaussian process | |
| CN115236997A (en) | Prediction control method, device, equipment and computer readable storage medium | |
| CN116674594B (en) | Longitudinal control method and device based on path planning | |
| CN114475575B (en) | Automobile control system and method and automobile | |
| CN113792410B (en) | Method and system for mapping vehicle control data to simulation environment | |
| CN114114929B (en) | Unmanned vehicle path tracking method based on LSSVM | |
| CN114179835B (en) | Automatic driving vehicle decision training method based on reinforcement learning in real scene |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |