Passenger car energy recovery control method and system based on cloud platform and neural network

Abstract

The present disclosure provides a cloud platform and neural network-based passenger car energy recovery control method and system, relating to the technical field of energy-saving automobiles, the method comprising: acquiring a fixed driving route and current GPS information of a vehicle, and acquiring road condition information in front of the vehicle through a network map; selecting a certain section in front of the vehicle, calling driving data and vehicle data of the vehicle through the section, inputting the driving data and vehicle data into an online neural network for analysis, outputting learning of driving behaviors of a plurality of vehicle drivers in a driving process on a specific section, and formulating a preliminary energy recovery strategy according to the output driving behaviors; correcting the preliminary energy recovery strategy, and inputting vehicle data, driving behaviors of the driver and current vehicle operating conditions into an offline neural network after correction is completed, and outputting a finally determined energy recovery strategy. The present disclosure can improve the energy recovery rate.

Description

A method and system for bus energy recovery control based on cloud platform and neural network Technical Field

[0001] This disclosure relates to the field of energy-saving vehicle technology, specifically to a bus energy recovery control method and system based on a cloud platform and neural network. Background Technology

[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.

[0003] Bus energy recovery is a precise and intelligent operating system. Under suitable conditions, the various components work together to achieve its best efficiency with the right energy recovery scheme. In particular, braking energy recovery requires the entire motor, battery and other key components to achieve the highest efficiency.

[0004] Invention patent CN111391672A proposes an adaptive energy recovery method for pure electric vehicles. Its main technical solution is to determine whether the vehicle information meets the conditions for braking energy recovery or coasting energy recovery, and adjust the corresponding energy recovery strategy in combination with the energy recovery level parameters set by the driver. This control strategy can greatly improve the energy recovery efficiency and can, to a certain extent, take into account the different requirements of different drivers for energy recovery and driving experience.

[0005] However, the inventors discovered that the above method has certain drawbacks. The setting of energy recovery level parameters is often based solely on the driver's judgment. However, when there are unexpected events during vehicle operation, it will affect the driver's judgment and reduce the vehicle's energy recovery efficiency.

[0006] Furthermore, existing methods that consider vehicle speed and mass in energy recovery require adding external sensors to the vehicle to collect various data, which undoubtedly increases the cost of hardware installation. Summary of the Invention

[0007] To address the aforementioned issues, this disclosure proposes a bus energy recovery control method and system based on a cloud platform and neural networks. By using road condition information of the road sections traversed by the bus, an online neural network, combined with the operational behavior of multiple buses during passage, determines an initial energy recovery strategy. The cloud platform then modifies the energy recovery strategy based on road condition information, and finally, an offline neural network determines the final energy recovery strategy, thereby achieving a higher energy recovery rate.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions:

[0009] A bus energy recovery control method based on cloud platform and neural network includes:

[0010] Obtain the vehicle's fixed driving route and current GPS information, and obtain road condition information ahead of the vehicle through a network map;

[0011] Select a road segment ahead of the vehicle, retrieve the vehicle's driving data and vehicle data for that road segment, input the driver's driving data and vehicle data into an online neural network for analysis, output the driving behavior of multiple vehicle drivers during the driving process of a specific road segment, and formulate a preliminary energy recovery strategy based on the output driving behavior.

[0012] The initial energy recovery strategy is revised. After the revision is completed, vehicle data, driver behavior and current vehicle operating status are input into the offline neural network to output the final energy recovery strategy.

[0013] When the vehicle is coasting, the weight and speed of the vehicle are taken into account to determine the magnitude of the motor's regenerative torque, and then the final energy recovery strategy is determined to control the vehicle's energy recovery.

[0014] According to some embodiments, the present disclosure adopts the following technical solutions:

[0015] The bus energy recovery control system based on cloud platform and neural network includes:

[0016] The cloud platform is used to obtain the vehicle's fixed driving route and current GPS information, and to obtain the road condition information ahead of the vehicle through the network map;

[0017] The neural network module is used to select a road segment ahead of the vehicle, retrieve the driving data and vehicle data of the vehicle passing through the road segment, input the driver driving data and vehicle data into the online neural network for analysis, output the driving behavior of multiple vehicle drivers in the specific road segment, and formulate a preliminary energy recovery strategy based on the output driving behavior.

[0018] The initial energy recovery strategy is revised. After the revision is completed, vehicle data, driver behavior and current vehicle operating status are input into the offline neural network to output the final energy recovery strategy.

[0019] When the vehicle is coasting, the weight and speed of the vehicle are taken into account to determine the magnitude of the motor's regenerative torque, and then the final energy recovery strategy is determined to control the vehicle's energy recovery.

[0020] According to some embodiments, the present disclosure adopts the following technical solutions:

[0021] A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned bus energy recovery control method based on a cloud platform and neural network.

[0022] According to some embodiments, the present disclosure adopts the following technical solutions:

[0023] An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the bus energy recovery control method based on cloud platform and neural network.

[0024] Compared with the prior art, the beneficial effects of this disclosure are as follows:

[0025] This disclosure utilizes basic bus travel path information provided by a cloud platform to adjust the energy recovery strategy, reducing the impact of unexpected events on driver judgment. Given that the bus's travel trajectory is basically fixed, the cloud platform can obtain road condition information ahead of the bus for use in the energy recovery control method.

[0026] This disclosure utilizes neural networks to learn the behavioral habits of bus drivers to determine an energy recovery control method. It employs an online neural network to learn the driving behavior of multiple bus drivers along a given road segment, and an offline neural network to learn the driving habits of bus drivers to formulate the energy recovery control method. The bus weight information collected is presented as the product of the number of passengers and their average weight to roughly calculate the vehicle's mass, eliminating the need for additional bus weight sensors and saving hardware costs.

[0027] The energy recovery strategy for passenger vehicle coasting adopted in this disclosure takes into account the passenger vehicle speed and mass, which can ensure the driver's driving comfort and improve the energy recovery rate. Attached Figure Description

[0028] The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.

[0029] Figure 1 is a flowchart of the energy recovery method according to an embodiment of the present disclosure;

[0030] Figure 2 is a schematic diagram illustrating the method for determining recyclability according to an embodiment of this disclosure;

[0031] Figure 3 is a schematic diagram of the control system structure according to an embodiment of this disclosure. Detailed implementation method:

[0032] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0033] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.

[0034] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0035] Example 1

[0036] One embodiment of this disclosure provides a bus energy recovery control method based on a cloud platform and neural networks, including:

[0037] Step 1: Obtain the vehicle's fixed driving route and current GPS information, and obtain the road condition information ahead of the vehicle through a network map;

[0038] Step 2: Select a road segment ahead of the vehicle, retrieve the driving data and vehicle data of the vehicle passing through the road segment, input the driver driving data and vehicle data into the online neural network for analysis, output the driving behavior of multiple vehicle drivers in the specific road segment, and formulate a preliminary energy recovery strategy based on the output driving behavior.

[0039] The initial energy recovery strategy is revised. After the revision is completed, vehicle data, driver behavior and current vehicle operating status are input into the offline neural network to output the final energy recovery strategy.

[0040] When the vehicle is coasting, the weight and speed of the vehicle are taken into account to determine the magnitude of the motor's regenerative torque, and then the final energy recovery strategy is determined to control the vehicle's energy recovery.

[0041] As one embodiment, in step one, the vehicle's fixed driving route and current GPS information are obtained, and the road condition information ahead of the vehicle is obtained through a network map;

[0042] Buses have fixed routes or pre-planned routes that do not change much. Therefore, we can obtain the vehicle's fixed route and current location information, and combine this with a network map to obtain road condition information ahead of the vehicle.

[0043] The road condition information ahead includes the road surface, congestion, and traffic light signals. The vehicle data retrieved for this road segment includes vehicle deceleration, acceleration, and overall transit time.

[0044] Vehicle data includes: steering wheel angle, accelerator pedal position, brake pedal position, speed difference between left and right wheels, and vehicle load information.

[0045] In step two, a road segment ahead of the vehicle is selected, and driving data and vehicle data of the vehicle passing through the road segment are retrieved. The driver driving data and vehicle data are input into an online neural network for analysis. The neural network structure integrates data patterns and participates in the overall rule formulation. The average value of the above data is calculated, and the average value data is finally output. The output shows the driving behavior of multiple vehicle drivers in the process of driving on a specific road segment. A preliminary energy recovery strategy is formulated based on the output driving behavior.

[0046] Specifically, the input nodes are the data uploaded by the vehicle information collection subsystem. The input layer, hidden layer, and output layer each contain one neuron. The overall structure is represented as follows:

[0047] 1. Construct a BP network

[0048] x = [x1, x2, ..., x m ] T

[0049] h = [h1, h2, ..., h n ] T

[0050] y = [y1, y2, ..., y k ]T

[0051] Where x is the input vector, h is the hidden layer vector, and y is the output vector.

[0052] Suppose the weight vector of the i-th hidden layer neuron is w i =[w i1 ,w i2 ,w i3 ,w i4 ,w i5 ,w i6 T, paranoia is b i The output of the hidden layer can then be represented as:

[0053]

[0054] Where f is the activation function, here we use the sigmoid function.

[0055] Suppose the weight vector of the j-th output layer neuron is v j =[v j1 ,v j2 ,…,v jn ] T ;

[0056] 2. Prepare training data, normalizing both the input and output targets to the range [-1, 1]:

[0057]

[0058]

[0059] Where x j,max x j,min These are the maximum and minimum values ​​of the input features, respectively; d j,max d j,min These are the maximum and minimum values ​​of the output feature, respectively.

[0060] The calculation process for forward propagation is as follows:

[0061]

[0062] 3. Calculation error E

[0063]

[0064] 4. Backpropagation error update network parameters

[0065] Δω ij =ηδ i x ij

[0066]

[0067] Where η is the learning rate, δ i Let be the error signal of the i-th neuron.

[0068]

[0069] Where z i Let f(x)' represent the weighted input of the i-th neuron, and let f(x)' represent the derivative of the activation function.

[0070] 5. Repeat step 24 until the network converges or reaches the maximum number of iterations, and then obtain commonly used values;

[0071] After completing the above steps, the obtained data are multiplied cumulatively. The steering wheel angle and steering wheel rotation speed are used in the calculation as reciprocals. The resulting values ​​are then multiplied by the road condition information, which is derived from network data.

[0072] The final energy recovery strategy is determined by the offline neural network system. The offline neural network system performs numerical calculations on the BP network after comprehensively processing the signals transmitted from the cloud control platform and the driver's operating habits (compared with the received online neural network information; if the difference is greater than 30%, the current driver's driving behavior is adopted as the criterion for determining the vehicle's driving state). After determining the current vehicle operating status, the energy recovery strategy is determined. The energy recovery strategy under coasting conditions needs to comprehensively consider the vehicle's weight and speed, and determine the amount of motor recovery torque by looking up a table, so as to ensure the smooth operation of the bus while maximizing the energy recovery utilization rate.

[0073] The energy recovery strategy involves setting various commonly used control parameters as corresponding calculation coefficients and substituting them into the energy recovery adjustment formula. The total range of the formula coefficients is evenly divided into three levels from small to large. In the first level, no energy recovery is performed. In the second level, the second motor performs braking energy recovery. In the third level, both motors perform energy recovery.

[0074] Among them, a preliminary energy recovery strategy is formulated based on the output driving behavior. The preliminary energy recovery strategy includes three types: if there are many traffic lights ahead, the road surface is bumpy, or there are many vehicles traveling at the same time and it is impossible to ensure smooth driving, the vehicle only operates heat recovery, and the motor is used for normal power output.

[0075] If there are few traffic lights ahead, the road surface is flat, or there are few vehicles traveling at the same time, and the vehicle is able to coast, the vehicle's heat recovery and mechanical energy recovery will proceed simultaneously.

[0076] If the road ahead is a long downhill section with low traffic volume and most traffic lights are green, vehicle thermal energy recovery and mechanical energy recovery will proceed simultaneously.

[0077] Furthermore, the initial energy recovery strategy is revised. After the revision is completed, vehicle data, driver behavior and current vehicle operating status are input into the offline neural network to output the final energy recovery strategy.

[0078] Secondly, the cloud control platform uses the network map to judge the road conditions ahead and corrects the energy recovery strategy. The cloud control platform monitors the traffic lights, road conditions, vehicle operation status and other information of the road ahead in real time. After comprehensive processing, it makes a comprehensive judgment. If there are few traffic lights ahead, the road conditions are a long downhill and there are few vehicles, it is recorded as 1.

[0079] If the road ahead is not uphill and there are few traffic lights but a lot of vehicles, it is recorded as 0.8;

[0080] If the road ahead is not uphill but has many curves and few traffic lights but a lot of traffic, it is recorded as 0.5;

[0081] If the road ahead is not uphill, but the overall road surface is uneven, it is recorded as 0.1.

[0082] The road surface information is recorded as 0 for any uphill situation;

[0083] After the judgment is completed, the data information is transmitted to the specific bus neural network system to participate in the cumulative multiplication.

[0084] When the vehicle is coasting, the weight and speed of the vehicle are taken into account to determine the magnitude of the motor's regenerative torque, and then the final energy recovery strategy is determined to control the vehicle's energy recovery.

[0085] In the offline neural network, the final energy recovery strategy is determined by the offline neural network system. The offline neural network system performs numerical calculations of the BP network after comprehensively processing the signals transmitted by the cloud control platform and the driver's operating habits (compared with the received online neural network information; if the difference is greater than 30%, the current driver's driving behavior is adopted as the criterion for determining the vehicle's driving state). After determining the current vehicle operating status, the energy recovery strategy is determined. Among them, the energy recovery strategy in the coasting state needs to comprehensively consider the vehicle's weight and speed, and determine the magnitude of the motor's recovery torque by looking up a table, so as to ensure the smooth operation of the bus while maximizing the energy recovery utilization rate.

[0086] Specifically, the system uses real-time network maps to assess road conditions ahead and adjusts the energy recovery strategy accordingly.

[0087] The final energy recovery strategy is determined by the offline neural network system. The offline neural network system performs numerical calculations on the BP network after comprehensively processing the signals transmitted from the cloud control platform and the driver's operating habits (compared with the received online neural network information; if the difference is greater than 30%, the current driver's driving behavior is adopted as the criterion for determining the vehicle's driving state). The energy recovery strategy is determined after considering the current vehicle operating status. The energy recovery strategy under coasting conditions needs to comprehensively consider the vehicle's weight and speed, and determine the amount of motor recovery torque by looking up a table, so as to ensure the smooth operation of the bus while maximizing the energy recovery utilization rate.

[0088] The energy recovery strategy involves setting various commonly used control parameters as corresponding calculation coefficients and substituting them into the energy recovery adjustment formula. The total range of the formula coefficients is evenly divided into three levels from small to large. In the first level, no energy recovery is performed. In the second level, the second motor performs braking energy recovery. In the third level, both motors perform energy recovery.

[0089] An offline neural network retrieves specific information about the driver's driving habits, current vehicle speed, and vehicle weight to determine the energy recovery control method.

[0090] Neural networks include online neural networks and offline neural networks.

[0091] Example 2

[0092] One embodiment of this disclosure provides a bus energy recovery control system based on a cloud platform and neural networks, including:

[0093] The cloud platform is used to obtain the vehicle's fixed driving route and current GPS information, and to obtain the road condition information ahead of the vehicle through the network map;

[0094] The neural network module is used to select a road segment ahead of the vehicle, retrieve the driving data and vehicle data of the vehicle passing through the road segment, input the driver driving data and vehicle data into the online neural network for analysis, output the driving behavior of multiple vehicle drivers in the specific road segment, and formulate a preliminary energy recovery strategy based on the output driving behavior.

[0095] The initial energy recovery strategy is revised. After the revision is completed, vehicle data, driver behavior and current vehicle operating status are input into the offline neural network to output the final energy recovery strategy.

[0096] When the vehicle is coasting, the weight and speed of the vehicle are taken into account to determine the magnitude of the motor's regenerative torque, and then the final energy recovery strategy is determined to control the vehicle's energy recovery.

[0097] The cloud platform control system combines the vehicle's GPS information, online maps, and the bus's fixed driving route to provide the control system with specific information about the road conditions ahead, including the road surface, congestion, traffic light signals, etc.

[0098] The cloud platform combines the vehicle's GPS information, online maps, and the bus's fixed driving route to provide specific information on the road conditions ahead, including the road surface, congestion, and traffic light signals.

[0099] The vehicle's built-in information collection system gathers data on steering wheel angle, accelerator pedal position, brake pedal position, left and right wheel speed difference, and vehicle load information, which is then uploaded to the vehicle control unit. Vehicle load information is primarily estimated through the bus station system, by multiplying the number of passengers checking tickets by their average weight to obtain a rough estimate of the vehicle's load.

[0100] The bus uses a combination of heat recovery system and mechanical energy recovery system. When the vehicle needs the motor for power drive, the heat recovery system is the main one. During coasting or braking deceleration, the two energy recovery systems work simultaneously for battery energy storage.

[0101] Example 3

[0102] A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned bus energy recovery control method based on a cloud platform and neural network.

[0103] Example 4

[0104] An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the bus energy recovery control method based on cloud platform and neural network.

[0105] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more flowchart illustrations and / or one or more block diagrams.

[0106] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0107] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.

Claims

1. A bus energy recovery control method based on cloud platform and neural network, characterized in that, include: The system obtains the vehicle's fixed driving route and current GPS information, and retrieves road condition information ahead of the vehicle through a network map; the cloud platform combines the vehicle's GPS information, network map, and the bus's fixed driving route to provide specific information on the road conditions ahead. Select a road segment ahead of the vehicle, retrieve the vehicle's driving data and vehicle data for that road segment, input the driver's driving data and vehicle data into an online neural network for analysis, output the driving behavior of multiple vehicle drivers during the driving process of a specific road segment, and formulate a preliminary energy recovery strategy based on the output driving behavior. An initial energy recovery strategy is formulated based on the output driving behavior. This initial strategy includes three types: if there are many traffic lights ahead, the road surface is bumpy, or there are many vehicles traveling simultaneously, making smooth driving impossible, the vehicle only performs thermal energy recovery, while the motor provides normal power output; if there are few traffic lights ahead, the road surface is flat, or there are few vehicles traveling simultaneously, allowing the vehicle to coast, thermal energy recovery and mechanical energy recovery occur simultaneously; if the road ahead is a long downhill section with low traffic volume and most traffic lights are green, thermal energy recovery and mechanical energy recovery occur simultaneously. The initial energy recovery strategy is then corrected. After correction, vehicle data, driver behavior, and the current vehicle operating status are input into an offline neural network to output the final determined energy recovery strategy. Specifically, when the vehicle is coasting, the vehicle's weight and speed are considered to determine the motor's recovery torque, and then the final energy recovery strategy is determined to control the vehicle's energy recovery.

2. The bus energy recovery control method based on cloud platform and neural network as described in claim 1, characterized in that, The road condition information ahead includes the road surface, congestion, and traffic light signal information.

3. The bus energy recovery control method based on cloud platform and neural network as described in claim 1, characterized in that, The retrieved vehicle driving data for this section of road includes: vehicle deceleration, vehicle acceleration, and overall transit time.

4. The bus energy recovery control method based on cloud platform and neural network as described in claim 1, characterized in that, The vehicle data includes: steering wheel angle, accelerator pedal position, brake pedal position, speed difference between left and right wheels, and vehicle load information.

5. The bus energy recovery control method based on cloud platform and neural network as described in claim 1, characterized in that, The system uses real-time network maps to assess road conditions ahead and adjusts energy recovery strategies accordingly.

6. The bus energy recovery control method based on cloud platform and neural network as described in claim 1, characterized in that, The final energy recovery strategy is determined by an offline neural network. The energy recovery strategy in the coasting state needs to take into account the vehicle's weight and speed, and determine the magnitude of the motor's recovery torque by looking up a table.

7. A bus energy recovery control system based on a cloud platform and neural network, characterized in that, include: The cloud platform is used to obtain the vehicle's fixed driving route and current GPS information, and to obtain the road condition information ahead of the vehicle through the network map; The neural network module is used to select a road segment ahead of the vehicle, retrieve the driving data and vehicle data of the vehicle passing through the road segment, input the driver driving data and vehicle data into the online neural network for analysis, output the driving behavior of multiple vehicle drivers in the specific road segment, and formulate a preliminary energy recovery strategy based on the output driving behavior. An initial energy recovery strategy is formulated based on the output driving behavior. This initial strategy includes three types: if there are many traffic lights ahead, the road surface is bumpy, or there are many vehicles traveling simultaneously, making smooth driving impossible, the vehicle only performs thermal energy recovery, while the motor provides normal power output; if there are few traffic lights ahead, the road surface is flat, or there are few vehicles traveling simultaneously, allowing the vehicle to coast, thermal energy recovery and mechanical energy recovery occur simultaneously; if the road ahead is a long downhill section with low traffic volume and most traffic lights are green, thermal energy recovery and mechanical energy recovery occur simultaneously. The initial energy recovery strategy is then corrected. After correction, vehicle data, driver behavior, and the current vehicle operating status are input into an offline neural network to output the final determined energy recovery strategy. Specifically, when the vehicle is coasting, the vehicle's weight and speed are considered to determine the motor's recovery torque, and then the final energy recovery strategy is determined to control the vehicle's energy recovery.

8. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the bus energy recovery control method based on a cloud platform and neural network as described in any one of claims 1-6.

9. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the bus energy recovery control method based on a cloud platform and neural network as described in any one of claims 1-6.

Images