CN119362962B - Motor drive control method for electric off-road motorcycle and electronic device - Google Patents
Motor drive control method for electric off-road motorcycle and electronic device Download PDFInfo
- Publication number
- CN119362962B CN119362962B CN202411920265.3A CN202411920265A CN119362962B CN 119362962 B CN119362962 B CN 119362962B CN 202411920265 A CN202411920265 A CN 202411920265A CN 119362962 B CN119362962 B CN 119362962B
- Authority
- CN
- China
- Prior art keywords
- road
- sub
- region
- area
- real
- 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.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P27/00—Arrangements or methods for the control of AC motors characterised by the kind of supply voltage
- H02P27/04—Arrangements or methods for the control of AC motors characterised by the kind of supply voltage using variable-frequency supply voltage, e.g. inverter or converter supply voltage
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides a motor driving control method and electronic equipment of an electric off-road motorcycle, and relates to the technical field of motor control, comprising the steps of obtaining a target off-road area of a target electric off-road motorcycle, calling a GIS (geographic information system) to extract a terrain roughness distribution map based on geographic position information of the target off-road area, and obtaining the terrain roughness distribution map of the area; the method comprises the steps of obtaining a cross-country sub-region set, configuring a sensing monitoring frequency set of the cross-country sub-region set, determining a real-time cross-country sub-region set, obtaining a real-time road feature set, and carrying out adaptive driving control on a motor of a target electric cross-country motorcycle by taking the real-time road feature set as a driving control analysis object. The motor driving control method solves the technical problems that in the prior art, the motor driving control of the electric off-road motorcycle has low response speed to the actual road condition and low control accuracy, and achieves the technical effects of improving the fitting degree of the motor driving control and the actual road condition and improving the response speed.
Description
Technical Field
The invention relates to the technical field of motor control, in particular to a motor driving control method and electronic equipment of an electric cross-country motorcycle.
Background
With the popularization of electric off-road motorcycles, the application demands of the electric off-road motorcycles in complex terrain environments are increasing. Electric dirtbikes are often required to cope with uneven, changing moistures on off road, and motor drive control systems are required to adjust output power in real time to accommodate different terrain. However, the motor control system in the prior art depends on a predetermined fixed control parameter, and cannot be effectively adjusted according to road topography in real time. Especially under the complex topography environment, the response speed of current motor control system is slower, is difficult to in real time to the ground condition of rapid change, leads to driving experience poor, and control accuracy is low, appears out of control or unstable condition easily.
Therefore, how to adjust the control parameters of the motor according to the road condition of the target driving area of the electric off-road motorcycle becomes a technical problem to be solved.
Disclosure of Invention
The application provides a motor driving control method and electronic equipment of an electric off-road motorcycle, and aims to solve the technical problems of low response speed and low control accuracy of motor driving control of the electric off-road motorcycle to actual road conditions in the prior art.
In a first aspect of the present disclosure, there is provided a motor drive control method of an electric off-road motorcycle, the method comprising:
Acquiring a target off-road area of a target electric off-road motorcycle, and calling a GIS (geographic information system) to extract a terrain roughness distribution map based on geographic position information of the target off-road area to obtain a regional terrain roughness distribution map;
Carrying out water diffusion segmentation on the regional terrain roughness distribution map, and carrying out subregion division on the target cross-country region according to the ridge lines generated in the segmentation process to obtain a cross-country subregion set;
Traversing the cross-country sub-region set to perform historical motorcycle driving out-of-control frequency analysis, and configuring a sensing monitoring frequency group set of the cross-country sub-region set according to an analysis result, wherein the cross-country sub-region and the sensing monitoring frequency group are in one-to-one correspondence;
Acquiring real-time positioning information according to a GIS positioning assembly of the target electric off-road motorcycle, performing position matching on the real-time positioning information and the off-road sub-region set, and determining a real-time off-road sub-region set;
Constructing an asynchronous extraction dual channel based on a real-time sensing monitoring frequency group corresponding to the real-time cross-country sub-region set in the sensing monitoring frequency group set, asynchronously extracting road images collected by a depth camera of the target electric cross-country motorcycle by utilizing the asynchronous extraction dual channel, and identifying road features of an extraction result by utilizing a fully-connected network layer to obtain a real-time road feature set;
and taking the real-time road feature set as a driving control analysis object to carry out adaptive driving control on the motor of the target electric cross-country motorcycle.
In a second aspect of the present disclosure, there is provided an electronic device comprising a memory storing executable instructions and a processor implementing any of the steps of the first aspect of the present disclosure when executing the executable instructions stored in the memory.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
According to the method, a target off-road area of a target electric off-road motorcycle is obtained, a GIS system is called for carrying out terrain roughness distribution map extraction based on geographic position information of the target off-road area, then the terrain roughness distribution map of the area is subjected to water flooding segmentation, the target off-road area is subjected to subarea division according to a ridge line generated in the segmentation process, an off-road subarea set is obtained, further historical motorcycle driving out-of-control frequency frequent analysis is carried out through the off-road subarea set, a sensing monitoring frequency set of the off-road subarea set is configured according to an analysis result, wherein the off-road subarea corresponds to the sensing monitoring frequency set one by one, then real-time positioning information is obtained according to a GIS positioning component of the target electric off-road motorcycle, the real-time positioning information is subjected to position matching with the off-road subarea set, a real-time subarea set is determined, an asynchronous extraction double-channel is constructed based on the real-time sensing monitoring frequency set corresponding to the real-time subarea set in the sensing monitoring frequency set, road images collected by a depth camera of the target electric off-road motorcycle are asynchronously extracted by utilizing the double-channel, road characteristics of the extraction result is identified by utilizing a full-connection network layer, and a real-time characteristic set is obtained, and then the real-time characteristic set is used as a driving control object of the electric motor of the driving motor. The motor driving control response speed is improved, the torque and the rotating speed of the motor are adaptively adjusted in real time, the electric motor car is enabled to be more stable and stable in performance in different off-road environments, the fitting degree of motor control and actual road conditions is improved, and the technical effects of driving experience and safety are improved.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
Fig. 1 is a schematic flow chart of a motor driving control method of an electric cross-country motorcycle according to an embodiment of the present application.
Fig. 2 is an internal structure diagram of an electronic device according to an embodiment of the present application.
Reference numeral illustrates a bus 300, a receiver 301, a processor 302, a transmitter 303, a memory 304, and a bus interface 305.
Detailed Description
The motor driving control method of the electric off-road motorcycle solves the technical problems that in the prior art, the motor driving control of the electric off-road motorcycle is low in response speed to actual road conditions and low in control accuracy.
Having described the basic principles of the present application, various non-limiting embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
An embodiment, as shown in fig. 1, provides a motor driving control method of an electric off-road motorcycle, wherein the method includes:
Step 100, acquiring a target off-road area of a target electric off-road motorcycle, and calling a GIS system to extract a terrain roughness distribution map based on geographic position information of the target off-road area to acquire a regional terrain roughness distribution map;
In one embodiment, the target electric dirtbike refers to an electric dirtbike that is traveling at a particular moment. The target off-road area refers to a geographical area where the target electric off-road motorcycle is expected to travel within a particular time, and typically includes a plurality of different terrain and road conditions. The geographic position information refers to geographic coordinates or regional positions of a target off-road region obtained through technologies such as a Global Positioning System (GPS), and the like, and is usually longitude and latitude coordinates. The GIS system (geographic information system) is a system for storing, analyzing and displaying geographic data, which can process geospatial information and generate various maps and analysis reports. The regional terrain roughness profile reflects the roughness profile of the surface of the target off-road region, typically involving factors such as grade, obstructions, undulations, etc.
In step S100, first, the area where the motorcycle is about to run is determined by acquiring geographical position information of a target off-road area of the target electric off-road motorcycle. The GIS system is then invoked with the location information, and the terrain roughness profile of the area is extracted by system analysis of the geographic data. The profile reflects the roughness of different terrain in the target off-road area, including the distribution of flat and rough terrain.
The data support is provided for the subsequent control decision through accurate terrain information, so that the motor driving control of the electric motorcycle under different terrains can be timely and accurately adjusted. Through the step, the system can know the terrain features of the target off-road area, and lays a foundation for the subsequent sub-area division and the dynamic adjustment of motor control parameters.
Step 200, performing water-flooding segmentation on the regional terrain roughness distribution map, and performing subregion division on the target cross-country region according to the ridge lines generated in the segmentation process to obtain a cross-country subregion set;
further, the regional terrain roughness distribution map is subjected to water diffusion segmentation, the target cross-country region is subjected to subregion division according to the ridge lines generated in the segmentation process, and a cross-country subregion set is obtained, and step S200 of the embodiment of the application further comprises:
Obtaining a terrain roughness lowest point of the terrain roughness distribution map of the area, taking the terrain roughness lowest point as a water diffusion starting point, performing simulated water injection into the terrain roughness distribution map of the area, judging whether the roughness difference between the terrain roughness distribution point and the terrain roughness lowest point is smaller than or equal to a preset roughness difference threshold value when the water surface passes through any one of the terrain roughness distribution points in the terrain roughness distribution map of the area, and if so, continuing to inject water into the terrain roughness distribution map of the area;
If not, generating an area dividing ridge line between the terrain roughness distribution point and the lowest terrain roughness point, and then continuously injecting water into the area terrain roughness distribution map until the water surface passes through the highest terrain roughness point in the area terrain roughness distribution map, stopping injecting water, and obtaining an area dividing ridge line set, wherein the area dividing ridge line rises along with rising of the water surface;
And connecting the regional division ridge line sets, and carrying out subregion division on the target cross-country region to obtain the cross-country subregion set.
In one possible embodiment, the method is used for performing flood segmentation on the regional topographic roughness distribution map through a process of simulating water surface rise, and the regional topographic roughness distribution map is segmented into a plurality of subareas. In this process, the water surface spreads from the lowest point to the periphery until a certain dividing condition is reached. The ridge lines generated during segmentation are used to define boundaries of areas of different topography. Dividing the regional terrain roughness distribution map by obtaining the cross-country sub-region set, and laying a cushion for setting the extraction frequency of road images for carrying out motor drive control on the cross-country sub-regions with different terrains.
In the embodiment of the application, the roughness difference refers to the roughness difference between two terrain roughness distribution points, and is used for judging whether to continue water surface expansion or generate a ridge line. The terrain roughness distribution points refer to each specific point in the regional terrain roughness profile, each specific point having roughness characteristics (e.g., slope, relief, etc.). The preset roughness difference threshold is a threshold set by a person skilled in the art according to experience or algorithm, when the difference between the terrain roughness distribution point and the lowest point is larger than the threshold, the water surface rising process is stopped temporarily, and a ridge line is generated to divide the area. The regional division ridge line set is used for dividing the target cross-country region into subregions, and the boundaries of all subregions are reflected.
In step S200, the terrain roughness distribution map is first subjected to flooding segmentation, and the lowest point of the terrain roughness is found from the map and used as a flooding starting point. Then, the simulated water spreads around the surface, and water is gradually injected. When the water surface passes through a certain terrain roughness distribution point, judging whether the roughness difference value between the point and the lowest point is smaller than or equal to a preset roughness difference value threshold value. If the condition is met, water is continuously injected into the regional terrain roughness distribution map, the water surface is continuously expanded, and if the condition is not met, a ridge line is generated between the two compared points, and water injection is continuously performed until the water surface passes through the highest terrain roughness point. The region-dividing ridge line set generated by the process divides different regions into a plurality of off-road sub-regions. Finally, all the generated ridge lines are connected with each other to complete the division of the target cross-country region into sub-regions, so that a cross-country region set is obtained.
According to the roughness difference of different terrains in the region, different subregions are efficiently divided, finer geographic data support is provided for subsequent monitoring and driving control, the control system is ensured to be capable of adaptively acquiring road conditions of the target electric cross-country motorcycle when the electric cross-country motorcycle runs in the different subregions, and control parameters of the motor are rapidly adjusted according to the characteristics of the different terrains.
Further, the region dividing ridge line sets are connected to each other, and the target cross-country region is divided into sub-regions to obtain the cross-country sub-region sets, and step S200 of the embodiment of the present application further includes:
connecting the regional division ridge line sets, and dividing the target cross-country region according to a connection result to obtain an initial cross-country sub-region set;
traversing the initial cross-country sub-region set to carry out sub-region area statistics to obtain an initial cross-country sub-region area set;
When the initial cross-country sub-region area set is smaller than or equal to a preset cross-country sub-region area threshold value, adding a corresponding initial cross-country sub-region into the cross-country sub-region set to be fused;
when the initial off-road sub-region area set is larger than a preset off-road sub-region area threshold, adding a corresponding initial off-road sub-region into a fusible off-road sub-region set;
And carrying out initial off-road sub-region fusion based on the sub-region positions of the off-road sub-region set to be fused and the fusible off-road sub-region set to obtain the off-road sub-region set.
Further, performing initial off-road sub-region fusion based on the sub-region positions of the off-road sub-region set to be fused and the fusible off-road sub-region set to obtain the off-road sub-region set, and step S200 of the embodiment of the present application further includes:
randomly extracting a first cross-country sub-region to be fused from the cross-country sub-region set to be fused;
According to the sub-region position, matching the neighborhood of the first cross-country sub-region to be fused from the fusible cross-country sub-region set to obtain a first fusible cross-country sub-region neighborhood;
Acquiring a first fusibility off-road sub-region neighborhood area set of the first fusibility off-road sub-region neighborhood;
Traversing and counting the distance from the neighborhood of the first fusible cross-country sub-region to the first cross-country sub-region to be fused to obtain a neighborhood distance set of the first fusible cross-country sub-region;
weighting calculation is carried out on the first fusibility cross-country sub-region neighborhood area set and the first fusibility cross-country sub-region neighborhood distance set to obtain a first fusibility cross-country sub-region neighborhood fusion coefficient set;
The first cross-country sub-region to be fused is fused into a cross-country sub-region which corresponds to the maximum value in the neighborhood fusion coefficient set of the first cross-country sub-region to be fused, and a first cross-country sub-region is obtained;
and carrying out initial cross-country sub-region fusion on the sub-region positions of the cross-country sub-region set to be fused and the cross-country sub-region set to be fused to obtain the cross-country sub-region set.
In one embodiment, the preset off-road sub-region area threshold is an area value predetermined by one skilled in the art, and is used to determine whether a sub-region needs further processing (such as merging). Firstly, the regional division ridge line sets are connected with each other, the target off-road region is divided according to the connection result, and an initial off-road sub-region set is obtained, namely a plurality of sub-regions which are preliminarily divided according to a terrain roughness distribution map. Next, these initial off-road sub-regions are traversed, the area of each initial off-road sub-region is counted and compared to a preset off-road sub-region area threshold. If the area of a certain subarea is smaller than or equal to a preset cross-country subarea area threshold value, the subarea is added to a cross-country subarea set to be fused, and if the area is larger than the preset cross-country subarea area threshold value, the area of the initial cross-country subarea is qualified and does not need to be actively combined, the subarea is added to the cross-country subarea set to be fused. And carrying out initial cross-country sub-region fusion according to the areas and sub-region positions of the cross-country sub-region set to be fused and the cross-country sub-region set to be fused, so as to obtain the cross-country sub-region set.
Optionally, randomly extracting a first cross-country sub-region to be fused from the cross-country sub-region set to be fused, and matching a neighborhood of the first cross-country sub-region to be fused from the cross-country sub-region set to be fused according to the sub-region position to obtain a first cross-country sub-region neighborhood to be fused. Acquiring a first fusibility off-road sub-region neighborhood area set of the first fusibility off-road sub-region neighborhood, and further respectively counting the distance from each fusibility off-road sub-region in the first fusibility off-road sub-region neighborhood to the first off-road sub-region to be fused, so as to acquire a first fusibility off-road sub-region neighborhood distance set.
And further, weighting and calculating the first fusibility cross-country sub-region neighborhood area set and the first fusibility cross-country sub-region neighborhood distance set according to weights preset by a person skilled in the art to obtain a first fusibility cross-country sub-region neighborhood fusion coefficient set. The first fusion cross-country sub-region neighborhood fusion coefficient set reflects the fusion quality of each fusion cross-country sub-region in the first fusion cross-country sub-region neighborhood, and the larger the coefficient is, the higher the fusion quality is. And selecting an optimal merging object according to the maximum value of the merging coefficient, merging the region to be merged with the merging object, and finally obtaining the optimized first cross-country sub-region.
And fusing the first cross-country sub-region to be fused into the cross-country sub-region which can be fused and corresponds to the maximum value in the neighborhood fusion coefficient set of the first cross-country sub-region to be fused, so as to obtain the first cross-country sub-region. Based on the same principle that the first off-road sub-region is obtained, the sub-region positions of the off-road sub-region set to be fused and the fusible off-road sub-region set are subjected to initial off-road sub-region fusion, and the off-road sub-region set is obtained.
Through refining and optimizing the preliminarily divided areas, the cross-country sub-area division is more reasonable and accurate through fusion operation, and therefore the adaptability and control accuracy of the electric motorcycle in different cross-country environments are improved.
Step 300, traversing the cross-country sub-region set to analyze the frequency of historical motorcycle driving runaway, and configuring a sensing monitoring frequency group set of the cross-country sub-region set according to an analysis result, wherein the cross-country sub-region and the sensing monitoring frequency group are in one-to-one correspondence;
Further, traversing the cross-country subarea set to analyze the frequency of the driving runaway frequency of the historical motorcycle, the step S300 of the embodiment of the present application further includes:
analyzing the historical motorcycle driving runaway frequency of the cross-country subarea set according to a preset runaway index set, and determining a subarea historical driving runaway frequency set;
Respectively comparing the historical driving runaway frequency of each subregion in the historical driving runaway frequency set of the subregion with the sum of the historical driving runaway frequency sets of the subregion to obtain a runaway coefficient set;
and multiplying the out-of-control coefficient set with a preset sensing monitoring frequency set to obtain the sensing monitoring frequency set.
In one embodiment of the application, the frequency of the driving runaway of the passing electric off-road motorcycle in the driving process is counted and analyzed in the history time of the off-road subarea collection, so that the frequency of the driving runaway of the subareas is known to be higher, and a basis is provided for subsequent monitoring and control. The sensing monitoring frequency group set reflects the frequency group condition that the target electric off-road motorcycle needs to analyze road characteristics when passing through different off-road subregions. The off-road subareas are in one-to-one correspondence with the sensing monitoring frequency groups.
Optionally, the preset out-of-control index set refers to a set of indexes preset by a person skilled in the art, and is used for measuring whether the electric off-road motorcycle is out of control in the off-road driving process. These indicators may include vehicle speed anomalies, wheel slip, under-power, steering failure, and the like. The sub-region historical driving runaway frequency set is a frequency data set of each cross-country sub-region with runaway event in the historical driving process, and the runaway frequency of different regions is reflected. The runaway coefficient set is a coefficient set obtained by calculating the ratio of the runaway frequency of each sub-area to the sum of the runaway frequencies of all sub-areas. Each runaway coefficient reflects the proportion of the risk of runaway for each sub-zone relative to the total zone. The set of preset sensing monitoring frequency groups is a set of sensor monitoring frequencies preset by a person skilled in the art.
In step S300, a historical motorcycle driving runaway frequency analysis is first performed on the set of off-road subregions, and the purpose of the analysis is to identify which subregions have a higher frequency of driving runaway through statistics on historical data. The step is to evaluate the historical runaway frequency of each sub-area according to a preset runaway index set so as to determine the sub-area historical driving runaway frequency of each sub-area and obtain the sub-area historical driving runaway frequency set. And further, calculating the ratio of the number of times of history driving runaway of each sub-area to the total number of times to obtain the runaway coefficient set. These runaway coefficients reflect the relative risk of each zone developing runaway.
Next, the greater the runaway coefficient, the higher the risk of motor drive runaway in that region, the higher the frequency of the sensing monitoring should be. Multiplying the runaway coefficient sets with preset sensing monitoring frequency set sets respectively to generate sensing monitoring frequency set sets corresponding to all the subareas one by one. Specifically, areas of high runaway coefficient (i.e., high risk areas) will correspond to higher sensor monitoring frequencies, and areas of low runaway coefficient (i.e., low risk areas) will correspond to lower monitoring frequencies.
The working frequency of the sensor is intelligently adjusted according to historical uncontrolled data, so that the monitoring of a high-risk area is more frequent and accurate, and the driving safety of the electric off-road motorcycle in a complex off-road environment is improved.
Step 400, acquiring real-time positioning information according to a GIS positioning component of the target electric off-road motorcycle, performing position matching on the real-time positioning information and the off-road sub-region set, and determining the real-time off-road sub-region set;
In one possible embodiment, the GIS positioning component refers to a Global Positioning System (GPS) or other positioning technology module equipped with a target electric dirtbike, which is capable of acquiring geographic location information of the motorcycle in real time. GIS (Geographic Information System) system is used to process and analyze these geographic information.
Firstly, geographic position information of a target electric cross-country motorcycle is acquired in real time through a GIS positioning assembly. Such location information may include parameters such as longitude, latitude, altitude, etc. of the motorcycle. Next, these real-time positioning information are position matched with the previously generated set of off-road sub-regions. The position matching process is to compare the current geographic position of the motorcycle with the boundaries of all cross-country subregions to determine the specific subregion where the motorcycle is located.
By this procedure, the system is able to determine at each instant the set of real-time off-road sub-regions in which the motorcycle is located. This information is critical because different off-road sub-regions may have different terrain features and driving requirements, and the real-time positioning information can help the system dynamically adjust motor drive control parameters (such as torque and rotational speed) of the motorcycle, ensuring stability and safety during driving.
Step 500, constructing an asynchronous extraction dual channel based on a real-time sensing monitoring frequency group corresponding to the real-time cross-country sub-region set in the sensing monitoring frequency group set, asynchronously extracting road images collected by a depth camera of the target electric cross-country motorcycle by utilizing the asynchronous extraction dual channel, and identifying road features of an extraction result by utilizing a fully-connected network layer to obtain a real-time road feature set;
In one embodiment, the real-time sensing monitoring frequency set is a set of sensor monitoring frequencies matched for real-time off-road sub-regions, which determines how frequently the system monitors the motorcycle driving road environment under real-time conditions. Asynchronous extraction dual channels are two channels that can work in parallel, but their extraction process is not completely synchronous. This way data processing can be accelerated and processing efficiency can be improved. The road image is image data collected by a depth camera of the target electric off-road motorcycle, and generally contains information such as texture, obstacles, gradient and the like of the road. The fully connected network layer is a structure of a neural network, which is generally used to process image data. In this layer, all input nodes and output nodes are fully connected, and feature learning can be performed on the extraction result through the connection. The real-time road feature set is a feature set of the road image processed by the fully-connected network layer, and the feature information is used for analyzing the current road conditions, including road surface conditions, obstacles and the like.
In step S500, an asynchronous extraction frequency of the road image collected by the depth camera is determined according to the real-time sensing monitoring frequency set corresponding to the current real-time cross-country sub-region set. The real-time sensing monitoring frequency set specifies how often the collected road images are acquired and extracted for proper control according to the topography. Then, an asynchronous extraction dual channel is constructed, namely, road image information is asynchronously extracted from two different data streams in parallel, so that delay possibly caused by synchronous processing is avoided.
These road images are captured by a depth camera of the electric off-road motorcycle, which can generate images with depth information so that the system can clearly perceive the road surface features. The image data is then analyzed and processed through the fully connected network layer to extract useful road features (e.g., rough road surfaces, obstacles, gradients, etc.). These extracted real-time road feature sets provide basis for subsequent motor drive control.
Further, based on the real-time sensing frequency group corresponding to the real-time cross-country sub-region set in the sensing frequency group set, an asynchronous extraction dual-channel is constructed, the road image collected by the depth camera of the target electric cross-country motorcycle is extracted asynchronously by using the asynchronous extraction dual-channel, and the extraction result is identified by using the full-connection network layer to obtain a real-time road feature set, and step S500 in the embodiment of the present application further includes:
extracting a first real-time sensing monitoring frequency and a second real-time sensing monitoring frequency in the real-time sensing monitoring frequency group, wherein the first real-time sensing monitoring frequency is smaller than the second real-time sensing monitoring frequency;
acquiring a plurality of sample road image sets and a plurality of sample first asynchronous extraction results obtained after the plurality of sample road image sets are extracted according to a first real-time sensing monitoring frequency, and constructing a first asynchronous extraction sub-channel;
acquiring a plurality of sample road image sets and a plurality of sample second asynchronous extraction results obtained after the plurality of sample road image sets are extracted according to a second real-time sensing monitoring frequency, and constructing a second asynchronous extraction sub-channel;
And connecting the first asynchronous extraction sub-channel and the second asynchronous extraction sub-channel in parallel to obtain the asynchronous extraction dual-channel.
Further, step S500 of the embodiment of the present application further includes:
acquiring a plurality of sample road feature sets corresponding to the plurality of sample road image sets;
And training a network layer based on the plurality of sample road feature sets, the plurality of sample first asynchronous extraction results and the plurality of sample second asynchronous extraction results until convergence, and obtaining the fully-connected network layer after training is completed.
In one embodiment of the present application, the set of real-time sensory monitoring frequencies includes a first real-time sensory monitoring frequency and a second real-time sensory monitoring frequency. The two represent different data acquisition frequencies of road images in real-time monitoring respectively, wherein the image extracted from the first real-time sensing monitoring frequency is used for monitoring the overall topography characteristics, and the image extracted from the second real-time sensing monitoring frequency is used for monitoring the instantaneous dynamic change.
And acquiring a plurality of sample road image sets and acquiring a plurality of sample first asynchronous extraction results obtained after the plurality of sample road image sets are extracted according to the first real-time sensing monitoring frequency. And training the frame constructed based on the convolutional neural network by using the back propagation algorithm by taking the plurality of sample road image sets, the first real-time sensing monitoring frequency and the first asynchronous extraction results of the plurality of samples as training data, and gradually adjusting network parameters, so that the network can extract images more accurately according to the first real-time sensing monitoring frequency until the training is converged, and a first asynchronous extraction sub-channel after the training is completed is obtained.
And acquiring a plurality of sample road image sets and a plurality of sample second asynchronous extraction results obtained after the plurality of sample road image sets are extracted according to the second real-time sensing monitoring frequency. And training the second asynchronous extraction sub-channel based on the same principle as that of obtaining the first asynchronous extraction sub-channel by taking the plurality of sample road image sets, the second real-time sensing monitoring frequency and the plurality of sample second asynchronous extraction results as training data.
Optionally, the first asynchronous extraction sub-channel is used for extracting the road image acquired by the depth camera according to a first real-time sensing monitoring frequency, so as to obtain a first asynchronous extraction result. And the second asynchronous extraction sub-channel is used for extracting the road image acquired by the depth camera according to a second real-time sensing monitoring frequency to obtain a second asynchronous extraction result. Then, the two sub-channels are combined together in a parallel connection mode, so that the whole work of asynchronously extracting the double channels is realized. The first and second asynchronous extraction results are used as inputs to the fully connected network layer, and the network can optimize its feature extraction capability and learn more accurate road features through training of a large amount of sample data. The purpose of this process is to ensure that the system is able to accurately identify and understand road characteristics based on changes in real-time environment, providing accurate information support for subsequent motor control.
In one possible embodiment, the deep learning training is performed by acquiring a plurality of sample road feature sets and a corresponding plurality of sample first asynchronous extraction results and a plurality of sample second asynchronous extraction results. The aim of network training is to optimize the weight of a fully connected network layer by utilizing the data, and improve the recognition capability of different road characteristics. The fully connected network layer is trained through a back propagation algorithm, and parameters of the fully connected network layer are adjusted gradually, so that the network can recognize road characteristics more accurately. This process continues during training until the network reaches an optimal state, i.e., converges. Through the training, the system can better extract useful information from the road image, improve the accuracy of road feature identification, and provide more accurate data support for motor drive control of the electric motorcycle.
And S600, taking the real-time road feature set as a driving control analysis object, and carrying out adaptive driving control on the motor of the target electric off-road motorcycle.
In one embodiment, motor adaptive drive control is performed using a previously acquired real-time road feature set as a drive control analysis object. Namely, according to road information (such as gradient, ground roughness, obstacles and the like) acquired in real time, the motor of the electric motorcycle is dynamically adjusted, and the torque and the rotating speed of the motor are changed. The adjustment can make the motorcycle better adapt to the current road condition, and improve the stability and safety of driving. For example, in rough terrain, the output of the motor may need to be increased to provide greater driving force, while on even roads, the output of the motor may be reduced appropriately to save energy.
In summary, the motor driving control method for the electric off-road motorcycle provided by the embodiment of the application has the following technical effects:
According to the application, the driving parameters (such as torque and rotating speed) of the motor are dynamically adjusted by acquiring the terrain roughness distribution map and the sensing data, so that the rapid response to different off-road terrains is realized, the stability and driving experience of the motorcycle under a complex environment are improved, meanwhile, the road characteristics are efficiently extracted by utilizing the deep camera and the asynchronous extraction double channels, the road conditions are accurately identified through the fully-connected neural network, the road conditions are mastered in real time, the accurate basis is provided for motor control, the historical uncontrolled frequency analysis is combined, the adaptive monitoring frequency is configured according to different off-road subareas, the efficiency and the accuracy of sensor data acquisition are improved, and the technical effects of improving the fitting degree of motor driving control and actual roads and improving the control quality are achieved.
In a second embodiment, shown in FIG. 2, which is a schematic diagram of an exemplary electronic device of the present application, the bus architecture is represented by bus 300 in FIG. 2, where bus 300 may comprise any number of interconnected buses and bridges, and where bus 300 couples together various circuitry including one or more processors, represented by processor 302, and memory, represented by memory 304. Bus 300 may also connect together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art, and therefore, will not be described any further herein. Bus interface 305 provides an interface between bus 300 and receiver 301 and transmitter 303. The receiver 301 and the transmitter 303 may be the same element, i.e. a transceiver, providing a means for communicating with various other apparatus over a transmission medium. The processor 302 is responsible for managing the bus 300 and general processing, while the memory 304 may be used to store data used by the processor 302 in performing operations.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411920265.3A CN119362962B (en) | 2024-12-25 | 2024-12-25 | Motor drive control method for electric off-road motorcycle and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411920265.3A CN119362962B (en) | 2024-12-25 | 2024-12-25 | Motor drive control method for electric off-road motorcycle and electronic device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119362962A CN119362962A (en) | 2025-01-24 |
| CN119362962B true CN119362962B (en) | 2025-05-27 |
Family
ID=94317519
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411920265.3A Active CN119362962B (en) | 2024-12-25 | 2024-12-25 | Motor drive control method for electric off-road motorcycle and electronic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119362962B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119789047B (en) * | 2025-03-11 | 2026-01-06 | 山东宜和宜美家居科技有限公司 | Adaptive Network Management Methods and Systems for Smart Homes |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111386560A (en) * | 2017-11-28 | 2020-07-07 | 奥迪股份公司 | Method for setting a fully automatic vehicle guidance function in a predefined navigation environment and motor vehicle |
| CN117333504A (en) * | 2023-12-01 | 2024-01-02 | 山东省国土空间数据和遥感技术研究院(山东省海域动态监视监测中心) | An accurate segmentation method for complex terrain remote sensing images |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156955A (en) * | 2014-08-04 | 2014-11-19 | 华中农业大学 | Method for detecting variation of high resolution remote sensing images |
| JP6940612B2 (en) * | 2016-09-14 | 2021-09-29 | ナウト, インコーポレイテッドNauto, Inc. | Near crash judgment system and method |
| CN118885981A (en) * | 2024-07-08 | 2024-11-01 | 辽宁科技学院 | A method for predicting land subsidence based on real-time parameter acquisition |
| CN118657347B (en) * | 2024-08-19 | 2024-11-05 | 山东路科公路信息咨询有限公司 | A resource scheduling method and system for highway maintenance information management |
| CN119122612A (en) * | 2024-08-21 | 2024-12-13 | 哈密鑫源矿业有限责任公司 | A smart mine control platform for mine management |
| CN118708991B (en) * | 2024-08-30 | 2024-11-01 | 四川精衡信建设工程检测有限公司 | Dam deformation monitoring method, device and equipment |
| CN118917673B (en) * | 2024-10-10 | 2025-02-11 | 浙江海数科技有限公司 | An intelligent monitoring system based on big data analysis of grassroots smart governance |
-
2024
- 2024-12-25 CN CN202411920265.3A patent/CN119362962B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111386560A (en) * | 2017-11-28 | 2020-07-07 | 奥迪股份公司 | Method for setting a fully automatic vehicle guidance function in a predefined navigation environment and motor vehicle |
| CN117333504A (en) * | 2023-12-01 | 2024-01-02 | 山东省国土空间数据和遥感技术研究院(山东省海域动态监视监测中心) | An accurate segmentation method for complex terrain remote sensing images |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119362962A (en) | 2025-01-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111192284B (en) | A vehicle laser point cloud segmentation method and system | |
| CN119362962B (en) | Motor drive control method for electric off-road motorcycle and electronic device | |
| CN112287832A (en) | High-resolution remote sensing image-based urban illegal building detection method | |
| CN108961758B (en) | Road junction widening lane detection method based on gradient lifting decision tree | |
| CN113743163A (en) | Traffic target recognition model training method, traffic target positioning method and device | |
| CN111324848A (en) | Optimization method of vehicle trajectory data for mobile lidar measurement system | |
| CN109214422B (en) | Parking data repair method, device, equipment and storage medium based on DCGAN | |
| CN112949519B (en) | Target detection method, device, equipment and storage medium | |
| CN110705342A (en) | Lane line segmentation detection method and device | |
| CN104794895A (en) | A Multi-source Traffic Information Fusion Method for Freeway | |
| CN112257772B (en) | Road increase and decrease interval segmentation method and device, electronic equipment and storage medium | |
| CN116828397B (en) | Track information acquisition method and device, electronic equipment and storage medium | |
| CN118254811A (en) | Vehicle driving intention prediction method, device, equipment and computer storage medium | |
| CN109886126A (en) | A Regional Vehicle Density Estimation Method Based on Dynamic Sampling Mechanism and RBF Neural Network | |
| CN115817529B (en) | Obstacle movement track prediction method, device, equipment and storage medium | |
| CN111694902B (en) | Difference path identification method and device, electronic equipment and computer storage medium | |
| KR102568112B1 (en) | Road slop information modeling apparatus and method based on ramer-douglas-peuker algorithm | |
| CN118942244A (en) | Traffic state reconstruction method and system based on sparse microscopic vehicle trajectory data | |
| CN117765525A (en) | A cross-modal distillation 3D target detection method and system based on monocular camera | |
| CN118068366A (en) | A method, system and device for detecting the adaptive drivable area of an intelligent wheelchair | |
| CN117709394A (en) | Vehicle track prediction model training method, multi-model migration prediction method and device | |
| CN110264725A (en) | The determination method and device of link flow | |
| CN118587710B (en) | Map element labeling method and device, electronic equipment and storage medium | |
| CN114519847B (en) | Target consistency judging method suitable for vehicle-road collaborative awareness system | |
| CN118269985B (en) | Vehicle lane change assisting method, device, storage medium and equipment |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |